Broadcom's CEO Hosts 2013 Analyst Day Conference (Transcript)

|
 |  About: Broadcom Limited (AVGO)
by: SA Transcripts

Broadcom Corporation (BRCM) 2013 Analyst Day Conference December 10, 2013 8:30 AM ET

Executives

Chris Zegarelli - Senior Director of Investor Relations

Scott A. McGregor - Chief Executive Officer, President and Director

Henry Samueli - Co-Founder, Chairman and Chief Technical Officer

Robert Americo Rango - Executive Vice President and General Manager of Mobile & Wireless Group

Daniel A. Marotta - Executive Vice President and General Manager of Broadband Communications Group

Rajiv Ramaswami - Executive Vice President and General Manager of Infrastructure & Networking Group

Eric K. Brandt - Chief Financial Officer and Executive Vice President

Analysts

Harlan Sur - JP Morgan Chase & Co, Research Division

Ross Seymore - Deutsche Bank AG, Research Division

Vivek Arya - BofA Merrill Lynch, Research Division

John W. Pitzer - Crédit Suisse AG, Research Division

Gary W. Mobley - The Benchmark Company, LLC, Research Division

Anil K. Doradla - William Blair & Company L.L.C., Research Division

Mark Lipacis - Jefferies LLC, Research Division

Blayne Curtis - Barclays Capital, Research Division

James Schneider - Goldman Sachs Group Inc., Research Division

Stacy A. Rasgon - Sanford C. Bernstein & Co., LLC., Research Division

Daniel L. Amir - Lazard Capital Markets LLC, Research Division

Christopher Rolland - FBR Capital Markets & Co., Research Division

Mona Eraiba

Joseph Moore - Morgan Stanley, Research Division

Timothy M. Arcuri - Cowen and Company, LLC, Research Division

Robert Sales

Operator

Welcome to Broadcom 2013 Analyst Day. Please welcome, Chris Zegarelli, Broadcom's Senior Director of Investor Relations.

Chris Zegarelli

Hey, good morning, everybody, and welcome to Broadcom's 2013 Analyst Day. I also want to take a minute to welcome those joining us on the webcast as well. We have a full agenda for you today, so we're looking forward to get right into the presentations.

Before we do so, I just wanted to walk through a couple of quick reminders. For your first is our cautionary statement. I'm not going to read this one to you, but there are 2 things to remember. First is that we may be making some forward-looking statements today, and those statements do carry risks and uncertainty associated with them. You can find the latest risk factors and our SEC filings in our latest 10-K, for example, and we don't take any obligation to update publicly any forward-looking statements made today. And I should remind you that Broadcom's results may differ materially from those statements.

Secondarily, we may be making some references the certain non-GAAP financial metrics today. Just be aware, you can find a full consideration in the Appendix to the books that you received at registration this morning, and you can also find a reconciliation on the Investor Relations section of our website. I also want to reminder you that you'll find some evaluation forms in the books that you received this morning. We really do encourage you to fill those out and return them to us at the conclusion of the event, out in the restriction area. As a thank you, you will receive a Broadcom power device as a thank you for filling those out. We do take your feedback very seriously, so please take the time during the event to fill out the evaluation form and return it to us.

Let me move on to the agenda, on what you should expect today. So we will be having -- we have a slightly different format for today's event than we've had in prior years. We'll start with a presentation from Scott McGregor, our President and Chief Executive Officer; followed by a presentation from Henry Samueli, who is our Co-Founder, Chairman of the Board, and Chief Technical Officer. Bob Rango will also present right after that, and update on our Mobile & Wireless business. After Bob's presentation, I'll invite all 3, Scott, Henry and Bob, back up on the stage for a combined Q&A session. After which, you'll hear from Dan Marotta, who runs our broadband business, as well as Rajiv Ramaswami, who runs our Infrastructure & Networking business. We'll follow-up with a Q&A with Dan and Rajiv after that. Eric will have the final presentation of the day. After which you will hear from -- and a Q&A session from the entire executive team, that's sitting right in front of me up here.

With that, we have a full agenda, so let's get right to it. Please join me in welcoming to the stage, Scott McGregor, our President and Chief Executive Officer.

Scott A. McGregor

Good morning, and welcome. It's really great to be here in New York, and thank you so much for providing a snowy day as a novelty for all of us in California.

Hey, we have some interesting stuff to talk about today. I'll start off talking about the strong core of our business and the different paths that we have for success going forward.

On my mind are a couple of things. One is the strength of our broadband and infrastructure businesses, and I'll talk about why they're strong and why they'll continue to be strong; how we apply that formula for success that we have in those other businesses to wireless and mobile. And then how we manage tightly through the investment cycle.

Our broadband and infrastructure businesses are truly world-class businesses. They have leading market position, a broad category of #1 market share positions. We do that by providing a complete platform, so not just pieces of a platform, but a complete platform with best-in-class technology across the board. This results in solid economics measured by profitability, measured by growth, measured by share. And we also believe they are multiple growth drivers that are going to take these businesses forward, so it's not just a time now that we have success. We believe that success will carry forward into the future. And I'll say a little more about each one of those.

In broadband, we've done a tremendous job of creating a strong end-to-end solution, and what that means is that we have devices in the central office. So for cable, there's CMTS systems; for DSL, DSLAM; for optical, OLPs, so excellent presence in each one of those. And then we've also achieved excellent presence on the client side, in the home or in the office consumer premises. And we've done that not just with modems, but building up set-top boxes from that and then filling it out with all the technologies that flow around that. For example, propagating across the home with technologies like MoCA and WiFi in that space, building up to complete media servers. And then adding additional things on top of that such as small cells that enable you to broaden into the telephony world, and Dan will say a little more about how we do that. But the key here is a strong end-to-end solutions. And in broadband, we have a leadership position in every single one of the last mile technologies. I think no other company can claim strength in one of these, let alone all of them at the same time. So a tremendous business, which gives us clear leadership in the broadband space.

In infrastructure, a similar story, excellent strength in the business here, again built by end-to-end solutions. So you can see here in the service provider solution I've shown, we have a presence in the macro base station, in the backhaul, whether it's fiber or microwave, wireless infrastructure, in the service provider networks, the core networks, the data centers, all the way back into the businesses and to the home and access. So covering all of these with technologies like switches and fiber and knowledge-based processors. A variety of #1 leading positions there, giving us a strong end-to-end solution. And we believe based on our calculations that 99.98% of all data traffic on the planet crosses a Broadcom chip. So good pervasiveness in terms of the structure.

So by creating great products like this and great solutions, we've been able to work with the leaders in the industry. Leaders like to work with leaders, and here you see the best of breed of the customers, which we count as our top customers for these businesses.

The reason we're able to attract and keep customers like this is we offer differentiation, and we differentiate in a number of different ways. In the residential gateway, we not only provide the modem and basic set-top box capability I talked about, but we provide all of the pieces that go around it. Let me say a few things about some of these. First of all, increasingly, these devices become a gateway for the home. They -- how you will access with your wireless devices, so we provide WiFi access points. Bluetooth, another way to connect there. We provide all of the pieces in the box to make the box simpler to build, such as power management. Other ways to get to the home, including MoCA and PLC. Transcoding is an important technology that allows you to take a set-top box, and take that content and view it on an iPad, or a tablet, or a phone, or other kinds of things, and we convert the format from how it comes into the house into the appropriate resolution for those devices. And then security. Very important for content security, as well as security for your own personal data. We don't believe anybody else puts together a complete solution like that for a residential gateway.

In the switch, a similar story. We not only do the basic switch capability, but we provide a lot of the things that the customers really count on for their differentiation. So for example, deep packet inspection is a way to do security in these devices, traffic manager, fabrics so you can scale into very large systems. Energy efficiency, a super important thing in the data center, so that you can get the low power, making it possible to reduce the cost of data centers going forward. So all of these are examples of how we provide a complete platform and differentiate Broadcom solution versus other customers.

So when we look at these chips, these are pretty powerful chips. The one on the left here is called the Trident II. This is the world's best data center switch. It's a pretty impressive device. It has over 4 billion transistors. It's pretty fast, 1.28 terabits per second. I think there was something on our earlier film about how many library volumes there are in the Library of Congress. And this chip, by itself, could transmit the entire contents of the Library of Congress many times in a second. The chip is quite capable, 32 lanes of 40-gig packet processing memory inside the chip. We don't believe anybody else could build a chip like this. It is the world's best data center chip, and it is in pretty much all of the data centers being built today as a result.

On the right-hand side is our latest gateway chip. This is an impressive chip. This is something that would go in a set-top box in your home, 1.8 billion transistors, very impressive level of integration. It has some new technologies like UltraHD and HEVC. UltraHD is the high-definition TV format going forward. It's equivalent to 4 HDTV screens built into 1, so 4x the bits. Because we have a new encoding technology called HEVC, you get 4x the bits for only twice the bandwidth. So some efficiency and encoding there, all built into this device. We have pretty powerful processors to run applications. This particular device has a quad-core 1.5 gigahertz ARM processor. But we also do our own processors. There's a dual VLIW DSP, very long instruction word DSP, that Broadcom designed in this box as well in this device. So this powers some of the most impressive set-top boxes you're going to see rolling out. Dan will say a little bit about the capabilities of this device. And Henry, when he gets, he's going to tell you how we built this device in multiple locations, something different than you'll see from other companies. So the world's best gateway and the world-best data center switch are examples of how we continue to win customers in this space.

As a result of doing devices like this, it gives us strong economics. On the left-hand side, here you can see that our broadband and infrastructure business together have about a 9% CAGR over the last 3 years, growing nicely in that space. And even better, we get leverage on that growth on the top line with growth on the bottom line. And you see a 12% CAGR here for the reportable operating income for these 2 segments combined. So good growth, even better growth and profitability in these businesses.

You may ask, well that's nice, how do these businesses grow in the future? What's next for these? And the answer is they've got a very bright future. In the home, we believe UltraHD and HEVC you're going to have a strong roll out over the next few years that will cause people to do a replacement cycle on their existing capital equipment, set-top boxes will get upgraded, content will get upgraded, it'll drive demand for these new devices. We have new technologies to get a bit faster to the home; technologies like G.Fast, DOCSIS 3.1, increasingly driving technology there. Small cells are an interesting opportunity to bring the cellular world combined with the world we have at home. And Dan will talk about the small cells, and how we're driving that forward. We've got a pretty impressive customer base on that so far, and I think that'll be an interesting driver for our growth.

Probably the biggest driver for our growth in the next year or 2, though, is going to be regional expansion. And Broadcom has done a tremendous job in the United States and Europe, and less so in some of the other regions. And so now getting into those other regions with custom-designed products for that is quite impressive. And Dan will share a little bit about what countries are particularly interesting there, and how we're going to drive that. And then last optical [indiscernible] growing forward.

In the data center, in the infrastructure space, we see growth in data centers. Cloud is an important driver. Many of these guys are building new, enlarged cloud data centers, and Broadcom is a strong participant in that growth. Service providers are deploying more networks now. For example, in China, they're going through multiple rounds of deployment of LTE base stations, Broadcom has content in those, and so we see that. We'll see that in multiple phases over the next few years, driving growth in that space.

For us, a new business, multicore processors in the network area. We did some acquisitions in recent years, and we're seeing good traction with customers in products in that space. Wireless infrastructure is interesting because we have an opportunity as all these base stations to deploy, not only to be in there as we are today, but to increase our content in the base stations, and we're driving that. And then lastly, the automotive market, a relatively new market for us, but one where we have a lot of products that apply, both in the wireless base, but in the infrastructure space we've taken our Ethernet products and created version applicable for automotive. And a small business in the next couple of years, but I think that one over the next 5 or 10 years, could be a new billion-dollar business for Broadcom, and Rajiv will say a little more about that.

So what we want to do is apply the success we've had in these other businesses in a similar formula to success in wireless. And there are 3 pieces of this that I'll share with you today. The first is technology leadership and complete platforms, so a similar story to what we had in terms of building complete platforms for these devices. We want to give you some clear milestones to our success in LTE, so you can see where we stand and how we're tracking in terms of that success. And then I want to talk about how we're going to continue to succeed in our connectivity business, and what drives that.

So first of all in the wireless space, we believe, again, in complete platforms. Many companies and competitors try to do just a piece of the solution. We believe you need to do the entire complete platform. So it's not just a modem, or not just a processor, it includes all the connectivity pieces, multimedia, RF, PMU and applications processors. You have to have also a good quality modem. In fact we have moved our quality of modem up to world class, and you can see a number of the features of that, Cat4, Cat6, power and space efficient. These have to be small, not only to enable small footprints in designs of the phones, but also to keep your cost low and your power low in that space. A carrier aggregation is an important technology feature going forward, as many of the carriers have fragmented spectrum bands. Envelope Tracking for efficiency, multimedia broadcast as people want to do more sports events and other kinds of things that are broadcast to phones, and Voice over LTE. These are features that you need to have for high-end phone going forward today, and we now have that capability in-house. So Broadcom, with all of these features today, has a complete LTE platform available for production now.

So why do we want to have the whole platform? It's really very simple. If we just sell connectivity in a phone, even if it's really good connectivity, we have an opportunity to get maybe $3 to $6 per device. By selling the complete platform, that includes application processors, modems, power management, et cetera, we have an opportunity to collect between $10 and $30 per device, a substantial multiplier on the opportunity for us.

So I want to talk about milestones for success, and I think this is important as you want to track our success in this business. For this year, we have some milestones. One was to close the Renesas transaction, which we did on October 1. Another one was to make sure that we had a modem that was globally certified for carriers. We've exceeded our expectations there, now having a modem certified all over the world, with actually the toughest carriers on the planet. We also want to make sure that we have multiple customer engagements under way. We have indeed succeeded with that, and I'll say a little more about that.

We've given you some 2014 milestones already, and here are a couple of new ones. We've said previously that we would have a dual-core LTE SoC in early 2014. We said we would sample a quad-core by the middle of the year, sample of leadership LTE-A, that's LTE-Advanced, fin [ph] modem by the third quarter of the year that we would tape out a variety of SoCs in the space, and that we would have strong traction and design wins with customer launches, which translates into meaningful LTE revenue. Now some of these will be easy for you to see. Meaningful LTE revenue, you'll see that in our quarterly earnings. Some of these are announcements. You'll see with products, you're going to see teardowns, other kind of things, which will help you evaluate these. But some you may not see. So for example, there's some design wins we make with customers we may not be able to announce, but they are important for us in tracking. So how do we look on these Q4 milestones? We are solidly on track, in fact we are better than solidly on track. And so today we're going to revise a few of these milestones to show how we're doing even better than we had told you previously.

So for the dual-core shipping, in early 2014 it's now early Q1. And why do we have confidence to say that? We have a customer ready to go. We have a customer who's given us production orders, so we're in the process of building these phones, and Bob will say a little bit more about that. The quad-core we had for the middle of 2014, we now have that in the first half. So we've moved that up by roughly a quarter. Sample leadership LTE Advanced, we've moved that up by roughly a quarter as well, now looking to have that in mid of next year. Why are we able to do this? Two things: One, the integration of the acquisition is going extremely well. The team is integrated, hard-working, doing well, number one; number two, our engineers are executing on track or better, and so some conservative forecast we've been able to bring in, reflecting the quality of that work. So I'm pleased to say that the milestones we've shown you previously we're doing as well, or better than what we previously told you.

Now connectivity is an important part of our business going forward as well. And we win in connectivity for 4 reasons: One is innovation and features, performance, power consumption, and footprint. So innovation and features, this is doing new things first. So for example, Broadcom led in 5G WiFi. We swept the stage pretty much for all access points and for other devices, providing the best technology out there. Bluetooth Smart, another example of innovation in the space. Performance is important, and it's not just 10% better, but in many cases we're like twice as fast as competitors. And Bob will show you an example where we're twice as fast as competitors on WiFi, and that's why they pick us versus the competitor. Power consumption is important. You see a lot of these devices out there, not only phones that want to preserve the battery as long as possible, but you have many of these new portable devices that really care about power consumption. And we believe we're 30% faster and 35% faster lower power than a lot of these other guys. And then footprint. Footprint is important because you want to have small devices. It helps you save on power, and it helps you save on cost. So these 4 things are all essential to our winning in connectivity going forward.

So our strategy for connectivity in smartphone, going forward, is straightforward. In the high-end phones for what we call hero phones or leadership phones or midtier phones, it's really those 4 things that I talked about on the previous page. It's new features which gives us differentiation, leading performance, power consumption and footprint. That's what makes people decide to use our stuff in phones. And we have a large percentage of all of the smartphones out there today, and we believe 75% of all the LTE phones out today use our connectivity for these reasons.

In the lower tier and midrange where people choose more often on the basis of price. For us it's important to get connectivity pull through with our baseband and full cellular platforms. And so our strategy there is as we win in LTE and continue to win in 3G, we drive the connectivity business across with high attach rates there. So overall, Broadcom leads in connectivity for mobile devices, and we expect to continue to do so going forward. We expect our business in connectivity to grow next year.

Now, a lot of people think that the mobile opportunity is the only thing around there. And there's a very interesting phenomena going on right now. If you look towards the end of this decade, we believe there'll be 5 billion or 6 billion smartphones and mobile devices that are connected. But even more important, there'll be 30 billion devices in total that are connected, which include a lot of new categories. They include wearables, wearables like a smart watch for example, like I'm wearing here. This is a Pebble Watch, for example. A lot of devices that will help you in fitness and health care. Google Glass will probably become very interesting as it transforms into other things. smart meters, smart appliances, toys, all kinds of things will be connected. We think this is a very large market, much larger in unit quantity than the cell phone space. Now what's interesting about that market is that it's powered by connectivity. And the things that you need in that market are lower cost, lower power, smaller footprint and unique capabilities in connectivity. So this is not a market that you're going to win with a cellular baseband, okay? Almost none of these devices will have cellular basebands in them. They're going to tether to a smartphone or to a WiFi mesh net, okay? This is also not a market you're going to win with a superfast processor in 10 nanometers, okay, because they don't need superfast processing. What they really need is excellence in connectivity. That's what makes the device win, okay? And those are the core strengths of Broadcom. So we believe this is really a good opportunity for us in the market going forward, and we intend to pursue this strongly.

I'd like to take a little bit about managing through the investment cycle. And there are 4 aspects of this: One is making sure our gross margins are strong; managing tightly our operating expenses; returning capital to shareholders; and tightly managing stock-based compensation. So I'm going to put up a little report card here on how we're doing in each of these 4 areas over the last few years.

In terms of non-GAAP product gross margin expansion. Since the first quarter of 2011 through last quarter, we're up 250 basis points. So in the phase of increasing our share in connectivity, in wireless, we've been able to grow our gross margins at the same time. In terms of non-GAAP R&D and SG&A expenses, we typically have a seasonal uplift in Q1, and then we expect after Q1 of next year to be stable or down each quarter. Basically meaning we think we've got the team we need to go build the products we have to do, and we aren't going to be adding headcount next year in any significant fashion to our businesses. That will allow us to trend down our R&D costs as the seasonal factors move out over the course of the year. And Eric will say more about that. Returning cash flow to investors is important, and we do that in 2 ways: One is dividends, okay, and then share buybacks. And since 2007, we have returned more than 100% of our U.S. cash flow, okay, in terms of either dividend or share buybacks. And this year in 2013, we have returned over 100% of our U.S. cash flow as well. So we think we're doing a good job in terms of returning our U.S. cash flow to our investors in that space. Lastly, stock-based compensation. Since 2007, that number is down 50%. And we took it down again in 2013. And I expect in 2014 it'll come down again as well. So I think we're managing tightly through the investment cycle and being responsible stewards for the shareholder value.

So conclusions. Number one, our broadband business and our infrastructure business, these are vibrant and growing businesses. They have a great future. Our cellular success is going to increase the value of the company. It'll drive economic value as we grow into that, and create value for shareholders. And thirdly, as a management team, we care about managing shareholder value, we're managing tightly through this investment cycle in the ways I showed you. Thank you very much.

Operator

Please welcome Henry Samueli, Broadcom's Co-Founder, Chairman of the Board and Chief Technical Officer.

Henry Samueli

Thank you, and good morning, and welcome to our Analyst Day. You're going to hear in the next few presentations from Bob, from Dan, from Rajiv, they're going to be talking about their individual businesses, their performance as a business, their products and technology. What I want to do here is take a little bit of a step backwards at a higher view, and discuss some of the issues ongoing in the semiconductor industry today; some of the dynamics, some of the challenges, some of the issues we have to deal with as an industry, and how Broadcom is very well positioned to deal with those.

First, let's talk about semiconductor industry dynamics. A lot of controversy these days over what's going on with Moore's Law. Is it slowing down? Is it ending? What about all the manufacturing costs associated with all these new process? How fast is it growing? And also the increased design costs. A lot of the issues that we have the deal with here. So let me go through all of these and give a little bit of our perspective on these issues.

First of all, Moore's Law and fabrication manufacturing technologies. We've lived for 40 years in a beautiful world in the semiconductor industry, with a relatively stable, simple transistor structure known as the bulk CMOS planar transistor with a polysilicon gate there. Served us well, shrunk generation after generation costs coming down. But starting in about the 40-nanometer generation we've seen complexity start to increase in each generation subsequent. First running into limitations on optical lithography, so we had to add a technology known as immersion lithography. That little layer of water between the lens and the wafer and straining the silicon. Then in the 28-nanometer generation, the polysilicon gate didn't work as well anymore, so we had to change the transistor structure with a metal gate. Then in the 20-nanometer generation, limitations on lithography, again with the wires, so had to do double patterning, adding layers and costs to the structure. Now in the 16-nanometer generation, the metal gate running out of steam, had to change the transistor structure yet again to a three-dimensional FinFET structure and going into the 10-nanometer generation optical lithography, probably going to run out of steam completely, and we'll have to switch over to extreme ultraviolet lithography, which is a very expensive manufacturing technology. So in each generation, it's stepped up in complexity and cost. So the question is, when does it end? Well, nobody really knows. But likely, it's into the next decade, into that 2020s generation, we'll get to probably 7-nanometer, maybe even 5-nanometer transistors. So there's still several generations to go. A lot of progress still to be made in the semiconductor industry. Someday it will end, we don't know exactly when, but someday because you're just running it a fundamental physical limits of a transistor. If you look at a 5-nanometer transistor, you'll have a gate length of 5 nanometer. Silicon atoms are spaced 0.5 nanometer apart. So in a 5-nanometer channel, you literally have 10 atoms of silicon. So you start losing the behavior of bulk material, hence -- that's why they were called bulk CMOS transistors, because you have bulk material behavior. Well now you're talking about atomic behavior. So eventually, you have to run out of steam, but the good news is it's still far enough away that it's not a problem we have to deal with today as designers in the semiconductor industry. We still have many generations of technology. Maybe 10 years from now, I'll have to stand up in here and figure out what the industry is going to do as Moore's Law comes to an end. But for now, plenty of runway left. But it is slowing, slowing in the sense, complexity is growing, rapidly -- cost is growing rapidly. And in particular, we're seeing a historical first here. The cost per transistor beginning to rise. Now again, this is a more controversial topic. But this particular data I'm showing here comes from a third-party analyst, IBS, but it is very consistent with the data we're getting from all of our foundry partners. We talk to all of them. They're all working on 16-, 14-nanometer, 10-nanometer. They're all providing us with cost projections going forward. And all of that data is very consistent with these third-party analyst reports, showing that 28-nanometer seems to be this global minimum of cost per transistor, and then it's slowly starting to rise beyond that. And that will have implications for our design choices going forward. Now it's not to say that these nodes are bad going into the future. They're great process nodes. You're clearly going to get higher-density transistors into the future. You're clearly going to get faster transistors going into the future. And you're clearly going to get lower power transistors going into the future. So all the performance metrics are getting better. The difference is you have to pay for it now. What a concept. It's becoming a rational industry. For the last 40 years, our semiconductor industry has literally been an irrational industry, like going into an auto dealer, getting a new car every 2 years, have a higher horsepower engine, an upgraded stereo system, better gas mileage, upgraded leather and a lower sticker price. It's irrational. Well, we have a rational business going forward. Better performance, better power, better density, everything, but you have to pay for it. So it's going to change your behavior a little bit. You're going to have to think a little bit more about how you partition the system, and I'll talk more about that.

And it's not only the manufacturing that's going up, it's the design that's going up. You're dealing with multibillion transistor chips. Scott showed you 2 examples of that are 4 billion transistor Ethernet Switch and our 2 billion transistor set-top box product. These are very, very complex devices. So if you look at it again, third-party industry projections here, also from IBS, but again, very consistent with our internal data as well, showing that in 28-nanometers, $575 [ph] million cost to design a chip. That's going up by a factor of 4 in the 16-nanometer generation to $300 million cost to design a chip. And if you're going to look at the various layers of where that cost goes, you can see the physical design aspects of the chip, relatively small. That's the red slice in the middle. Where the money is going is in the software because every one of these chips has multicore processors, DSP engines, huge amount of firmware and software to develop. The validation of the IP, the verification of the relayouts, a lot of complexity in the -- just the validation of the design because of the multibillion transistor complexities. So if you translate that into the implication on your products, well, it depends on the business here. If you look at an infrastructure business, your lifetime volumes of chips are much, much smaller than are obviously in the consumer space. You might sell a million chips lifetime of a typical infrastructure product. So if you have to amortize a $300 million design cost in 16-nanometer, that's $300 a chip cost amortized due to design, which could be larger than your manufacturing cost. So very substantial impact. So you have to have a business that can support the margin infrastructure, so that you can justify the ROI investment to develop a chip where you're spending $300 per chip on design. But that also applies even in the other businesses. Even if you look at the broadband business, maybe there's 10-plus million, 20 million units sold lifetime. So the design cost would be $30, $15 a chip, fairly substantial relative to the selling price of the chip going forward in '16 and beyond. And even in the consumer space, even if you sell 100 million units of a chip. In consumer, a chip itself sells 100 million units typically is a sub-$10 chip, but that means you'd be spending $3 a chip on design. So it doesn't matter what product you're selling, the design costs are now a significant fraction of the total costs of the device, layered on top of the manufacturing costs. So clearly it's going to impact your design partitioning decisions and choices to move forward into these advanced nodes for certain products. And again, I'll talk a lot more about that later.

All right. So besides cost implications, what are the other implications for the semiconductor industry for these dynamics that we're undergoing here in recent years? Number one, scale. Scale is critically important because you need to acquire a very wide and deep IP portfolio. Scott talked a lot about that. That's one of our strengths as a company. But where IT breadth and coverage across all your product lines is essential to develop the these complete platform solutions and complete platform solutions are essential for success in this business. Number two, as a result of the rising costs, both on the design and manufacturing side. Your process node selection and your platform integration strategy will change as Moore's Law slows. Again, it's a mindset shift. You're going to have to think twice about: Do I want to port this design to the next node? Or do I want to keep it in the current node longer? A lot of different trade-offs will occur as a result of these rising costs. And third, which really affects the design industry, and it's affecting it in a good way, a very positive way, is that you're now going to achieve your cost reductions from better design engineering rather than better process engineering. And this has been the mantra of the analog semiconductor business for decades. This is how they operate. If you look at all the analog companies out there today, they're very well-respected, very highly valued, great multiples, great margins, but they all use very old process technologies. They're not at the leading edge. They're on old nodes, fully depreciated fabs, very low variable manufacturing cost per wafer. Their value add in the system is design innovation. Well, we're not going to see that value get pulled into the SoC business as well, that design innovation because you're going to be in a process node such us 28 much longer. So the biggest complaint I get from my engineers when they talk about cost reductions, I ask them, "How come you didn't make the chip smaller?" They'd say, "Well, I didn't have time. You told me I had to run to the next process node to get the cost out of the design. Well, now finally, they can stay in a process node longer, optimize it, be innovative and create value in that respect, which is good for companies like Broadcom who focus on design innovation.

All right. So what are the Broadcom advantages when we deal with these dynamics in the industry? I'll talk about all 3 in particular: portfolio breadth, obviously; our platform solutions, clearly; and design engineering efficiency. All 3 I want touch on.

All right. Scott presented in his presentation slide that we use quite often in our corporate overview. It's actually my favorite slide in the corporate overview, making the statement that 99.98% of all Internet traffic passes through a Broadcom chip. The reason we can make a statement like that is because of the breadth of portfolio coverage of our product. It's truly unmatched in the communications semiconductor industry. In spite of the fact that you've got, I think, like 60-plus different boxes here, I had a very difficult time creating this slide because there is so much technology breadth and coverage in the company. It's hard to pick and choose what to put on the slide like this, so I had to pick the highlight areas of each of the businesses. But it's truly an incredibly deep portfolio. You can't imagine where you are in a network, whether you're at home, in the office, whether it's on the infrastructure side. Wherever you turn, there's a Broadcom chip somewhere in the network from the switches and the infrastructure, to the WiFi and baseband and connectivity products, to the set-top box broadband access. You'll hear from all of the guys about the great product technology. But this is what creates 99.98%. And on top of that, the gray box on top is also critically important to enable our products as our investment in central engineering infrastructure. We're doing our own custom cell libraries, our own custom memories, our own custom analog, our own S building blocks, our own [ph] IOCL, control every aspect of the chip design internally.

So let me give some examples. You'll hear more about that from the individual presentations, but I just want to give highlights because it's really a strong message. Networking. Here is an example of a customer platform that will be going to production shortly, a core network switch built entirely out of Broadcom silicon. Why out of Broadcom silicon, because we offer best-in-class in every one of those boxes on the previous page. It's not enough just to have a product in that space. You have to have the best-in-class product in order to be selected by your customer/partners. So here you'll see, for example, and this is a great example because it also shows the successful M&A integration of Broadcom. This product is brought together by several technologies that came in through acquisitions, as well as core developments in the company. At the very bottom, we have 4 Ethernet Switches that came from our core legacy teams. And above that, they're all attached to 4 knowledge-based processors that were developed obviously by the NetLogic team that came in through acquisition a couple of years ago. And then at the top and the middle, we have some additional Traffic Manager packet processors that came in through the Dune Networks acquisition that, I guess, we did about 4 years ago now. And then in the upper right, an XLP multicore processor from NetLogic. So you have a beautiful integration of all these technologies through the acquisitions, through Broadcom organic into a platform here that sells into the market of many thousands of dollars of silicon content in a product like this. So -- and recognize the fact that today, these chips are mostly 40- and 28-nanometer. We haven't yet gotten to the 16-nanometer generation. So these products will -- in 16-nanometer, have all the great features they like. They're going to have higher port counts because of the additional transistor densities. They're going to have lower power dissipation per port, they're going to have faster speeds per port. So all of the beautiful things you want in terms of performance you're going to get in 16-nanometer. But that product will be, unfortunately, or fortunately whatever you want to look at it, more expensive as a result of the design and manufacturing cost rises in 16 nanometers and beyond.

Mobile. Clearly, a lot of discussion in mobile about our position here, and I think we have a very strong position going forward, especially in the LTE space with the recent completion of the acquisition of Renesas. Here is an example of our LTE reference design platform; clearly, has world-class connectivity IP. I don't think there's any dispute about that. Again, you'll hear more about that from Bob. There are more chips on the back side of the board, so it's not obvious from looking at the front side where all those chips are. But we have WiFi, GPS, Bluetooth, NFC, all the connectivity on that platform. World-class power management and RF capability, and through, now world-class AP multimedia and through the recent acquisition, now world-class cellular modem IP. So all of the pieces in place for a complete LTE platform with world-class technology in each of the segments of that platform. And you, again, will hear a lot more about that from Bob.

In the broadband space, same thing, very strong platform solutions here. Here's an example of the UltraHD set-top box gateway. Scott showed you the chip at 7445, which is right there in the middle on the left-hand side, which is the UltraHD decoder with the high-efficiency video coding support, quad-core CPU, et cetera. Then on the upper right side, we have the cable modem front end, and then we have another receiver front end, as well in the middle. Then we have the integration of the WiFi, dual-band WiFi, 2.4 gigahertz, 3 by 3. And then, a 5 gigahertz, 3 by 3, on that platform. And also, integration of networking products to give you the Ethernet Switch in the upper left-hand corner. So in the Bluetooth on the lower side for remote control functions. So it's a great example of the merger of technologies from all the different business groups into a single platform that's sold in the broadband business. And silicon content here is in the high tens of dollars for a platform like this. Obviously, the volumes are much higher than in the core networking space.

And if you look at that SoC, that 7445 [ph] that Scott showed earlier and is shown here again, about a 2 billion transistor chip and 28 nanometers takes a lot of IP to assemble a chip of that complexity. That IP has to come from multiple different teams in the company. In this case, about 15 different teams around the world work in collaboration to develop a chip of this complexity. And this has been our mantra from the day we founded the company of teamwork, collaboration, IP reuse and sharing. And just to show the extent of that in the company, really, is part of our core fabric as a culture in Broadcom.

Last year alone, we had 15,000 instances of IP sharing. How do we know that? Well, we've built this database in the company that we used to manage all of our IP sharing. Every time you create a piece of IP that is to be used by another group, you have to check it into the database. And then if the group needs it, they check it out of the database. So the database can monitor all transactions that occur in the company and document all the reuse that occurs. And here are the individual transactions summarized in this chart. And you can see here literally thousands of sharing instances between the groups. So it's not just 3 independent silos designing chips. You have one cohesive, intertwined team sharing IP, reusing IP between the business groups, between central engineering with external IP providers as well.

And another very important point, as I mentioned earlier, when dealing with the slowing of Moore's Law, is the issue of design innovation because that's where you're going to extract value in the future if you're not going to get cost reductions out of process node transitions. So this is a historical perspective, but this will also apply in the future as well. But it's a great example of how design innovation can get you even better than Moore's Law. And Dan will talk a little bit more about this as well in his presentation. It's such a key critical component of our whole broadband product line. It's called our Full-Band Capture front-end.

So if you look back into our first generation of front-end tuner for a set-top box back in 350-nanometer, we designed that front-end tuner in a very traditionally way much a how cell phone's designed today with this so-called heterodyne-based mixer, filter-type architecture, RF front end, classical RF. While we've transitioned that Over the years, to a new technology entirely different called Full-Band Capture where you digitize the entire gigahertz spectrum and use digital receiver technology to tune your channels has huge advantages not only in power and area but in performance. And Dan will give an example of that in his talk. But looking at scaling you get. In a perfect world with Moore's Law, the best you usually hope for in a digital scaling environment is a factor of 2 for generation. You don't always quite get there, but best case it's a factor of 2. Here we've shown a 65% area reduction for node, not because of process transition only but because of design innovation layered on top of that. And power, typically Moore's Law gives you maybe 30% power reduction per generation. We've doubled that because of the architectural advances in innovation. And now going forward beyond 20-nanometer, which is the last one shown here, it becomes even more important because you're going to be in nodes longer, hence you're going to rely on that design innovation to get you that cost advantage.

The next point goes to platform partitioning, which is driven home by the issues I discussed with Moore's Law slowing and costs going up. If you look at how the business groups are going to look at their platforms going forward. For example, in infrastructure. you still need the legacy nodes much like the analog companies are always designing in the legacy nodes. When you have unique analog functions, whether it's power management functions where you typically need high voltage hence can't go to the very small transistor structures so you do the legacy high-voltage processes, or whether it's an esoteric RF front end, microwave or millimeter wave power amplifier, whatever, transceivers, you often will need a specialty RF process. So there's always going to be a business for those types of technologies. But the sweet spot in terms of our business will be the mainstream low-cost process, which will become 28-nanometer. And a bulk of those designs will end up in that node for a long time. In the infrastructures space, now the Ethernet transceivers, the small medium business, SMB processors and switches will likely stay at 28 for a long, long time. But on the other side of the coin, because infrastructure products are so critical in terms of die size, critical in terms of power, critical in terms of speed and performance, we have to move those to aggressive nodes as fast as we can. So Rajiv's guys are aggressively designing 16-nanometer chips as we speak today. They'll drive it to 10-nanometer very soon. They have to be in the leading-edge nodes. So the switches, the multicore processors and knowledge-based processors clearly have margin infrastructures that can support the extra manufacturing costs, the extra design costs because they need to have performance. That's a good thing. When you go to the Mobile & Wireless space, same issues with power management, front end modules and the legacy technologies. 28-nanometer will be a mainstream technology for our connectivity products, Bluetooth, WiFi, GPS, NFC will do just fine in 28 for a long time to come. Cellular RF likely as well. The modem depends on the segment of the market. There will be many segments of the market that will do extremely well with 28-nanometers. Then where are you going to push the advance nodes and pay for the extra performance? Clearly the application processors we need higher speed processors, what we think we do, so we will design them. High-performance graphics engines, clearly, need high speed there. And the modem that's a very high end. Once LTE Advanced goes into the gigabit speeds likely for power reasons, we'll need advanced nodes. So there will be partitioning, but I think the bulk of the products will stay in 28 for quite a while. In broadband, same thing. The front end stuff and the legacy. The gateways, the cable modems, the home networking, some of the mainstream video processing subsystems will stay in 28 for a while. And then you'll push the envelope no application processors and graphic subsystems, and maybe the very highest-end video subsystems or set-top boxes. So we are rethinking partitioning going forward. But that's not a bad thing, that's a good thing because again, it's all about cost optimization. You find the node that works best for your product.

So in summary, the semiconductor industry, clearly, is undergoing some interesting dynamics with Moore's Law slowing, increasing complexities, increasing costs, not a problem. We can deal with that. That's life in the industry and there's lots of good ways to approach and deal with. And Broadcom is extremely well positioned and has strengths in the areas necessary to deal with the semiconductor industry dynamics. Our scale, our portfolio breadth, second to none, our best-in-class IT, our design efficiency, our corporate culture of IP, we use all strengths that play well into this dynamic industry that we're in. So we believe that going forward, we are very well positioned to be a leader as a result of all these strengths as a company. So thank you very much.

Operator

Ladies and gentlemen, please welcome to the stage, Bob Rango.

Robert Americo Rango

Good morning. Thank you. It would be my pleasure to return to my hometown of New York to give you an update on Broadcom's Mobile & Wireless business today. A lot's happened since last year, and I'm going to go through that today. I'm going to talk about how our new products in wireless connectivity and in cellular baseband are going to create a platform leadership for Broadcom and enable us to go after the new generation of smart connected devices that are coming to the world, including smartphones and tablets.

My agenda is pretty simple. I'll spend about 35 minutes in the presentation. I'll spend just a few minutes about favorable market dynamics. I think everybody in the audience knows that the Mobile & Wireless business has got a very favorable dynamic in terms of market growth around the world, developing regions all consuming a tremendous amount of Mobile & Wireless products. Then I'll quickly go into our leadership positions and our new product in wireless connectivity. I'll show you how there are multiple avenues for us to continue to grow our wireless connectivity business. And then, of course, I'll cover our status on cellular baseband, both 3G and 4G and our new LTE chips coming into the marketplace.

Okay. So just a few minutes on market dynamics. Today, there's 1.9 billion smartphone subscriptions, and that's going to grow into 2019. It's going to triple to 5.6 billion. So a very good market dynamic, but very interesting data below that is that mobile data traffic is going to increase tenfold, okay? So while the number of smartphone users in the world goes up, the amount of data that they're consuming is also going up at a more rapid rate, a 10x increasing in the amount of data traffic, okay. And then more interestingly than that, the type of data traffic is very interesting. So 50% of that data traffic will be video, okay? So why is video important? Well it's important because a lot of people look and watch that video indoors. And when they're indoors they're using WiFi, a strength of Broadcom. And with all of this video, wireless operators are going to need a massive amount of spectrum, of licensed spectrum complemented by our unlicensed spectrum, both of those things play into Broadcom's strength in WiFi and in LTE.

Okay. And then moving forward, connectivity. We see a massive trend emerging in the Internet of Things. This is every consumer's device, every consumer's electronics device and every consumer device connected to the Internet to make it smarter, to make it easier to use and to make it more productive for the end user. What we see in here is everything like smartphones, automotive, and it's a huge opportunity. Scott mentioned 30 billion units, translate that into dollars and you come with $1.2 trillion of opportunity from the Internet of Things. And Then within the Internet of Things, there's another interesting emerging market called wearables, okay? A 10x Increase in wearables over the next 3 to 4 years. Wearables are another interesting market as Scott mentioned because they play right into Broadcom's strengths. Our connectivity solutions are very small, highly integrated and enable a lot of new use cases. And those use cases are needed to enable these very interesting wearable paradigms that are coming our way

Okay, so let me switch to connectivity. Okay, I mentioned earlier multiple avenues to growth. Of course, one of them is the Internet of Things, just connecting everything to the Internet using WiFi, using Bluetooth. Okay, that's one avenue to growth this new TAM expansion. Probably hundreds of millions of units of TAM expansion from the Internet of things. The second avenue for growth is for us to drive new features into the mid and high-end smartphones, okay? This allows us to enjoy a higher ASP and of course, the higher end of the smartphone businesses is where a lot of the growth is. And then finally, the third avenue to growth is winning with your own cellular platform. When we win with our 3G and our 4G cellular platform, it always carries Broadcom's connectivity. So as we gain share in cellular, we'll gain share in connectivity. So multiple avenues to grow, okay? All driven by our continued conductivity technology leadership.

So let's talk about that leadership. In 2013, we were able to achieve some very interesting things. LTE phones, for example, one of the most important segments of the smartphone business, growing quite rapidly as we know in the U.S. and starting to grow now in emerging nations. In fact, still a very nascent market in my opinion since in the U.S., LTE penetration's only hit 22%, and in emerging markets, only 2%. But here we are, with these 4G LTE phones from Apple, Samsung, HTC, LG and Google. Different base bands, different application processors, but one thing that they have in common, which is Broadcom connectivity. These handset vendors need and want to differentiate their phones in this space, and a lot of them do it with Broadcom's wireless connectivity technology and our combo chips. ABI Research recently came out with a report that Broadcom that has 75% share in wireless connectivity in the LTE smartphone segment. Quite a compelling number for us.

Now how do we do that? In 2013, Broadcom was first to roll out a new important WiFi standard called 802.11ac what we refer to as 5G WiFi, okay? We were the first to market, we rolled it out into the router space in a 3 by 3 MIMO configuration, 3 transmitters, 3 receivers for maximum performance and literally cleaned up in the router space. You can see the names of the router companies launched with us, including Apple, Belkin, D-Link, Linksys, Buffalo, NetGear. And then we quickly followed up with 1 by 1 combo chips and 2x2 combo chips that support 5G WiFi, okay? And we did a very good job of penetrating the smartphone space. And then in tablets and PCs, we also launched there. So a very good example of it, what I call an exemplary rollout for Broadcom to rollout this new technology. But there's more to come. This particular technology rollout is probably only 20% to 30% complete. And enjoy an ASP uplift when we sell an 802.11ac solution versus an 11n solution. So there's a lot more road to go on the deployment of 5G WiFi for Broadcom, okay?

Now Scott mentioned why do we win in connectivity? I'm going to go into a lot -- a little more detail. I know that the financial community likes quantitative data, so I want to provide a little more deep dive into the quantitative reasons on why we win with wireless connectivity, okay?

So at the top of this chart, 2x throughput performance, 2x throughput performance versus the competition, 35% lower power consumption, 35% lower die size. Now I'll admit die size is kind of hard to compare these days because some of our competitors have integrated part of their wireless connectivity solution on their cellular SoC. So we took the liberty of comparing just the RF subsystem to our RF subsystem, which we think is representative of the efficiency of the implementation, and our RF subsystem is about 35% smaller than the competition.

And then residual BOM [ph], in other words the measure of how much integration we can achieve, 35% lower than the competition. How did we achieve this, you might ask. Well, Broadcom is the only company that has integrated 2.4-gig and 5-gig PAs on our combo chips. So as PA is being integrated means our customers don't need to buy them and put them into the residual BOM [ph]. And we've done a better job in just integrating the PAs. We've actually, in some cases, integrated the switches in the front end. So a lot of integration on our analog and our RF achieved this kind of residual BOM [ph] advantage over our competition, okay?

Now complementing those -- these quantitative advantages that we have are a number of what I would refer to as qualitative advantages, okay? Broadcom open-sourced our Bluetooth and NFC stack and profiles into the Android community and now are a standard part of the KitKat release that's coming out shortly from Google. And we also open-sourced some key parts of our WiFi driver. So we're considered to be a very favorable partner on connectivity to the Google Android operating system.

And then we're one of the few companies with NFC. We have a quad-combo solution that includes WiFi, Bluetooth, FM and NFC, complemented with the discrete NFC solution that is significantly lower power than our competition. In standby mode, it's up to 10 -- 100x lower power than our competition, smaller footprint and advanced features. One of the advanced features I'd like to reference is the size of the antenna. Handset manufacturers have a big challenge fitting all the antennas for WiFi, Bluetooth, cellular, all the multiband 2G and 3G into a handset. When they put NFC into a handset, they want the antenna to be small. The size of that antenna is usually a differentiation for us, so we could support the smallest NFC antenna, okay?

And then within our wireless connectivity solutions, we implement extended features. I'll give you just a few examples, beamforming in the 802.11ac standard, an LDPC code. I'm going to show you in a moment what those kinds of things can do for us on throughput performance. And then finally, we support drivers for every ecosystem, okay?

So if this wasn't enough, I'm going to go into one more level of detail, again quantitative information about why Broadcom is usually chosen for wireless connectivity for these mid- to high-end phones, where handset makers need us to differentiate their products. Okay, here's a chart, transmit performance of Broadcom in red versus some competitor in blue. 142 megabits for Broadcom, 97 megabits for our competition, roughly a 50% performance improvement.

But most of the users, like what are you -- what some of you are doing right now in browsing the web, are spending most of their time in receive mode, receiving data from the Internet, okay? In that case, we're almost 3x advantaged over our competition, okay? 188 megabits per second versus 75 at an average -- what a receiver might see in a room like this at negative 55 EVM [ph] . So on average, I'm giving the competition the benefit of the doubt here, 2x the throughput performance versus our competition, okay? Like I said, this will be a big advantage for us going forward. We plan to maintain this lead with a new generation of combo chips coming out in 2014, okay?

Now one interesting thing that's happening is the 2x2 configuration of our combo chip have been adopted in the tablet space. Arguably, the tablets are consuming more content from the Internet than smartphones, a bigger screen, people might want to watch more video on their tablet. So the manufacturers of these devices upgrade from a 1x1 configuration, 1 receiver, 1 transmitter, to a 2x2 configuration. And what we see over the course of 2013 is Google or Apple or Amazon, in this case, all using Broadcom products, have put 2x2 WiFi into a smaller screen device. They started with 10. They moved to 9 inch, 8 inch and more recently 7 inch.

We think this trend's going to continue, okay? We think this trend will continue down to phablets and into smartphones, okay? The reason is video consumption. 2x2 WiFi provides higher rate and higher rate of range. So if you're further away from the access point, it'll work that much better. And in a lot of use cases, it's actually lower power because you're on the air less time, okay?

So we see this happening into the smartphone business in 2014. Now keep in mind, if we're successful with this, a number of our competitors who have integrated a 1x1 configuration into their baseband are going to need to spin those basebands in order to keep up with the WiFi cadence. Again, year after year, we've mentioned that the cadence of connectivity moves differently than the cadence of cellular baseband, and this is going to prove that again in 2014.

So moving on again to that Internet of Things I mentioned earlier, connecting everything to the Internet. This is going to be a big opportunity for Broadcom. And within that space of the Internet of Things are these wearables, the body-borne computing and sensing devices. And what we see, of course, is that connectivity will be the avenue by which all of these devices are connected, okay? You'll need WiFi to connect to the Internet, to communicate data up to the cloud to be processed in your smartphone, Bluetooth to have the wearable connect directly to the smartphone, GPS to give you contextual awareness so the smartphone knows where you are, whether you're at work or whether you're at home, and NFC for simple pairing, okay? So all of this require Broadcom strengths in highly integrated, low power and enabling new use cases with our connectivity devices.

And then how do you service a model? You might ask, "How do you service a model like the wearable, where there's hundreds of companies?" There's not a clear leader yet in the wearables space. What we did was create a model to go service all of the various companies behind the door of this wearable segment, okay? What we did was we created an SDK, a software development kit, that enables companies that don't know how to do wireless, okay? There's a lot of companies in the world that know how to do wireless like the Samsungs and the Apples, the HTCs. But then in the wearables space, a lot of new companies don't know how to do it, so we have enable them. And we have to enable them without a lot of engineering support, okay? So we do that by providing this SDK. We do it by going through a set of distributors. And in fact, we actually enjoy a higher ASP because we're going through a set of distributors because some of these companies are lower volume, but there's a lot of them. And we provide online tools and support. So a model that's scalable.

And to give you a sampling of where we are with this WICED initiative -- we call it WICED, which stands for Wireless Internet Connectivity for Embedded Devices. We're currently engaged with 400 companies, and I'll give you a sampling of those companies: 7 of the top 8 home appliance companies, 14 different fitness companies, 7 different thermostat companies, okay? So that's just a sampling. Many different companies who are building automotive appliances, so automotive accessories, okay? A lot of different companies in this space, and we're supporting them with a model that we believe scaled and is a low-touch model for Broadcom so that we can leverage our WiFi and our Bluetooth technology into the space.

Okay. So let's switch gears. Let's move over to our cellular baseband business. I've mentioned over the course of many years that having a complete portfolio is very important. Complete portfolio allows the customer to use one of our chips but then quickly switch over to other devices, in other words preserve their investment. Well, with our acquisition, which closed about 71 days ago, we now have access to the 4G marketplace, with leading 4G LTE IP. And I'm going to give you an update on how we're executing on that 4G LTE IP right now, okay?

So let me start with what we did in 2013 with our 3G roadmap, okay? We launched single-, dual- and quad-core HSPA+ devices into the marketplace, and that allowed us to broaden our footprint in places like China. But now with the acquisition we completed about 71 days ago, we also have 2 -- we have a family actually of cellular SoCs for the 4G LTE space, a dual core and a quad core. So now we're able to service our customers with a complete product line. ASP is increasing as you move from the left to the right. Many different configurations, evolved modems and a product line that's really targeted at the sweet spot of the cellular market, right, the emerging market for smartphones, both 3G and 4G, okay?

So let's talk a little bit more about some of this. Let me start with what we were able to do, proving that point about roadmap and roadmap leverage, having a customer start with one of our chips and then proliferate the roadmap. Let's talk about Samsung, the biggest smartphone company in the world, roughly with 1/3 of the smartphone market today. We now have multiple models in production from Samsung, single core, dual core and quad core; screen sizes ranging from 3-inch to 5.8-inch, okay? So I think this partnership with Samsung is very strong. And you can see, again, that roadmap leverage in this slide, okay, where having more than 1 product in the roadmap really makes a big difference and allows our handset partners to proliferate our technology across a range of handsets.

And let me talk beyond Samsung and what's happening with our 3G business. So right now, we have 2 Tier 1 companies we claim as our customers, ZTE, which is shipping the phone you see, our single-core HSPA platform, into places like T-Mo U.S.A.; and TracFone in the U.S.A., which is this largest MVNO, which is a virtual mobile operator; and Samsung, which we just discussed. What I'd like to announce today is our third Tier 1 OEM customer, HTC. And I'm holding in my hand and you could see on the screen, a picture of the Desire 601, which uses Broadcom's quad-core HSPA+ SoC, plus quad-core or quad-combo chip that I mentioned earlier, okay? So HTC is going to be our third, what I would consider, Tier 1 customer.

But then, beyond the Tier 1 customers that we're supporting with our cellular platform, we have been able to proliferate our technologies in places like China and emerging markets, okay? So if you look at the chart below, you'll see the number of handsets has extended. Last year, when I was here I talked about 44 different handsets from companies in places like China. This year, that number has exceeded 100, okay? So 100 different handsets from Broadcom, enabled by our turnkey design, which I'll go through in just a moment, and you can see some of the companies on the chart, okay? Companies like K-Touch, G'Five and then some leading India brand names like Karrbon and Micromax; carrier-branded phones from places like Idea Cellular, which is one of the top 4 cellular carriers in India.

So single-core, dual-core and quad-core phones all coming from servicing the developing regions of the world. And I anticipate that this trend will continue because our turnkey, as I'll show you in a moment, has progressed quite a bit from the early days of our turnkey design.

So let's switch back to 4G for a moment. Let's go into a little more detail on those 2 SoCs I showed you earlier. We're going to switch to a 3-digit part numbering scheme because I think a lot of the communities, both the customers and the financial community, thought our 5-digit part numbering scheme was too complicated. So we're simplifying our part numbering scheme and going to 3-digit part numbers. So EOS2 will now be referred to as the M320 design, M320 SoC. And hopefully, our 3-digit part numbers will be much easier for you to understand moving forward.

Let's talk about the M320. Available now, it's ready to go to production. This device is a 28-nanometer SoC, dual core that supports the 2 most popular implementations of LTE today, FDD and TDD. Recently, you heard that the China carriers were granted frequency spectrum using TDD. Most of the U.S. carriers are using FDD. This chip can support both. It also supports backward compatibility to all the relevant 3G standards, and most importantly, it supports new features that are coming in the marketplace as opposed to features that have already been deployed. Two examples of that: category 4 LTE. Most of today's networks that are deployed are using Category 3, which is 75 megabits per second. The world is moving right now to category 4, which is 150 megabits per second, and these chips will support that. Voice over LTE or VoLTE as we refer to it, again, a feature that hasn't been deployed yet, but will be deployed, and these chips support voice over LTE, with SRVCC fallback, by the way, which is the only type of VoLTE that could be deployed into the network, okay, dual-core A9 at right at 1.2 gigahertz.

When Scott was up here, he talked about our ability now to wring out some of the conservativism in the numbers we cited at the time of our acquisition. Now that we're in production with the M320, we're able to see some of the early yields. And now we're able to announce a new product called the M320+ [ph], which improves the CPU speed by 25%, so a dual core running at 1.5 gigahertz, which will sell for a slightly higher ASP than the dual core. So we're already creating a family of LTE solutions for 2014, 1.2 gig dual core, 1.5 gig dual core and then, of course, the quad-core version of that same device called the M340 [ph], okay?

So for those companies that need to go quad core for the retail marketplace, we also have a solution. And as Scott mentioned, this chip now we feel confident sampling in first half of 2014, okay? So all this stuff is on track. So I feel good about 2014. And I want to make sure that everyone understands, our customers are reacting very well to this -- creation of this roadmap. Our customers really want an alternative supplier on LTE.

So let's talk about that turnkey design. So 71 days after the acquisition, I can hold up our turnkey design for the M320 or the EOS2, and this is a picture of the design on this slide, okay? What makes a turnkey different than a reference design? This turnkey is ready to go to production. A customer could change the ID, but we've already optimized the bill of materials. We've optimized the bill of materials with suppliers that our China handset partners like to use in China, okay? And we've gotten further to preserve their investment. I mean, the turnkey is all about making it easy for our handset partners to adopt the Broadcom complete solution. It's all about investment preservation and ease of adoption.

So I just mentioned, the family of LTE solutions, the M320, the M320+ [ph], the M340 [ph], okay? We made it even easier by making all these 3 chips pin compatible, same motherboard to support all 3 of these chips. And then we've gone further than that. We have also done software investment preservation. Once a customer or we port a customer's UI to our framework, the first handset gets done. We can quickly move to handset #2 and handset #3, a lot of reuse. And that's where the leverage is going to come from, reuse on both the hardware and the software platform. We've done it for 3G already, with 4 different turnkey designs that you can see proliferated to over 100 handsets now in places like China, and we'll do it for the 4G family of products you just saw, okay? So the turnkey -- fifth-generation turnkey. We've learned a lot about the first 4 turnkeys we've produced for all of our 3G designs. We've optimized it. We've optimized the cost. We've optimized the design, and now we're going to apply all that knowledge to our 4G turnkey.

Okay. So let me go and talk about why would customers pick us for LTE, okay? Well, we have a compelling roadmap. We just showed you that. We have a turnkey that's ready to go. But let's talk about again some of those quantitative advantages that we have over the competition. So our power consumption is 20% to 30% lower than our competition, and our die size is more efficient. It's smaller than our competition by 25% to 30%. Again, you might ask, "How could Broadcom make such a bold claim in terms of both power consumption and die size?" In order to substantiate this, you need to look at the legacy of the modem team we just acquired.

The modem team we just acquired is the Nokia modem team located in Europe and in Finland, and they've been working on modem technology for 15 years. In fact, they've sold billions of 2G and 3G modems. They were there when the 4G LTE standard was created. They were probably 1 of 2 semiconductor companies at the table when the LTE standard was defined in 2004 and in 2005. So they were at the table, they've had a number of years to work on this and to optimize it. In fact, their IP is also very portable. Their IP, their modem IP found its way, of course, into a lot of TI ASICs that were done for the Nokia phones that many of us carried 7 or 8 years ago.

And that same IP then went into an STE device called the U8500, and that U8500 was already qualified at places like Samsung, HTC, Sony, Yulong, Antim [ph], places in China, Europe and of course, Nokia, okay? So this modem IP is -- has a long legacy of history, and the team really knows what they're doing. I'm very happy to report that integration is going very well. The team is very motivated to make sure that this thing is a success. In fact, the team thinks that this is the best partnership for them. Our connectivity and their modem technology is going to be a very formidable force in the marketplace.

Now complement that with all these quantitative things we just talked about, ready today with VoLTE, ready today with carrier qual. We've already got carrier qual on many of the most difficult networks like AT&T, like NTT docomo, like SoftBank in Japan and a lot places in Europe, including Vodafone and Orange and EE, okay? The 2 most prevalent types of LTE supported with this chip at Cat4 rates complemented with our complete platform. So these are the reasons that our customers are very happy with us right now because they really like what we've done with this 4G LTE story.

Now let me continue because, last year, we talked about a Fin [ph] modem. This year, we're going to embellish that Fin [ph] modem with more features and sample it in mid-2014. Category 6 LTE, which is now 300 megabits per second, this will include carrier aggregation for the carriers that don't have contiguous spectrum. This will be our first chip with TD-SCDMA for China Mobile. So we've been working on TD-SCDMA now for a number of years, and that technology will port over to the new modem architecture. All of these chips are complemented by Broadcom's RF subsystem, which can do quad-band 2G, peta-band 3G and as many bands on LTE as you could want, okay? Voice over LTE, LTE broadcast, advanced transmit technology like Envelope Tracking, okay?

So you might ask, what makes you -- why should you believe that Broadcom could deliver a chip like this. Here is the reason. The modem team that we acquired was working on this technology when we acquired them. The modem architecture is very modular. Unlike some of the new implementations on LTE, the way our team architected this thing is in a very modular fashion, so the 4G category 4 will [ph] not need to be changed in order to put Cat6 [ph] on top of it. And we've already taped [ph] out a chip with a significant amount of this category 6 IP. We think this chip is going to compare very well with anybody else's modem out there in 2014. So that product will be available and sampling in mid-2014.

And then, of course, there's going to be more chips that are coming. I talked to you about the SoC roadmap, which really will hit the sweet spot of the LTE marketplace, phones from $100 to $300, different screen sizes, different IDs. And there's another series of cellular SoCs coming. And Scott mentioned the tape-out of these chips are things we're measuring very closely. The team is very focused on execution. So within these new series of SoCs, we're going to improve the ISP, the image signal processing IP, 3x the performance to improve the quality of the images that you take because everybody wants a good camera in their smartphone.

We're going to improve the speed of the pixel processing, and we're going to make sure we have these advanced features, things like signal, smart imaging for things like beautification [ph] and augmented-reality apps, okay? So our imagining IP is going to improve significantly. Now keep in mind that Broadcom's ISP technology was in the first Nokia 42-megapixel phone, and Broadcom's ISP technology was the first ISP to make it through the Samsung image labs in 2013 -- in 2012, okay? So our ISP is only going to get better from here.

And then on video, of course, strength of Broadcom. Obviously, video permeates every part of the company, 2x to 4x improvement. We're going to be able to support UltraHD configurations, increased frame rates to 120 frames per second in certain configurations and implement more efficient codecs like the H.265 codec, which is 2x the efficiency -- I'm sorry, about 50% the efficiency of the prior version, which is H.264 [ph], so better video compression, okay? So our video IP is going to improve.

And we're spending a lot of time and paying a lot of attention to CPU and graphics in this new SoCs that are going to tape out in 2014. We're going to do SoCs with more application processors. We're not going to say how many more, but certainly more than 4. And we're going to increase the performance by 3x. Where Rajiv can put 20 cores [ph] on a chip in the mobile and wireless space, we need to have a very careful tradeoff of power and performance, and we're going to do the right thing here. But I do want to make mention that Broadcom is an early access partner to ARM for their 64-bit class of processor and their V8 [ph] instruction set. So we announced that over a year ago, and that process -- that new processor will be a part of our new roadmap in 2014.

And then finally, graphics. Again a 5 to 10x improvement in our graphics performance, support of the latest graphics libraries, like OpenGL ES [ph] and GP GPU, okay? We're going to maintain our advantage in graphics, where we're smaller sized and lower power. So we're going to pay attention to all of this IP. When you consider this kind of IP added to the new modem IP we have, this will be a very exciting set of cellular SoCs that will emerge in 2014.

So back to platform control. Platform control is very important. Our customers want us to deliver a complete platform to the marketplace. And great mobile and wireless technology allows us to complete many different platforms, and I'm just citing a few here, starting with the end [ph], set-top box or cable modem or DSL gateways, okay? The IP that you need are -- is LTE, WiFi, Bluetooth and NFC, all technology that allows us to complete the set-top box platform.

So back to platform control. Platform control is very important. Our customers want us to deliver a complete platform to the marketplace. And great mobile and wireless technology allows us to complete many different platforms, and I'm just citing a few here, starting with the Hand set-top box or cable modem or DSL gateways, okay. The IP that he needs are -- is LTE, WiFi, Bluetooth and NFC, all technology that allows us to complete the set-top platform.

Moving to IoT and wearables. Great WiFi, Bluetooth, NFC, GPS and touch are all needed to complete that platform. Let me just make a reference that just yesterday, we announced the latest generation of our GPS technology that can receive signals from 5 different satellite constellations, 88 different satellites in all, and provide the best positioning solution to all of the apps that we all use every day that rely on position to give you the best information. That's the kind of IP that's necessary to make a success in the IoTs and the wearables.

And then finally, the biggest market of all, smartphones and tablets, okay. We just talked about all the IP that we have at Broadcom to go win in this space and now the 4G technology that we have ready to go into production immediately.

So in summary, we have made a lot of progress on our complete platform solutions with the addition of our 4G LTE IP and the acquisition we completed 71 days ago. Our customer relationships are strong. We services about every leading handset company with our connectivity solutions, and we have more and more engagements on our cellular platform, okay.

And now, what I'd like to do is to talk about one more thing, okay. Scott mentioned some very aggressive goals for us to achieve in 2014. And I'm very happy to announce, because of our intense focus on execution, our first LTE phone from the world's largest smartphone manufacturer, Samsung, I'm holding it in my hand, and there's a picture of it on the screen, okay. This is a phone that's based on Broadcom's M320 or the EOS2 platform that was referred to previously, Category 4 with a global launch coming in early 2014. So this phone will go to production in early Q1 2014, based on Broadcom's LTE solutions. So again, pulling in the milestones that we talked about at the time of acquisition. So I'm very excited about this. I'm very happy that our friends at Samsung allowed us to hold up this phone, a phone that's going to go to production in Q1 of 2014.

So with that -- so with useful presentation, I'm going to invite Chris to come back up onstage. Thank you.

Chris Zegarelli

All right, we'll now conduct our first Q&A session. I will then invite Scott and Henry back up onstage here with Bob. And as we go through the questions, just to remind, let's start with this one question per person and let's -- introduce yourself, just your name and your firm name and then we can go on to the question.

Question-and-Answer Session

Chris Zegarelli

Let's just start right up here up front with Harlan from JPMorgan.

Harlan Sur - JP Morgan Chase & Co, Research Division

So you talked a lot about the platform approach. And so just to clarify, with the announcement of Samsung ramping in Q1, are these phones using your baseband and connectivity products? In other words, is it truly a platform solution to kind of kick the launch off? And then on these new LTE phones that you're ramping in Q1, maybe you could try and quantify the rough dollar content per smartphone?

Robert Americo Rango

So to answer your first question is, yes, every phone that I showed in the presentation here has that Broadcom cellular SoC, has Broadcom connectivity, okay. And as far as the unit volume goes for the Q1, it's really hard to predict the success of any one phone. But I can -- I feel confident that once you get to that first phone, as I mentioned in my presentation, that we can proliferate the technology into more phones and more IPs into more regions of the world. So that's really the success formula for us is to start with one phone and then to proliferate it across one handset manufacturer's product line. I think on the LTE platform, okay -- it's a -- on the LTE platform, the dollar content for these kinds of devices will probably range from $18 to the $22 range, depending on the customer base, okay. So compare that to our 3G solutions, which probably are in the $10 to $15 range, these are significantly higher. And of course, that'll change depending on the customer and the actual device that we talked about, whether it's quad core or whether it's dual core.

Henry Samueli

You're lucky Bob answered that question. I think Eric and I would have said 2 digit.

Chris Zegarelli

Okay, let's go up here with Ross Seymore.

Ross Seymore - Deutsche Bank AG, Research Division

Ross Seymore from Deutsche Bank. Actually, a question for Henry. From the board level, how do you define success and failure in handsets with any sort of granularity? We've -- investors have been frustrated about what happened on the cellular side with Broadcom for a while. It's great you're showing the LTE phone. Bob, I think it's a great first step, but we've seen traction in 3G that's then gone away either with design loss or with ASP pressure coming in. So from the board level, how do you define that strategic effort?

Henry Samueli

Sure. Good question. Well, clearly, you need a set of milestones. You can't define success unless you know specifically a set of milestones that you need to achieve. You saw on this presentation, both from Scott and from Bob, a clear set of milestones, product development milestones, customer design win milestones. We have internal revenue goal milestones. So all of those are going to be measured on a regular basis. And that's how we judge every one of our business, how much customer traction do you get, how many product new design wins do you get, revenue ramps. I mean, every business in our portfolio, the executive team reviews on a regular basis and we make sure they're on track. So we do have quantitative measures in place, many of which will be targeted for 2014 evaluation.

Scott A. McGregor

And on track to create shareholder value. Let me just make sure that's clear as well.

Chris Zegarelli

How about Vivek, right up here up front as well.

Vivek Arya - BofA Merrill Lynch, Research Division

Vivek Arya from Bank of America Merrill Lynch. Actually, related question to that. I think, Scott, in your presentation, you had this very meaningful revenue. Would 5% of your Mobile business be considered meaningful revenue in LTE this year? And maybe just a larger question is on the competitive landscape. You seem to have a very strong product portfolio, but Intel also put out its product. They also claim to be in a Samsung device. Marvell had an announcement recently. So it's just going to be a very competitive market. So maybe, Scott, if you could tell us what meaningful revenue means. And if maybe, Bob, if you could help us understand the competitive landscape in LTE. Is the design win pipeline enough in terms of value and in terms of the breadth of customers that you can actually -- we can look forward to meaningful revenue?

Scott A. McGregor

Meaningful revenue for me means 9 digits of revenue from LTE next year, and that correlates to probably tens of millions of phone. So that's probably the ballpark we're looking for in there.

Robert Americo Rango

Yes. As far as the competitiveness of the solutions, I think you're going to hear a lot of announcements and a lot of chatter about LTE announcements in the next year. And I think that the proof's going to be in the pudding. When you look at the devices, first of all, you need to determine are they stable, are they ready to go to production. LTE is much more complex than 3G technology. And it's going to take anybody longer to do it, which is why I made such a strong point about the legacy of our modem IT. So I think the time is going to really tell in 2014. I think that Broadcom will be one of a very small set of companies with both a thin modem and cellular SoCs dual core and quad core to be able to go after that 4G emerging marketplace. And there's a lot of growth to come in 4G, as I mentioned in the presentation. So 4G in the U.S. is obviously a big thing, but a lot of countries around the world are going to still deploy LTE. So there's a lot of runway. And while most of these carriers are going to be very compelled to move their spectrum over to 4G because it's more efficient, right. There's more spectral efficiency in the 4G OFDM carrier than using the 3G technology. So they'll even reform some of their spectrum over from 3G to 4G. So anyway, I see a lot of growth, and I think we're in a really good position right now with these devices that we have.

Scott A. McGregor

Vivek, to add on what Bob said, I think it's stability, pretty much as Bob said. That's very hard. People can talk about shipping something. Stability takes a long time to get and something we have an advantage. I think also, remember, Broadcom's pretty conservative about announcing our products, okay. We tend to announce when they sample our ship, whereas other folks tend to announce before they even tape out the product. So be careful when you compare roadmaps of different companies, okay. What we talked about today are production-ready, okay, or at least sampling, okay. Some of the other folks haven't taped it out yet. We've seen some announcements just in the last week. People announcing things and they haven't even taped up the chip yet. If we announce everything we haven't taped out yet, there will be a much longer list of stuff we could've put out there.

Chris Zegarelli

Let's go in the back then, John from CS, I think.

John W. Pitzer - Crédit Suisse AG, Research Division

Yes. It's John Pitzer with Credit Suisse. Scott, for many of us in the audience, I think we're less concerned about the technical milestones around the baseband business. You will get that right and drive unit share. I think we're more concerned about what the economics of that market might look like 12 to 18 to 36 months out, especially given how price aggressive 3G has become. And so can you talk about kind of the competitive landscape from a margin perspective? And everyone seems to think this market's going to consolidate to Qualcomm, MediaTek. And then what if it's just a more competitive market and the economics just aren't that attractive longer-term, can you address that?

Scott A. McGregor

I think you're right that it is competitive economics out there. And margins are certainly challenged. 4G has gone very fast. People are often using price in order to get into that market. And for me, it means that it really touches to that point that I made in my talk, which is you just can't sell basebands and make money in this market. You got to sell the whole platform, okay. And so the margins on all the other pieces besides the baseband actually aren't bad, okay. That's pretty good part of the business. So if you can sell the entire package, I think you need to look at platform margin or the pieces that you can contribute to that, not just the baseband. So the baseband's probably going to be tough economics, okay, for the next few years. I do think there's going to be consolidation. If Bob's team does what they say they're going to do, I think we're going to be a strong player in that. I think a lot of other guys are going to shake out. And I think we're going to be making money on all those other pieces, okay, while we're doing battle on that baseband piece.

Chris Zegarelli

There's a question over here on the left.

Gary W. Mobley - The Benchmark Company, LLC, Research Division

Gary Mobley with Benchmark. I had a question about your 5G WiFi. Could you share with us with your estimate as to the penetration for 11ac in routers versus smartphones? And not all of the leading smartphones vendors have adopted 11ac per recent teardowns. And I'm just wondering what the considerations will be for the next generation of some of the leading smartphones and moving to 11ac and how meaningful that could be for Broadcom.

Robert Americo Rango

Thank you, Gary. So I think the question on 5G WiFi is a good one. I still think this is a trend that's going to continue for the next 2 years. My estimate of the penetration for a 5G WiFi today is 20% to 30%. I mean, it's -- on routers, it's probably the same percentage. It's going to increase over time. We're coming out with a new generation of 5G WiFi chips, just like we did with 11n, continual improvements, cost downs, lower the cost ownership, increase the integration of the device. So this is going to be a trend that will continue for at least another 2 years as we cut over from 11n to this 5G WiFi technology.

Scott A. McGregor

And, Bob, that's a meaningful driver in our growth and connectivity next year.

Robert Americo Rango

Yes. I mean, there is an ASP uplift for us when we sell 5G WiFi over 11n.

Chris Zegarelli

Perfect. And there's a question in the back, left back corner.

Unknown Attendee

This question is for Henry. Henry, I'm -- just kind of at a high level. You had talked about increased cost per transistor and design cost at 28 nanometer, I think, 4x over what we saw at 40. I often get the question, who's going to bear the brunt, do you think, of these higher cost? Is there -- is it the foundry? Is it the chip designer? Can you pass this on to the OEM, or will the end customer, at the end of the day, get these higher -- and what are these costs? Who's going to bear the brunt of this, I guess?

Henry Samueli

Sure. First of all, minor correction, the 4x cost went from 28 to 16. You said 40 to 28. It's 28 to 16. So the 4x increase is from $75 million to $300 million. So you operate in a business we've had -- a business model that we've had since inception of the company 22-plus years ago, where we've seen a gross margin model in the 50% to 52%. And costs have changed over the years in every single process node we've designed in, back in 30 -- 350 nanometer going to today. So cost changed, ASPs changed. One constant we've held in the last 22 years of the company is gross margins. So there's no reason to project any change going forward. As cost change, ASPs change. The end customer pays for value. In the end, every product you buy, you're willing to pay for value. And the value is going up in these newer generations, faster, smaller, cheaper, lower power, all good things which justify additional value and additional cost being paid by the end consumer. So ultimately, the end consumer product prices will go up to bear the cost.

Scott A. McGregor

Well, Henry, wouldn't you say that if you can't get the value, you're not going to migrate to the smaller feature size?

Henry Samueli

Exactly.

Chris Zegarelli

I see there's a question on the front row.

Unknown Attendee

For the 5G WiFi AC, it seems like for the handsets which is -- the handset market is, right now, controlled by the 2 big players. One of them seems to have adopted it and the other didn't for -- the adoption curve is no longer determined by the technology or what you're pushing, but by -- rather by the bigger players. So your revenue stream is slower. If I can look at the previous generations, the determination now is more price determination for the bigger players. I don't want to add another $10, forget it, even if it's better. I will do it next year. How can you push the adoption curve to follow the technology more closely to get this higher ASP and improved revenue?

Robert Americo Rango

I'll be happy to take that question. So it's interesting that you said that the 2 big handset vendors, only one of them adopted 11ac. But if you look closely at the product line of the other big vendor, you'll notice that they have adopted AC in their non-mobile product lines, like in their Macs. So I think they see the value in our 5G WiFi solution already, and I think they will -- I can't speak for what they're going to do, but it would seem a natural course of events for them to proliferate it across their product line.

Unknown Attendee

So just a delay.

Robert Americo Rango

There is a delay because, obviously, we've had one big OEM adopted in 2013. I think moving forward, what -- the way we facilitate the adoption of new technology is through our complete platform, right, because we need to offer a set of solutions to our customers that they can differentiate, that helps them differentiate in the marketplace. So as we drive the cost down, the cost of ownership, we'll get more adoption.

Chris Zegarelli

And there's a question in the front row here in the far left. And just a reminder, introduce yourself, please, before you ask the question.

Anil K. Doradla - William Blair & Company L.L.C., Research Division

Yes. This is Anil Doradla from William Blair. Henry, it's kind of a big-picture question to the semiconductor industry. You talked about software being a bigger component in the semiconductor chips. But unfortunately, the industry has never been able to price this software in their chips. I mean, it wasn't too long ago before you would just look at the die size and price it. Going forward, do you think the industry will be able to resolve that?

Henry Samueli

That's a good question, but it really ties into the whole platform solution. Again, if you can -- software is a very sticky aspects of our design. In fact, Bob will tell you that one of the primary reasons he wins a lot of these connectivity solutions is because of the stickiness of our software. So software can be part of that platform bundle that gives you leverage and winning designs that has value to the company. And sure, it's an R&D investment. It's a significant R&D investment. But in the end, it will generate that incremental revenue by the stickiness of the design, and that's the benefit to the company.

Unknown Attendee

[indiscernible], are you thinking that [indiscernible] because it's the best that you could offer and its stickiness, plus if you just [indiscernible] get some premium [indiscernible].

Henry Samueli

Well, the question is how much premium can you generate by the software platform, and that will be determined over time. Clearly, as these become -- as bundles become stickier and stickier, you can create more value there and, hopefully, generate value in the marketplace by selling it appropriately. So time will tell. But you're right. It's an issue that there is continually increasing investment in software, no question.

Chris Zegarelli

Let's see, there's a question right near your left.

Mark Lipacis - Jefferies LLC, Research Division

Mark Lipacis, Jefferies. Henry, you showed the charts of transistor cost leveling out. Intel has made the argument that they are able to stay on the transistor cost curve. And my question is would you consider using them as a foundry? And would you be able to benefit from those lower transistor costs that Intel is arguing for?

Henry Samueli

Sure. Well, Intel has made the public statements many times that they're trying to expand their foundry business, and Broadcom is not tied to any particular foundry. We use multiple foundries. We would love to expand our foundry business with other customers. And especially if they offer us lower prices, we'll be first in line to latch on to it. So certainly, we're open to discuss foundry business with any company out there and especially ones that would be willing to offer us lower prices, be happy to.

Chris Zegarelli

Let's see, there's a question in the back. Right over here as well.

Blayne Curtis - Barclays Capital, Research Division

So Blayne Curtis of Barclays. Scott, this is your comment, but maybe Rob as well. You said you were going to grow connectivity next year. If you could talk about the assumptions behind that. And then secondly, on the platform, I think you're stressing that you're going to sell entire platform. I think you've seen Qualcomm developed a platform. And there's plenty of good players in power and audio, and they've been able to have a very high attach rate there. Why is it different in connectivity outside of maybe some high-end SKUs that may want 2x2 or some high-end features? Clearly, you guys are the leader there. I don't think anybody would deny that. But where the volume is, how do you keep that attach rate with connectivity?

Scott A. McGregor

Let me start and then, Bob, why don't you join in? I mean, I think there are a number of reasons that will contribute to our growth in connectivity next year. We have probably the broadest connectivity portfolio of anyone out there. And others have tried to develop some of the pieces, but we still have many more pieces than they do. And within each of the pieces, a much wider set. So many folks will have like a connectivity offering or 1 or 2 connectivity offerings. Bob's team has put together quite a range of mix-and-match things that are all compatible and play together. Breadth includes things like NFC. You'll see us announcing other pieces of connectivity over the course of the next year. So there's many more pieces of the cell phone that you can still get in that space. So I don't think anybody challenges us from a breadth of all the connectivity solution there. Bob?

Robert Americo Rango

Yes. I think the other part of the question was how do we differentiate with our platform. Let me make that into a 2-part question. I think on our platform, the way we differentiate, of course, is we're moving our features up in terms of making sure that we're at the leading edge of technology, where there is more profit margins and our customers are going to need us because of the complexity of the solution. And our connectivity will complement -- our connectivity leadership will be a way for us to differentiate our basebands and our cellular SoC platforms. Also, keep in mind that when you talk about some of the new emerging spaces like wearables, the connectivity solution is the platform. As Scott mentioned, you don't need a quad core in a device that's only going to measure the number of steps you're take during the day. You might just need a small CPU embedded in a lot of our -- in a connectivity solutions. We already have a CPU. We can make it accessible to the end manufacturer so they can run their RTOSs and small operating systems right on our device. So in that sense, the connectivity chip is the platform.

Blayne Curtis - Barclays Capital, Research Division

Looking in the forecast for growth, are you assuming you're going to hold share, or are you making it up with the LTE games, where you'll have good attach rate with those, or ASPs or such? Any color?

Robert Americo Rango

All 3 avenues, okay. All 3 avenues, I think, will allow us to grow our business in 2014, higher attach rates of things like 802.11ac 5G WiFi, okay. Our success in our own baseband and then higher ASPs for things like 2x2 WiFis. So those 3 avenues that I mentioned are all the kind of the underlying fundamentals that allow us to make that statement, growing our connectivity business in 2014. And there are some other things, too. I mean, one of our key handset suppliers now is one of the largest carriers in the world. The last I heard, there's been some new introductions into China Mobile, and I think that'll also help drive our connectivity business in 2014.

Chris Zegarelli

Perfect. We have time for one more question. Let's go right here in the middle with Jim from Goldman Sachs.

James Schneider - Goldman Sachs Group Inc., Research Division

I was wondering if you can talk about the thin modem strategy. Clearly, it seems to be a lot more competitive in that space heading into next year. So could you maybe comment on whether you think the thin modem is a really viable revenue driver in terms of scale for 2015 sales, or should we think of it more as a stepping stone to kind of the next-generation advance platform for multicore LTE SoCs?

Robert Americo Rango

I mean, I think it's both. I certainly think it's both. I mean, it's going to be IP that we take forward into our cellular SoCs, as you mentioned. So certainly, that's the case. But just speaking to the market for thin modem technology, we obviously know there's a lot of handset companies who have their favorite app processor, and they're not about to move off of that app processor. And that not only applies to the company in Cupertino, but it applies to companies in other places in the world like Korea and China. There's a lot of company -- a lot of handset companies have their high-end phone where they want to pick the app processor that goes with it. Perhaps, that app processor has some features in it that they desire, and they need a thin modem to complement that. So I think the thin modem is a way for us to deliver the value of the modem IP without any of the other pieces, and it's certainly a high-value component. I mean, the modem is probably the most complicated part of the entire SoC subsystem. So delivering that is a high value proposition for us.

James Schneider - Goldman Sachs Group Inc., Research Division

So do you think that could be a material revenue contributor in 2015?

Robert Americo Rango

We won't comment on 2015, parts of forward-looking statements. And maybe Scott would like to make a comment on it. I really can't predict -- I can't tell you 2015.

Scott A. McGregor

Nice try, Bob. No, I think -- Jim, I think we've got a really competitive LTE-advanced thin modem there, and our hope is to win some significant business with that. It's too early to forecast how much that is. We got to win that business first. So maybe a year from now, that might be a question you'd get a better answer on that. But certainly, it is our goal to win significant business with that product.

Chris Zegarelli

Perfect. Thank you very much everyone for the first Q&A session. We will come back into another session after you hear from Dan and Rajiv. So again, thank you very much. And we'll move to the next presentation.

Unknown Executive

Please welcome Dan Marotta, Executive Vice President and General Manager of Broadcom's Broadband Communications Group.

Daniel A. Marotta

Okay, tough act to follow there. Hopefully, I'll be able to keep everything going here. Good morning, everyone. My name is Dan Marotta, and I'm responsible for the Broadband Business at Broadcom.

And what makes a best-in-class business, right, market share, growing faster than the market, beating the competition, right very simple and, obviously, multiple growth drivers. Anybody can come up with those. I think we know where the new revenue is, but you need the investment dollars, the team, all the stuff Henry talked about to go chase and you have to execute.

So I think there's no question the Broadband Group is the undisputed leader in the access and the pay-TV markets. There's no question about that. We've been doing that for 20 years, slowly but surely getting into new market spaces. And we have more to go. We've always won by technology. I mean, we have some very, very good marketing guys, no question about it. We win by engineering. We win by getting it done first. I think it's a very simple formula.

I think the business results show that. And I think this year has been an outstanding year financially and execution-wise. We have outgrown the market. We've significantly outgrown our competitors, you see here, 1,500 basis points. We've just had an excellent year, both financially, execution-wise, lots of innovations in the market as we made our transition to 28 nanometer, starting in last CES.

We've introduced a raft of products. We're not first movers in each process node, but we used it strategically to launch key products and continue to drive for integration, continue to roll out our own products. We've launched our multimode small cell products, very innovative product keep are our small cell business on the leading edge. I'll talk about that later. And then of course, it's a race. Don't get wrong. This is a very serious race, and we always intend to be first. And we've had a lot of first this year, listed some here, and that's how we win. You cross the finish line first. There's -- hope it comes back and answer any questions later.

Another view on kind of what Henry showed earlier in terms of what do we do here for the company, what are we doing for broadband in terms of leverage. I think the broadband group does very well, but we can't get along without the rest of the company. And sometimes, one of secrets of the company, in my opinion, since I've joined here 11 years ago, is just the IP sharing, the scale of the company, the engineering culture. I think we don't talk about it enough and what it's done for our execution across-the-board. I think everybody knows we have a leading IP, but we've got the people to be able to put that together and get it done first. And we're -- at least in my business, we're able to apply it to all our end markets, which I think is another little bit of a secret on the R&D line, to be able to get that leverage. People talk about software. We have a lot of software in my business, and you want to be able to leverage that across all these end markets. They're very similar. And I think that's also been a contributor to the success of the business.

Today, we literally have hundreds of concurrent projects going across these operators and many more. I couldn't fit in this page. You could fill this whole page up with Indian and Chinese operators. That's how many they are. And there's no company in the world today, in my markets, that is executing on that type of scale. And that's what it takes to be successful. You'll see many competitors come in with some acute entry strategies and try to target some of the big guys, the Comcast, the echoes, the directs. But to make a real business, a sustainable business like we have today, you have to go global, and it takes literally -- there's hundreds of projects going on right now to make this into a substantial business.

As an example, why do we win? Here's something I showed last year, right, real execution. Here's an example box that you might see at a North American operator. So here's the chips that would make it up. Not only are there 5 chips from the competition, right, it's not one competitor. It's actually 5 competitors. And you can see the knockout effect, lots of additional memory. Everybody has got to have their own memory subsystem. You take a look at something that we're deploying today, we've got a cable modem front-end and we've got an MPEG SoC. This is not even the one Henry showed, the latest one. This is the previous generation. And of course, we've got less memory. That's cost, right. Memory is expensive and generates power, board space is smaller. And of course, we've got more features. So we've got all those key attributes, cost, board space, but we've also got more features in there. I think that's the sort of how-we-win story. And of course, we're able to consolidate all the process, right. You've got 5 companies all trying to optimize their profit story. And of course, we're able to do it in one corporation.

Let's go through some of the growth drivers. I talked about it earlier. I think these are very obvious. I don't think any of the competition doesn't know about these today. But of course, you need the investment dollars. You need the team. You need the IP, the sales channel. I mean, there's a lot of factors that go in to be able to execute in these markets.

Let me go through a couple of them quickly. Of the Indian market, everybody knows probably the biggest TV market in the world in terms of content, middle-class viewing audience, this has been a great market for us. We're actually shipping into 6 of the 7 direct-to-home satellite guys. They're listed on the bottom. We've -- it's been a very successful story. It's a tough market, not an easy market in terms of ARPU. But it's transitioning. And you can see here, I'm showing the market moving now to PVR and HD. That's what we've been waiting for, okay, and that benefits us. We built a lot of specific products to deal with all the nuances of this marketplace, have a great relationship, great team there to help support this transition. And just probably 1.5 year ago is now the cable business in India starting to ramp up. Lower end, but it'll move the same way as satellite did to higher end. Before long, you'll be -- they'll be doing HD. And of course, in this case, it's a much bigger TAM than satellite. And so this has been a great story for us.

Another area of the world that's gotten a lot of press recently, South Africa. MultiChoice is an operator that I'm highlighting here. They're in Durban. And they have a big terrestrial offering that they propagate all through Africa. And this has been a huge deal for us this year. We've shipped 2 million terrestrial set-top boxes. So that's a 2-chip solution with a DVB-T2 and an MPEG SoC in the back end. We have a single chip coming. So we'll keep improving our competitiveness here. And this has been a really big deal. And this will just keep growing next year. This market is also very dynamic. And like many parts of the world, the appetite for TV content, the large population that Africa has makes it very, very interesting for us, and we've done an outstanding job here.

And last, but not least, an area the world, as we come up to the World Cup in the spring and 2 years later the Olympics, a lot of focus on Central America, South America, Mexico. An operator I'm going to highlight here is Net, which is the big cable operator in Brazil, part of América Móvil. And this has just been a great story. It's been a great sales story, a marketing story, a technology story. Two years ago, these guys didn't even know us. And now, they're putting our name and our logo on their investor presentation, okay. That's technology, leadership, them seeing a lot of value in us as a company to be able to deliver advanced solutions. I think that's -- this is a real big deal. And when you're working with the leading operator, I think that's pretty satisfying to see that when you go from a dead start 2 years ago on your back to be able to have this type of notoriety and relationship with an important operator is a big deal, a big deal for the team as well.

New features, right, we have talked about all these new areas. But this is something Henry touched about. This is Full-Band Capture. And all the comments here are just amazing. You can see here comments from Comcast below. So Comcast has about 1.5 million cable modems today that actually have our Full-Band Capture for -- and deployed. And this is an example I pulled out of a deck they have. I think this is important somewhere. We've all messed around with these splitters. You can't read which is the import and which is outport. You connect them up. In this case, you had a customer on top right here. This is bad. This is -- the guys connected it wrong. It's got a lot of attenuation. You're not getting a good video signal. They were able to diagnose this remotely and call this guy up, get him to put the in where the in goes and the out where the out goes. And of course, you got a better flat band all across this band here. This would have taken at least a day or 2 to drive a truck around this neighborhood and figure this out. And so from an operational perspective, here's something that's delivering real operational results. Here's something that, 2 or 3 years ago, we had as an idea and a technology, now it's deployed in the field, and a major operator like this is actually utilizing it. And you just look at these comments and all the emails and accolades, and this is -- it's quite satisfying. It's just amazing. People try to copy this. Of course, the imitation is -- it's appreciated. But we're continuing to do this. We'll put this in cellular. We'll put this in other areas as well. We're working on that, so stay tuned.

Another feature, vectoring, this is something that can increase the bit rate and the reach. And so if you're an operator who has a network, you've invested in that network and you want to go ahead because the competition increased the reach in radio network. Deutsche Telekom is a big operator that's looking to deploy this today. Take a look at their investor deck. You'll see they want to put out 20 million to 24 million lines of vectoring in their network, mainly to increase reach. They've got lots of competition. And so this is a great technology that we pioneered. We worked -- this is an example, a customer I showed last year ,where we developed 3 FPGAs with them. We've gone into production with this chip, huge reductions from the ASIC to actually a standard product for them. And then you can look at the ASP increases and the performance that it brings to the operator with no network investment other than changing out the ends of the network. So it's a big advantage for a subscriber line operator and, of course, the supplies to guys like ATT, not just Deutsche Telekom.

Another feature Scott mentioned UltraHD and just all the buzz around it. It's a new TV technology. Not all UltraHD chips are created equal. And this is something you have to pay attention to. What we're focused on, what operators are focused on are p60, 60 frames per second. That's what makes sports content look good. If you don't have that, that's something an operator is going to purchase. They're also looking at 10 bits per color element. Today, most MPEG system use 8 bits per color element. We're focused on 10 bit. And then finally, HDMI 2.0. You don't have HDMI 2.0, you don't have p60. So just be careful out there. Not all technology is created the same.

All those features -- here's an example of all those features wrapped up together. We know this guy, don't forget his name, right. And Rogers want to make sure their customers remember them as well. So here's 8 tuners, home networking that Rogers in Canada and Toronto area is deploying right now. And we've got 2 more generations of product right behind this. So the innovation keeps going.

How do all those features translate? You guys always hear about features, features, features, right? This is an outstanding financial result. You do not see this in the chip business. Look around. Here's '10, '11, '12, '13, broadband ASP's growing. You guys know the story in the chip business, right, prices go down. And I think this is the feature story. This is the execution story. You do not see this. Look at our competition. Look at any company you want to. When even a modest increase like this, also in the face of driving into geographies that have lower feature sets, higher volumes but lower ASPs, so even in the face of that type of share capture, we're modestly growing our selling price. That's a big, big deal in the chip business.

Over-the-top, let me explain this chart a little bit here. I continue to say over-the-top has been complementary to our business today. It's been driving operators. We're making a business out of it. We're selling these chips to -- into these products. So we think it's important. You can see here what I've outlined. Here's pay-TV subscribers and here's OTT subs -- loosely subs. I mean, this has got low-end guys watching Rob Ford videos on YouTube, how entertaining is that. You've got the middle tier that maybe do some purchases. And then you've got the actual subscriber that subscribes to Netflix. And so that's all in this number here. And you look at these and they're not killing each other. They're coexisting together. You've even got operators that are putting Netflix on their network. And so I continue to -- we see this in the market that these are coexisting. And then we see phenomenon like the Netflix driving all these video traffic, and that feeds back into our broadband business. That's what's driving this modem business, pushing for DOCSIS 3.0, for VDSL, for vectoring for PON is video traffic. So we continue to see that with the competition and in this market space being complementary. And of course, we're involved in it, too. We want to make a business out of it. The more, the merrier.

Continuing to build momentum in small cells. Our customer base here continues to grow. We have our 3G LTE product that we're bringing into production as we speak. We're selling the 3G-only product today in one operator. And I want to highlight, as an example of a real innovation, is an operator called Free located in Paris, right. They're kind of competitive local exchange carrier, if you will, in the old days. And here's our 3G modem, and they've actually put it into this carrier, which is a PCI Express card and actually slips into the back of the set-top box that they deploy with us. And so what Free has done is they're building a whole network of small cells using their residential footprint to offload traffic from Orange or France Telecom's network onto their own network in order to make their business model work. And they chose our product as the power is low, the performance of our cellular product is very, very good and they see our roadmap with LTE and 3G as being very important and, of course, the support and the know-how. So I think outstanding result. You can see other foreign factors. Here's the Alcatel Vodafone unit, kind of plugs in the wall, kind of interesting form factors, so continuing to innovate in this area.

And so just to wrap up, leading the market and achieving substantial business performance. I think this year has been the best year ever, record revenues, record profit, outgrowing the market substantially. I think we're very proud of ourselves in the Broadband Group. We think we have a lot of work to do. The management of Broadcom is -- I think we're very hard on ourselves. We always want to do better, and I think that's part of the culture of the company. And there are a lot of growth drivers out there. They're not secrets. And the question is do you have the money to go invest in them and can you get the job done and do it in a way that, I think, makes sense for investors. I think we've been doing that, and we'll continue to do it.

Thanks.

Unknown Executive

Please welcome Rajiv Ramaswami, Executive Vice President and General Manager of Broadcom's Infrastructure & Networking Group.

Rajiv Ramaswami

Good morning. I'm here to talk to you about our Infrastructure & Networking businesses. And by way of a quick introduction, the group is a variety of products starting from physical layer products for optical and copper media, switches, knowledge-based processors, multicore processors and a bunch of other products. And we target primarily service providers, data centers, enterprise and small and business end markets. About 40% of our revenue comes from the service provider market. Data center has been the fastest-growing segment, has grown from about 20% to now 30% of our business and then about 20% is enterprise and the rest is the SMB and the other category.

Now what I'd like to walk you through here in the next 20 minutes or so is, first, to show you how we are a best-in-class business infrastructure. And second, we're going to talk through all the growth opportunities that we have ahead of us. So let me start with the state of the business as it stands today. Now for anybody to build a best-in-class business, you look at the different categories here. You look at your technology leadership. You look at market leadership. And then finally, you are developing business. And so let me walk you through each of the 3 pieces here for our business.

So starting with technology because that's the underpinning of what drives everything. And let me start with our fastest-growing segment here, which is data center. Now if you look at the data center market, and you look at the infrastructure, it's all about compute, network and storage. And we have leadership position today in the networking portion of the data center, but opportunities over time to expand into the adjacencies in compute and storage, and these are all coming together as well over time.

Now you look at this picture. You look at a server rack on the left, and that's the conventional server rack, velvet rack server, sometimes late servers. And then you have these newly emerging micro servers coming out in the middle here. And then you have the aggregation and core layers that's switching towards the right. And if you look at our product portfolio, we have the best-in-class, the broadest and the deepest portfolio when it comes to data center. So starting with controllers that sit inside the servers, those are the NetXtreme products to the physical connectivity, whether it be copper, 10GBASE-T, or whether it be fiber, whether it be switches at the top of the rack or at the aggregation layer, or the control plane processors that go along with every switch that's deployed in the network. And every one of these, we have best-in-class products. And it's not just the best of the products, but also the velocity at which new products are coming out every year from us. Those products highlighted in pink there are all new products that we'll introduce this year.

Trident II, for example, is now just getting into production. Number of customers have launched their products already. It's starting to be deployed in volume as we speak. And we're also starting to see, in many cases, the acceleration in terms of, from the time we have a product ready to the market to the time our customers are taking products into the market. Trident II, as an example, we sampled our final silicon back in summer and it followed the production as we speak today. So again, a very strong position today in the data center market.

If you look at the service provider network, this is perhaps the most complex network that we sell into just because it's so diverse. It's got many, many different boxes starting from the access area, whether it be mobile or whether it's fixed access, going into the backhaul portion of the business and then, eventually, into the aggregation and the core network layers. With many, many different products needed. Over time, we built out a portfolio here where we have optimized solutions for every step, every box in this network. And again, as you can see here, the velocity at which we're introducing new products, again, the products shown in pink over there, is pretty high. So these are all products that we introduced this year. We continue to have a lot of momentum behind the service provider business as well.

Now the final aspect of technology that I want to cover here is multi-core processors. Again, this is a relatively new market for us. We entered this through our NetLogic acquisition. And if you look at where we are today, earlier this year, we launched our XLP II device. This is a mid-space machine. It's the highest-performance communications processor out there in the industry today. It has -- it's a 20-CPU machine with 80 volts of CPUs, 100 gigabits of networking, 100 gigabits of security, full hardware virtualization. And there's a thing called network function virtualization that's emerging in the carrier market where more and more carrier applications that were running in the network are now running on server cloud CPUs. This is the perfect device for those types of applications because not only does it have the CPU horsepower needed to run those applications, but it has all the offloads that are essential to run these networking applications efficiently.

Now at the low end of the portfolio, we have ARM-based engines today. We have -- these are standard ARM cores and we wrapped them -- wrapped around the cores again all the networking accelerators and offloads that are needed to optimize them for specific applications. And this is, again, a rapidly growing business for us.

Now where are these things going? So this is what we have today. We'll continue to introduce MIPS machines as we go forward. We have other MIPS products in development. And we recently announced our high-end custom CPU as well. So the new CPU architecture that we're working on is a 3-gigahertz plus server-class CPU. It will be one of our first products delivered in 16 nanometers. As Henry mentioned, this is an area we're focused on. And it's, again, going to have Intel Haswell type of CPU performance, combined with all the networking offloads that are going to make it a best-in-class CPU design in the industry. And all of these, since we've announced the architecture, we're seeing a lot of interest from customers for a variety of applications for this particular CPU.

At the lower end, we are the only licensee for the ARMv8 architecture, and we're going to be taking those standard calls and hardening them and also then combining them with all the other networking goodness and putting them into SoCs. So if you take this all together, we have the broadest portfolio in the industry for processes optimized for communication applications today.

Now where does this put us in terms of our customer base and our market position. If you look at where we're at here today, I mean, we sell to pretty much every OEM that's in this business. And you can see the list here. It's a very diversified customer base. As you can see, they're everywhere. And we also continue to display the ASICs at all of these customers.

And then in turn, if you look at where these products eventually get deployed, they're deployed in the largest carriers and the largest data centers in the world, across the board. And this is where you heard Scott talk about how, essentially, 100% of networking traffic goes to a Broadcom chip, and it's timely because our chips are in every one of these carriers and data center developers in the world. So a very strong position here.

Now what does all of this lead to in terms of market? So in terms of the financials, if you look at our last several years of growth since 2009, we've been growing at about 18%. Now some of you may wonder, well, 2009 to 2010 was an outlier. And indeed it was an outlier. But even if you take 2009 aside, we've still been doing double-digit growth.

And if you look at this year again and how we are performing, we expect to be over double-digits this year in terms of growth, while our peer group is going to be flat to slightly down about 1%. So clearly, all of this, the strong product portfolio, the better customers, the fact that we're in the right markets is factored in our performance.

Let me move next and talk about the growth drivers and look forward rather than looking backwards. Now again, we have a pretty broad set of drivers here to propel the business forward. So you look at data center, service providers and new markets for us and multi-core processors and wireless infrastructure and automotive. The -- all of these together represent many, many growth opportunities. And let me walk you through each of these. So starting with the data center. At the macro level, CapEx spend in cloud data centers continues to go up at a fairly amazing rate here. You can see the rate of growth here. And there's a war going on right now between Google, Amazon, Microsoft, trying to build out these large multi-tier and data centers to attract enterprise customers. And enterprise customers are migrating their applications over to public clouds at a rapid clip, or they're building their own private clouds internally to replicate what's being done in the public cloud. This is obviously a huge growth area, and this CapEx spend has been a great tailwind for our business.

But this is not about the CapEx spend coming here as well, but it's also about the fact that we have a best-in-class product portfolio and that there are many, many market transitions that are driving the growth here. So if you look at data center, again, many of you heard about some of these buzz words. You heard about software-defined networking, the ability to disaggregate the network, the ability to have more programmability in the network. You heard about network virtualization. You heard about virtualizing servers, with VMware pioneering this. But now it's also about what's driving the network so that you can actually host multiple customers in a common data center and share the infrastructure across multiple customers and segregate those customers. So that's why you need network virtualization. You heard about OpenFlow, which is, again, a way to program the network and decouple your control plane and data plane functions. You heard about the need for more programmability in the network, need for more instrumentation in the network. All of these are driving transitions at a rapid clip. And the question is, who can execute in terms of delivering on these features at a rapid clip? And so if you look at our history here, today, we are on Trident II, and we've delivered an initial clip at 24x10 gigabit a while go, moved to Trident Plus and then now Trident II. And we have continued to execute at a rapid clip, delivering all these features. And this is really why here we have been able to successfully display almost ASICs everywhere to the point where these chips are now de facto in every data center product that gets built, not just for the top-of-rack layer, but also for the spine layers, all the aggregation layers as well. So this growth, and you can see how a 10-gigabit revenue has been growing. And so, in addition to all the other features, there's also, of course, scaling up in capacity. So growing from -- your gigabit to 10 gigabit, 10 gigabit to 40 gigabit and then now driving 100 gigabits as the next step. So I think it's fair to say that this growth is going to continue and, again, driven by 2 factors here: One is, of course, the macro spend on CapEx in terms of all these data centers, and the second is the continuous displacement of ASICs out here.

As an example, Cisco, for example, recently said, with their launch of Insieme, that they're now adopting a merchant-class ASICs solutions where it used to be all ASIC in the data center. So that's obviously upgrade benefit to us.

Now looking at the service provider market. Here, the CapEx story is a bit more muted. You can see that we don't expect carrier CapEx to grow substantially over the next several years. And it hasn't been the case historically as well, and so you have probably a low single-digit growth CAGR. But it you look at where they're deploying their CapEx, it's in the area that you see in green below. So they're deploying it in carrier ethernet. They're deploying it in wireless infrastructure. They're deploying it in routing and packet optical. And they're taking it away from -- wire is taking it away from Sonet/SDH. And what -- where we stand, of course, is we're [indiscernible] Exposed to all the areas that they are spending money. And of course, this is helping us grow. And if you look at our historical growth in the service provider market, we've been growing at about 21% CAGR in that market.

And so, again, we think we're exposed to the right set of parts of the service provider market. And again, as I say, a lot of ASICs spend still to be had in the service provider market, which we will go after, either convert them to merchant or try to provide value in terms of being an ASIC provider to some of these opportunities as well. So again, good opportunities looking ahead in service providers.

Now we've talked about multi-core processors earlier. And I do want to point out again that there is a very large market opportunity for us here, it's about $3 billion of annual spend. And if you look at the segments, it's right across a variety of segments. You have networking control plane, where every switch that gets sold, there is a processor opportunity for the control plane. You look at secure appliances, things like security, load balancers, packet inspection platforms, continent inspection platforms. All of those, again, require a lot of deep embedded processing. You look at wireless infrastructure. You look at base stations. Those require embedded processing. Storage, high-end storage systems also require embedded processing.

Now at the low end of this market, which is, again, relatively new for us, we're in consumer network-attached storage, we're in consumer retail routers, we're in point-of-sale terminals. So these are all market opportunities for us at the low end of the portfolio. And so, here, we have the opportunity to have a process of portfolio that's going from, say, a few dollars of ASP to north of $1,000 of ASP. That's the range of products that we have today in our portfolio to go after this market.

Now looking at wireless infrastructure, again, we have to take a long-term view of this market. The 5 players that really matter in terms of having the majority of share of this market, you see them listed at the bottom. We've got Ericsson, Huawei, Nokia Service Network and then, of course, Alcatel-Lucent and ZTE bringing up the rest here. And if you look at the market here and you look at the base stations, you look at the backhaul and then you look at the radio heads and, of course, also, the mobile packet core, all of those, there's over $1 billion being spent in that every year, and in terms of semiconductor spend for us.

And how are we going after that? So first thing is we have on the backhaul solutions available today, whether it be microwave backhaul, whether it be fiber-based backhaul, we have a leadership position today in the market. And again, for us, particularly in microwave, it's a matter of displacing in-house solutions that most of these guys are still doing, and we'll over time work our way into those.

If you look at -- for the base stations and the radio heads, what's happening is that all of these OEMs are now moving to slightly different architectures in terms of how they're partitioning their systems. Some of them moving to more of a pizza box-type basis and architecture. Others are going to a modular architecture. And depending on how functions are combined in platforms, some people have separate processing combined with a single processing. So our approach here is to go to the market with the base of IP that we have. We have embedded processing capabilities. We have IP for radio heads such as the Digital Pre-Distortion technology, and we have DSPs in-house, and then put together customized solutions for each of these 5 folks. And of course, this is a long-term engagement, but we're in there to stay, and this is a 5- to 10-year growth story for us.

Meanwhile, as we do that, as of today, what we see is [indiscernible] in a couple of these base stations for LTE. So you see kind of mobile, for example, ramping up their LTE base stations. The 200,000 base stations that they talked about, they are actually rolling them out as we speak. And we are supplying to 2 of the providers that provide those solutions for them with our processors. But as I said, long term, the game is about providing these customizable solutions for this entire market. And again, we think this is a great opportunity for us to work on.

Now moving on to the last big market that we want to talk about is automotive. We've talked about this, and as Scott mentioned, this could be $1 billion opportunity for us in the long term. Again, I want to emphasize the design cycles here are very slow. We've been working on this already for a few years. And this year, we saw our first car go into production. 2014 BMW X5 is out there in production. It's got Broadcom ethernet in it. But it takes time. And these guys were the first to go into production. And if you look at your opportunity here, it's the network everything inside the car. You start out with the 360 camera system. That's what -- if you look at this adaptive driver assist, which is an option that you can buy under BMW, that's using Broadcom chips. But you can go from there. You can go to infotainment, the ability to hook up all the audio/video inside your car, and then onboard diagnostics. And there's many different networks inside a car and the opportunity here is to consolidate all of those networks over a standard, ethernet-based standard, and that's really the opportunity. And of course, one of the benefits of going with ethernet is that we reduce the cable weight by substantially leading to better fuel economy.

Now how are we pursuing this? Well, over the last few years, we've been working on creating a de facto consortium to establish a standard -- de facto standard for automotive ethernet. It's called OPEN, One-Pair Ether-Net, alliance, and what it's got at this point is every auto manufacturer is a member of this consortium. And so, all of these folks are interested in and actively at this point looking at all of our ethernet in their cars. And again, like I said, this is a -- it takes time. It takes -- these cars have 4 to 5 design cycles. But I think it's a great opportunity for us long term. And over time, we can expand into this market and expand the product base that we sell beyond just switches in size to other products.

So where does this all leave us here? So I want you to walk away with -- I hope I've convinced you that we have a best-in-class infrastructure business here based on the portfolio products that we have, the technology leadership, the position in the market in terms of, again, leadership products and on the segments we've been playing, combined with the financial results that you're seeing and also our -- outperforming our peer group. As we look forward -- I talked to you about all of these growth drivers in data centers, service providers, processors, wireless infrastructure and automotive. And if you look at it, it's a combination of secular drivers as well as macro drivers. If you look at some of the secular drivers that we have here, slowly replacing ASICs and gaining share in these new markets are both secular drivers for us, regardless of what happens to the markets themselves. And then if you look at positions where we already have strong share, for example, in service provider and data center, perhaps, and switching, those again are determined by some more macro trends. But together, you take those 2 together, and you look at the diversity of the customer base that we have, the diversity of the markets that we're playing and the growth drivers we have, I think they're looking at a business that can be very healthy for many, many years to come. So let me stop there and have Chris come up.

Chris Zegarelli

We'll now do a question-and-answer session with Rajiv and Dan. Please join me on the stage. Let's start here at the back.

Unknown Attendee

Actually, similar question for both, maybe starting with Dan. So your business did quite well. It seems to go in cycles, right? You have a very strong year then you sort of have a modest year, but profitability was very strong. As I think about the next couple of years, is it reasonable to think of your business as a mid single-digit grower at similar levels of profitability? And then I have a similar question for Rajiv. You said that you have grown double-digit for the last many years. Is that the kind of run rate that can be sustained, all else being equal, and the profitability levels that you experienced this year? So not asking for guidance, but just do the trend support those kind of growth rate and profitability?

Daniel A. Marotta

Yes, I'd say yes. I think that we don't see the business structure changing within what we're doing today. The business is, it can be lumpy, as you said. And I think one thing to bear in mind is a lot of these platforms take years to get to production. Even the Rogers box I showed, that's actually 3 years of work. So that generates a lot of this behavior. And so the R&D pipeline gets very, very deep. But I agree with your assessment of the kind of gross rates, and I think the business structure is in pretty good shape, even in the face of a lot of expansions throughout the world.

Rajiv Ramaswami

Without looking at forward-looking numbers here, but I think our goal is to outgrow the market and outgrow our peer group. And to the extent that you believe in all these drivers here, I think we're in a good position to get to some of the numbers that you were talking about.

Unknown Attendee

And Anything on profitability? Because profitability was very strong in 2013. Is it reasonable to expect you could maintain those levels?

Rajiv Ramaswami

I'd say we continue to invest in R&D and we're very careful about how we monitor investments. And again, as long as the revenue growth continues, profitability will be there.

Chris Zegarelli

Perfect. Let's move up here in the front with Stacy up here in the front row.

Stacy A. Rasgon - Sanford C. Bernstein & Co., LLC., Research Division

Stacy Rasgon, Sanford Bernstein. A question on the infrastructure business as regards to Intel. How do you guys think about, I guess, maybe the shorter-term opportunity with Intel as their own products are driving a good amount of data center growth versus maybe the longer-term threat as their pushing their own products much more broadly into the adjacent markets, including networking, both on the processor side as well as the switch side? And then they have their -- this kind of integrated rack scale server that's incorporating all of this with their products?

Rajiv Ramaswami

Yes. So Stacy, certainly, Intel is a formidable competitor. We compete with them today in processors, specifically for communications applications. And so far, what we've seen is that we are able to optimize our processors for networking applications. Intel has tended to just sell their server-class CPUs into networking applications. While they have great processing performance, they're not necessarily optimized for networking applications. So we've been able to compete pretty effectively with Intel in that market. Now of course, now Intel is also now talking about getting into the networking market. And they bought Fulcrum. And they brought a variety of assets, whether it's Cray or whether it's QLogic to come -- now it all depends on where the rack architecture goes. Our view is that the rack architecture should be modular. You're not going to buy everything. Proprietary architecture is from one person. You're going to want to build these things out of modules, and you're going to go -- continue to have best-in-class, you have a best-in-class CPU, you have a best-in-class interconnect, you have best-in-class controllers. And we also think that ethernet is the right interconnect mechanism within the rack. And there are some other things proposed like proprietary, as well as PCI Express within the rack. So at this point, we think that ethernet is the appropriate interconnect architecture. So as long as we continue to drive this standard-based modular architectures and continue to execute on the products, I think we're in good shape.

Stacy A. Rasgon - Sanford C. Bernstein & Co., LLC., Research Division

Do you think it's still true that they haven't correctly optimized their processors for networking? I mean, they have specific SKUs, with Jasper Forest, with the Autorun...

Rajiv Ramaswami

Yes, so they're starting to now -- so certainly, we're starting to see them build more products for the networking market.

Chris Zegarelli

Perfect. And we can stay in the front here, just 1 row behind Stacy.

Daniel L. Amir - Lazard Capital Markets LLC, Research Division

Daniel Amir from Lazard. I have a question more about small cells. You guys talked about it in the past. Clearly, it looks like potentially big opportunity. How -- what milestone should we be looking at there? Maybe you can give us a better idea how we should look at this market. Would it be meaningful in your business? Or is it really going to be driven still on your traditional drivers of the set-top box market and some of the old [indiscernible] stuff?

Daniel A. Marotta

Yes, I mean, it's a good question. I think the market is still pretty nascent. And I think we don't give revenue guidance for the business. It's much smaller, obviously, than set-top will ever be. But I think it's an important add-on business for us today, and it helps us drive technology into these other channels. So I think it will be a contributor. It has process -- business structure that's very different. It's more infrastructure. I think it's important for that category, anytime you can add something with that kind of profit structure. And that's very interesting for us, given the complexity. So I think we'll guide you guys with releases. We'll see at the trade shows. We'll show you some of the technology, but I think we'll steer your attention to the right places with the right press releases and the design wins that you can pay attention to. And I tried to highlight one today that we think is an important example of our technology that's being deployed.

Chris Zegarelli

Perfect. We have a question here, halfway back.

Christopher Rolland - FBR Capital Markets & Co., Research Division

Chris Rolland, FBR. So maybe if you could talk about Trident II, who the customers are, are they mostly OEMs or white-box switch guys? And how do you sort of see the profile of Trident II in terms of uptake playing out?

Rajiv Ramaswami

So as I mentioned earlier, Trident II today has been adopted by every OEM, almost all the customer that you saw on that list, as well as white-box reference platforms out there. So as you imagine, some of the large cloud operators, of course, have migrated and are deploying white-box solutions today, and they will be using Trident II. But again, all the OEMs out there in the market, many of them, you've seen at least 4 or 5 announcements, you saw Cisco, you saw Arista, you saw Extreme, you saw Brocade, just many, many OEMs out there today with the Trident II-based products that are in production, as we speak. So we've always seen the ramp, and it's well underway.

Christopher Rolland - FBR Capital Markets & Co., Research Division

What would you expect that profile of the ramp to look like? Has there been sort of a pent-up demand for II?

Rajiv Ramaswami

I mean, there certainly was pent-up demand. Trident II has been one of the fastest in terms of getting to production from the time of your sampling. So typically, our products take about a year to get into production. Trident II, from the final release, within 3, 4 months going into production. So there's clearly pent-up demand. And it's, again, driven by the fact that people are migrating to 10 gigabits and top-of-rack solutions. And again, Trident II is also now being used, not just in top-of-rack, but also on the aggregation list. So we clearly see the pent-up demand and have been -- we have been servicing that over the last quarter.

Chris Zegarelli

There's a question up front here from Harlan.

Harlan Sur - JP Morgan Chase & Co, Research Division

Harlan Sur with JPMorgan. Question for Dan. One of the largest cable pay-TV service providers looks like they're getting ready to pull the trigger on their next-generation platform, their IP-based platform, which includes Newton [ph] clients and a gateway hardware architecture. Question is how big of an upgrade cycle is this for Broadcom in 2014? Obviously, set-top churn is always a good thing for Broadcom, but is this a big enough upgrade cycle to drive good incremental revenue growth for your business in 2014?

Daniel A. Marotta

Okay, good question. I think I know what you're talking about. And it already is doing it for 2014. And I think, from the standpoint, a lot of these products we've had in place have been working out for a while. As you know, software is quite complicated, but also it's very advanced. And I think that's what I was trying to get at with the OTT competition. There's the case where that offer is -- today, we're going to build something better than that, put together this IP structure. And I think it is making an impact in 2014. That's already started. So good news.

Chris Zegarelli

Let' see, there's a question in the back row there.

John W. Pitzer - Crédit Suisse AG, Research Division

John Pitzer at Crédit Suisse. Rajiv, a couple of quick questions. First, is there any way to quantify the market opportunity for SDN for you? What does it do to your served available markets as we see SDN become more pervasive? And then secondly, there was a question about Intel getting more aggressive on the networking side of the processor. I was kind of curious, when you look at the more traditional server market, there's an argument to be made that a lot of the applications are being driven more by the IOs than the raw compute power. And so given kind of your IO expertise, is there any sort of opportunity for you to try to go after some of the server market a little bit more aggressively over time?

Rajiv Ramaswami

So on the first question, on SDN. I look at SDN as one of the growth drivers. It's not the only growth driver, but I mean -- and there's a number -- SDN, as viewed at a number level, when different people talk about SDN, they're referring to different things. So for example, if you look at some of the use cases, the first use case is building overlay networks with SDN. So that's what makes it to virtualization. So that gets typically, again, implemented over a much silicon-type infrastructure, so that's a good driver to the extent that people build these virtualized data centers. The second opportunity is to provide -- enable more of a disaggregated approach where you have providers of -- service providers of boxes, providers of software and providers of services. And so that kind of disaggregation is still in the early phases. Some of the large cloud guys have gone there. And that could be also wrapped up under the SDN number land that, again, presents a growth driver for us. The other aspect of SDN is, typically, it comes -- associated with it more programmability or configurability, more instrumentation capability, and to the extent that we can implement those things, those functions in our product like we have been doing that, again, presents a great opportunity for us to solidify the position. So overall, I think it's just -- I look at this as one of many things that's driving our data center penetration and growth going forward. The second question on the IO and servers, we certainly do see an opportunity to perhaps come up with optimized servers with specific workloads. Because at the end of the day, it used to be -- you settle down for a server and you run everything, all your workload around the server. And as you look at how these workloads are now being optimized and you have different workloads now, cloud for example, Google has their own search applications to be optimized. For networking applications, for example, a lot of the CAD applications are very specific workload. You're running at packet core now on a server. So those things do require, in my view, more network functions, more acceleration capabilities. For example, there's no reason why you want to do packet encryption or decryption on a CPU. That's just a waste of cycle. So you're better off doing that in hardware or by deep packet inspection. So we see an opportunity to, therefore, take these communications processors that we are building. And in the context, for example, of NFV, whether they get deployed in the network or whether they get deployed in the data center, they could be optimized for running these virtualized communication applications.

Chris Zegarelli

Perfect. There's a question there in the way back up in the third area.

Unknown Attendee

Question for Rajiv. Rajiv, seems like there's a shift from MIPS to base processors in the multi-core. Just wanted to get your views on where you are competitively compared to your other players in the space in terms of the shift? And second question is, why are you doing that 16 FF? You showed in the side that 16 FF is 2x more expensive than 20-nanometer? They're limited to questionable scaling benefits, limited performance benefits, why not a 20-nanometer?

Rajiv Ramaswami

Yes, I think to answer the first question, we certainly do see more interest in ARM in the networking market. Now in many ways, MIPS is actually ahead in our markets because they have 64 bit. They have high-performance MIPS in the market. So they have a lot more of an entrants customer base on MIPS related to ARM. But again, looking forward and looking at the ecosystem, I mean, there seems to be more and more of a preference to ARM, which is why we're moving to ARM as well. In terms of how we see we are positioned, I mean, again, if you look at the various announcements out there, there's people -- many people going after ARM for server applications. You've seen a lot a list, 7 to 8 players talking about ARM servers. That's not our focus. Our focus is embedded applications and networking. And again, for that, what we're trying to go after is a real high-performance core. Start with a high-performance CPU core. Like I said, 3 gigahertz plus core, the highest-performance core, surround that with all the extra data. Now why did we pick 16-nanometer as opposed to 20? Because I think the processors, particularly at the high end, is one where you want to squeeze every ounce of performance you can at the lowest possible power. So power-optimized performance is very clear -- critical for a lot of our applications. And in terms of time frame, at this point, given where we're at, we felt that we didn't want to be behind when we came out with our products. So that's why we picked 16 over 20.

Chris Zegarelli

There's a question up here in the front row.

Mona Eraiba

Mona Eraiba, Trust Company of the West. A question to Rajiv. What is the difference in the design for the public cloud companies, the Web 2.0 and the traditional or the proprietary data centers? And what is the benefit for you [indiscernible]? And if you could also elaborate your gross respect [ph] to more CPUs or to more -- are you per rack or per unit? What works for your -- to drive demand for Trident II?

Rajiv Ramaswami

Okay, so first of all, I'd say that the public cloud and private cloud enterprise data centers are getting more and more closer, as we speak. It used to be -- with the advent of virtualization, I think the -- you may be using VMware more on the enterprise and you may be using KVM or Linux-based virtualization in the cloud. So I think the workloads are kind of starting to become similar for both public cloud and private cloud. And therefore, the way these data centers are getting built are starting to be similar. Now the data centers and the public cloud tend to be larger, larger scale, much bigger scale into the -- compared to an enterprise data center. So if you're looking at data centers that will have tens of thousands of servers and public cloud, whereas a large enterprise data center may have a few thousand servers. So there's a distinct difference in scale. Now in terms of architecture as well, these are coming together. So again, I'm getting a little technical here, but cloud data centers are migrating more to a Layer 3 architecture with a simple leaf-spine model. The enterprise is traditionally being a multi-tier model with Layer 2 at the access and then Layer 3 aggregation in core. But again, those architectures are becoming fairly simple. Now in terms of what's driving demand for Trident II, it's definitely the need for more scale, not just in capacity, of course, with servers going to 10 gigabits per second, clearly, you need more 10 gigabit ports. If you think about 40 servers per rack, okay, requiring two 10-gigabit reconnection speeds, that's 80 for 10 gigabits right there, okay? And then moving up from the access to the aggregation layer, you're looking at 40 gigabytes, so 100 gigabytes going forward. So all of that drives the need for more capacity. But in addition to just capacity, it's also about other features that we're talking about. For example, for network virtualization, or for SDN, when we came out with our original Trident product back in 2010, at that point, network virtualization was fairly early along. And so since then, there have been multiple standards that have been established for network virtualization in terms of the ability to create overlays in the network. That's something called BFN that was established by VMware and VTR which was established through Microsoft. And so now, the new chips need to support these. The other chips didn't have them; Trident II has all of these things. So again, that's deployed as an upgrade cycle even for people who had Trident chips before. So combination of features and capacity and scale are both driving demand.

Chris Zegarelli

That's perfect. And we'll leave it there. We have one more Q&A session after Eric's presentation. So again, thank you very much to everybody, and we'll move to the next presentation. Thank you.

Operator

Please welcome Eric Brandt, Broadcom's Executive Vice President and Chief Financial Officer.

Eric K. Brandt

Good morning, everyone. So as the last presenter, I'm going to try to tie together some of the economics of what was said by the various presenters earlier today, Scott, the various general managers. And to do so, the theme of my discussion really hits 2 points, right? The first is around the nature and the power of the economics of the company that we've got. Certainly, as you've seen in the world-class businesses we have in broadband and in infrastructure, but even inside of Mobile & Wireless, that enable us to make the significant investment in LTE. Now one of the things I'd like to do also is sort of give you an example of the nature of the leverage where we have invested in businesses, which have been subcritical mass and actually have turned out to actually grow into fairly significant businesses and drive significant leverage in the P&L.

The second topic I'd like to talk about, and Scott touched on it, really is how we are managing tightly as we go through this investment cycle. Certainly, you can see that in the performance of the core businesses. You can certainly see that when you sort of peel apart the underlying economics of the business. And I'll talk about a number of different things, including portfolio management, operating expenses, et cetera, as I get to my discussion on managing tightly as we go through the investment cycle.

So let me start with the first point in terms of the powerful economics. And so what I'd like to do is I'd like to peel apart in the economics of the P&L. We'll look at it in a couple of different ways. Starting at the corporate level and then looking at the strength of the core businesses and the nature of the LTE investments, so you can get a sense of what the underlying economics and performance of the businesses are as we operate them.

So on my first chart, what I've done is I've taken a 5-year view, and you can see the performance of the corporate component of the business. And so the corporate component of the business, revenues up on a compound annual growth rate, about 13%. Cash flows are up roughly 13%, notwithstanding the fact that probably, over the last 3 years, the industry, certainly in our sector of the industry, has been relatively flat. I think what's more interesting probably is that chart in the middle. And so I've spent a lot of time for those of you I see on the road, I see many of you on the road talking about gross margin and gross profit. Why? Because it's the gross profit dollars that define the economic health, the gross margin of the company that define the economic health and enables the company to make the investments that it makes in R&D. And so what's interesting about that chart in the center, as you can see over the time period, not only does the gross margin expand over 100 basis points, but probably more importantly, that the business, the Mobile & Wireless business, which is -- typically runs below our corporate average of gross margin has expanded in that time frame almost 1,500 basis points, 1,450 basis points. And that becomes important when you look at the actual performance of the individual businesses and the actual performance that they've shown in their operating margins as they go forward.

So let's peel that back one more notch. And so if you look at the individual businesses, if you look at the corporations, you'll see that the operating margin over the last 5-year period expanded about 70 basis points. But underneath that 70 basis points, you can actually see operating margin expansions that are broader than what is seen on the corporate level because of that mix shift to the Mobile & Wireless side of the business. So broadband and infrastructure are actually up 170 basis points, almost 2 full operating points for those 2 businesses. And Mobile & Wireless is up a pretty strong 630 basis points over the time period. Now again, because of that 1,450 basis points mix shift, you get a much more muted picture at the corporate level. But if you look more closely at Mobile & Wireless, and that's where I'd like to drill into in a little bit more detail, it doesn't tell the whole picture, right? And so one of the things you have to look at is what's the performance over the time period and where is that investment that we've been making in LTE? And I can say the acceleration of that investment really began in 2010, certainly, through to today, with the acquisition we made of Renesas that we closed about 2 months ago. And you can see, underlying that 630 basis point growth is actually a step-up much higher than that and then a drop-down as we made that significant investment in Mobile & Wireless.

So let's look at that in even more detail. So if you look at that 940 basis point decline -- or 960 basis point decline in Mobile & Wireless, what that's made up of is principally an investment in the baseband side on the LTE side. Now I can tell you that, in fact, that investment is somewhere between 1.5x to 2x the dollars of the operating margin drop that you see in that time period, reflecting really the underlying strength of the connectivity business behind the investment in LTE. So just like broadband and infrastructure are very strong businesses, so is that connectivity business, which Scott talked about and Bob talked about. We're up here talking about the platform itself.

The other point I want to make is at the bottom of the chart. Last year, I said this business was probably 3 to 6 points of corporate operating margin in terms of the loss associated with our cellular business. Today, I would say it's probably closer to 3 to 8, and Q4 probably a smidge above that, right, as we sort of have the period of time where we've absorbed the maximum amount of the cost structure associated with Renesas with no revenue in Q4.

And so you can see the nature of the amount of investment the corporation is bearing and that Mobile & Wireless is bearing. And I think I also put because of the fact that it represents roughly 50% of the overall mix of the business, you can see the number of operating margin points that, that investment reflects in Mobile & Wireless, so maybe some sense of the health of the rest of the Mobile & Wireless business outside of that investment.

So we get asked, "Why are you making this investment? Gee, it's high risk. What's the real value of it?" And so whether you look back in Broadcom's history to our investment in DSL, when people said you'll never make money in DSL or in Bluetooth, which was probably almost on the portfolio cutting floor probably several times, wireless LAN, which was almost on the portfolio cutting floor as well, there were numbers of businesses that we have made investments in, where we have, I think, accurately predicted the competitive dynamic. Not in all cases, but certainly, in DSL, Bluetooth and wireless LAN, where we've accurately predicted the competitive dynamic and then the ensuing economic dynamic that goes along with that. So what I've tried to do is put the wireless LAN example off for you. Now I've run through 2011, we don't measure it as a wireless LAN business anymore. We measured it much more as a market segment in connectivity. But be as it may, this is where we have the best data.

So if you look back at 2006 when Broadcom was really starting to really build its position in wireless LAN, there were probably 8 competitors in the market. So what I've tried to do is put the wireless LAN example up for you. Now, I've ran it through 2011. We don't measure it as a Wireless LAN business anymore, we measure it much more as a market segment in connectivity. But be that as it may, this is where we have the best data.

So if you look back at 2006, when Broadcom was really starting to build its position in wireless LAN, there were probably 8 competitors in the market. There were 2 very big competitors in the Intel and Agere. The market share of the top 3 competitors was about half of the market. If you roll the clock forward, right, to 2011, 5 years later, the market had consolidated, right? It consolidated around us, it's consolidated around the Atheros. It consolidated around TI and maybe a little bit in Marvell and then a number of, well, collation competitors that were really not technology players in the space. So, really, 4 technology players at that time. And the market share that existed in the 3 largest players was close to the 95%.

And as the market consolidated closer to 95% across those 3 larger players, the gross margin began to expand. And the gross margin of our business over that time period actually expanded approximately 500 basis points over the time period. Now, the operating margin impact of that, the leverage drop through impact of that because of the critical mass of the R&D investment and the need -- and not having the need to substantially expand R&D once you've reached critical mass. Similar to where we are in the investment for LTE, the operating margin expansion was more than 3x that. So, roughly 1,800 basis points. And so while not a perfect example, certainly, an illustrative example of the nature of the leverage assuming we deliver on the milestones we've set, and I think you can see that we are showing good traction on those milestones and are in fact slightly ahead, of the kind of leverage we could see from our investments in the baseband.

Let me shift gears. So now I want to talk a little bit about managing tightly in this window. And so if you look, and you look at the performance of the wired businesses, what you can see in the wired businesses is you can see world-class profitability, world-class market positions and I think with some perspective on how we're doing in mobile and wireless, you can actually see some underlying real strength in mobile and wireless as well. And why does that come? That comes because we continue to manage tightly. We manage tightly in terms of the portfolio process we follow, in terms of the way we think about our operating expenses, and I'll talk to you more about how I think that will play over the course of the year. We've committed to you in terms of a set of objectives on stock-based compensation and moved aggressively on stock-based compensation and certainly, the topic is euro in terms of return on capital, we've been aggressive on our view of return on capital in terms of how we think about our U.S. cash flow and the ability to actually return capital.

So let me talk a little bit about portfolio management. So I think I showed this chart probably first in 2008, it actually had 3 colors on it at that time. We were probably closer to 18 businesses and we had splitted into 4 quartiles, now we use 3, and there's -- today, there's 15 and, yes, I do know there's 2 sevens and 2 tens on there, that's because they're tied. So it's not just that I can't do math, it's actually because they're tied.

And so what you'll see here is that we run the same portfolio process on an annual basis such not that we don't check the milestones of every business and the traction of every business as we grow across year, but we run a very formal process on an annual basis. And you'll see here that once again, the 2 businesses that are principally right on the bottom have been targeted to reduce investment and no, neither one of those 2 turns out to be cellular, they're actually other businesses, and we're going to redirect that investment to other parts of the company so that we can actually continue to grow our business, continue to have the best returns, continue to drive performance in the places that we believe produce the best returns for the company. And so, one of the ways you can do this is because you actually have a zero-sum view in the way you actually take those -- the view of your investments.

Moving to operating expenses. So what I've done is I've put up a trend of our operating expenses, and I think it's important, again, to go back and just reiterate the point. So I started back in Q4 of last year and rolled it through sort of Q1 of next year. And I do that for a couple of reasons. The first and probably most important one is, just to remind you again, that we do have a normal step up associated with fringe and merit and it's in the vicinity of about $40 million, typically, that we see due to the accounting for fringe and merit that we see step up in Q1. Now some of that rolls into Q2 because the merit increase usually occurs in the third month of the first quarter and you get 2 months in the second quarter, but you typically see it there.

Over this time period, from Q4 of '12 to roughly Q4 of '13, we will have added roughly, including the impact, the netted impact of the Renesas acquisition, probably in the vicinity of 1,400 people. So what that does is that does create a growth in operating expenses sequentially that we see across 2013, some of which you just don't see.

Now, the biggest step of that this in the step that we provided guidance for between Q3 and Q4 of 2013 where we guided originally up $40 million to $60 million. And today, we took that down to $30 million to $50 million and taking that down about $10 million as we continue to manage tightly. So you do have that effect. Now, as we roll into next year, we will have the same effect of the annual fringe and merit step up and call it, ballpark, $40 million. And then through the rest of the year, plus or minus the wrap around effect of the merit increase in Q2, I expect, because as Scott mentioned to you, we have put in place our resources for next year. In fact, if anything, we'll probably trim some of the transition services that we've got coming in from Renesas, offset by increased tape-out cost in 16-nanometer and a variety of other things, that our operating expenses over the course of next year will be flattish. Flattish to down-ish. They'd bounce a little bit in the quarter and trying to predict these numbers which are in the $700 million range, plus or minus $10 million or $20 million, it's pretty tight but fundamentally, we continue to believe that we'll manage the business very tightly as we go into next year, excluding acquisitions or anything else that could sort of substantially change things, but the plan of record looks something like this.

Stock-based compensation. So, the first time I stood up here, there was a lot of discussion about stock-based compensation, and since so many of the Oracle use my Ghostbusters joke, which is, at the time, our stock-based compensation was above the covers, 4 feet above the covers. And hopefully now, we are down sort of down to where the covers are, and actually have dropped stock-based compensation down to -- down about 50%, not just as a percentage of revenue but also in terms of the actual grants per employee as we sort of level set our compensation more towards the cash side and gotten things more aligned with the market. And I think that as you look at these numbers, we will continue to have that number drift down over time, and certainly into 2014, as we continue to move and align our stock-based compensation more with the peer set and competitive group that we actually compete for, for talent.

Finally, cash flow. So I need to digress for a moment just to make a point and I believe most of you understand this, but I want to say it anyway, which is, companies are only permitted to buy their own equity or pay dividend, U.S. companies, not -- non-U.S. companies don't have this constraint. U.S. companies, which we are, are only permitted to buy their own equity an issue dividends out of their U.S. cash. And hence, their U.S. cash flow. And as a company before Broadcom, our U.S. cash flow is roughly 30% to 35%, on a free cash flow basis, of the total cash flow. So the nature of our structure has a lower U.S. cash flow perhaps than other companies. We do benefit from a lower tax rate, but we have a lower U.S. cash flow.

If you look at the time period from 2007 to today, we have returned more than 100% of our U.S. cash flow. And that's because we were carrying excess cash on our balance sheet. We have returned a fair amount of that excess cash. You can see the flattening in this chart from 2012, and that flattening reflects the acquisition of NetLogic where we were not in the market buying shares. And so as a company, we have tried very hard to be responsible with our U.S. cash flow in terms of returning it to our shareholders. And our plan, net of debt service, is to continue to do that. We tried to issue dividend somewhere between 40% and 50% of our U.S. cash flow. We tried to buy back with the remaining amount of money, plus or minus M&A that's going on. But fundamentally, we are committed to returning capital both to our equity shareholders and to our debt holders, and so our plan is to return meaningful amounts of cash to our investors as a percentage of our U.S. cash flow.

So in closing, really 3 points. This is a company which has extraordinary economics in its core businesses. They're best-in-class market position businesses, they're best-in-class economics positions, they're best-in-class technology positions. And those economics enable the company to make meaningful investments in new areas and new opportunities in one of the largest or will be the largest semiconductor markets going forward in terms of growth and opportunity in the wireless space and in particular, in LTE.

Assuming we hit our milestones, assuming we're on plan, and you can see today that we are even slightly ahead of plan, we believe this investment would drive meaningful leverage through our P&L and meaningful value for our shareholders, and hence, why we're doing it. Now, we know we're in an investment cycle. We know that we are trying very hard to build a position in a market which is larger, perhaps than the other invest that we've made in terms of DSL and Bluetooth and Wi-Fi. And so as a result, we'll continue to manage tightly. We're going to manage tightly on the portfolio, we're going to manage tightly on our operating expenses, we're going to manage our stock-based compensation tightly, and we're going to be very, very careful and thoughtful about how we return capital to our shareholders in terms of dividends and share buybacks.

So with that, let me invite Chris back up here and the rest of the management team up here for a broader set of Q&A.

Chris Zegarelli

Perfect. So we'll now conduct a final Q&A session. There's a question in the back, right over here.

Unknown Analyst

Thank you. First question, maybe, for Bob. The Samsung phone, the LTE phone, is that going to be a worldwide phone or is that going to be specific to just 1 region?

Robert Americo Rango

I would consider it a global launch, doesn't mean it will go to every single region like the U.S. but it will certainly go to places like Europe, South America.

Unknown Analyst

Okay. And then for Eric, with the updated guidance this morning and the upside in networking and infrastructure, what specific area within that drove that upside?

Eric K. Brandt

Well, we continue to see strength in the data center business, as Rajiv was talking about. I think we're hopeful longer term on the service provider. But principally, as we said, it was across all businesses and a little bit stronger in the infrastructure side.

Chris Zegarelli

Okay. There's a question up here. Vivek from Bank of America.

Vivek Arya - BofA Merrill Lynch, Research Division

A question for Eric. Actually, 2 questions. So first is, could you just again help us understand what the business model is? Are you comfortable with the long-term operating model that you had set out before and how soon do you think the company can get back into that model range?

Eric K. Brandt

We are -- we haven't change our operating model and I think we'll be right around -- if you plug in the guidance we gave this year, we'll probably be right around 20% if you net out the effect of the acquisition assumption that we took in the fourth quarter. So we are in an investment cycle. I think we are probably in that investment we are making in cellular at, what I will call, near the bottom of that ladle between Q4 and Q1. And assuming we see that meaningful ramp in revenue as we sort of get through the middle of next year and the back part of next year, I think you'll see this operating expense cost in mobile and wireless begin to shift down towards the lower end of the range.

Vivek Arya - BofA Merrill Lynch, Research Division

Got it. And as a follow-up, one of the charts that you showed, the optimizing investment. I'm trying to understand the revenue CAGR. So the top growing parts of the business are sort of shown at the bottom of where they are generating value and some of the higher value parts have much lower revenue CAGR. So what should we make of that?

Eric K. Brandt

So what typically happens there is that, as you can imagine, if you look at a number of those metrics, they favor businesses that drive significant economic value or economic power. And so the first column looks at revenue CAGR which tends to typically get smaller parts of the business. Frequently, we find many of those companies, those LOBs in the company wind up being sort of a promise for the future and a continued promise for the future. So they frequently wind up sort of always having higher revenue growth rates, but not moving up very high in the economic ladder. And so the way to think about that chart is that chart doesn't make decisions. That chart forces conversation. And so to the extent businesses have stayed in that spot of the chart for multiple periods of time, then when we actually do a portfolio balance, you actually have to begin to start making adjustment for what you believe the forecast accuracy of those businesses are and the degree to which they could drive economic value. And so it does -- it's one of the metrics that's really designed to favor the smaller businesses, so you can actually see where they are.

Chris Zegarelli

And let's just stay right behind the back with Ross.

Ross Seymore - Deutsche Bank AG, Research Division

Question for Eric, Ross Seymore from Deutsche Bank. Actually, 2 questions, quickly. One, the clarification on the OpEx cut. Is that now the new lower base off of which we step up in '14 or is there some temporary aspects?

Eric K. Brandt

No, it's the new base. It's the new base you should assume for the step up.

Ross Seymore - Deutsche Bank AG, Research Division

Great. And then a follow-on on that topic. If we think about fourth quarter next year exiting the year, with what you said in the past you're going to trim on the Renesas side, about $40 million or so, would you expect the OpEx, although flat sequentially through the year, of that first quarter, could actually be down year-over-year?

Eric K. Brandt

Year-over-year, '13 to '14?

Ross Seymore - Deutsche Bank AG, Research Division

Correct.

Eric K. Brandt

No. I don't. And the reason why I don't is because, as I mentioned, we hired probably in the vicinity of 1,400 people, including the Renesas people from the beginning of the year to the end of the year and the slope of that and the wrap around effect of that will not, in most -- in all likelihood, mitigate the actual expense delta between year-on-year. So remember, you get the slope, so you only get half of the cost in the current year and because we have a fairly large step up in Q4, you actually get more of that into next year.

Chris Zegarelli

There's another question. It's a little bit further down in the same row.

Joseph Moore - Morgan Stanley, Research Division

It's Joe Moore from Morgan Stanley. So last year at this meeting, you guys were fairly enthused about the BCM LTE chip. And the feature set looks pretty similar to the chip that you're talking about, sampling on the [indiscernible]. So I understand the benefits that Renesas gives you in terms of the midrange LTE, but can you talk about the high end and have you sacrificed anything from the time line for features like carrier aggregation relative to where you thought you were?

Eric K. Brandt

Yes. I can answer that. So certainly the decision to move away from the BCM approach and move on to the Renesas mobile approach was really driven by the recognition that the LTE markets happening a lot faster than any one in the industry had expected. And the modem that we acquired was just more stable and more mature than the one that we're working on. So it's really just a pragmatic business decision to move ahead. As I explained at my presentation, I think the IP that we acquired is actually in a much better shape than the IP that we gave up with the BCM acquisition. And the team that we acquired had started earlier, have worked on a lot of the advanced features long before we have started on. So it was just, in my opinion, a pragmatic business decision to go get into the market as quickly as possible and to start proliferating that roadmap story that I talked about. Our customers are already on 3G. That same framework now is going to get loaded on our 4G chip and they can easily go from our 3G solutions to our 4G solutions. It's a good story.

Chris Zegarelli

There's a question right here in the middle, on the right.

Timothy M. Arcuri - Cowen and Company, LLC, Research Division

It's Tim Arcuri at Cowen. You've put up slide, Henry, that was talking about less of your revenue pushing the bleeding edge. And I'm wondering if you can quantify, maybe if you look at today, how much the revenue is pushing bleeding edge and if you look out say, 2 to 3 years, how much will push bleeding edge? And maybe, Eric, how much could that help margins over time?

Henry Samueli

It's hard to say in a fraction of revenue. I was looking at it more of a -- from a product portfolio perspective. So more of our products are going in. Different products, obviously, much different revenue profile. So I didn't analyze it from a revenue perspective. But yes, the trend is absolutely true that we will stay in 28 longer over time and you won't push the bleeding edge on as many products, but Rajiv, for example, will push the bleeding edge on numerous of his high-end products which generate a substantial portion of his revenue in his business. So I don't have a cut as to the quantitative answer to your question, but the trend is definitely there.

Eric K. Brandt

So today, a little over 40% of our revenue is in 40-nanometer or a little under 40% is in 65-nanometer. Typical trend is from the time we start shipping in production of a node, that it takes about 3 years to cross over the prior node. And that was true in 65. And that was true in 40, and well, absent any other data, I would assume that, that will be true in all 28. And now that we are beginning to ship in 28, I would expect 3 years for those 2 lines between 28 and 40 to cross over. So '14, '15, probably in '16. In terms of gross margin, we've done a number of things to work on gross margin. I think within the businesses, and broadband is a great example, broadband has aggressively gone after improving the die size, the efficiency of the chip, looking for opportunities to shrink and improve the underlying margins of business, which enabled them to fund the R&D and the cell funding in a lot of respects in terms of what they do in the market. And wireless LAN is an example. I think the other place we've seen it is as we've grown our infrastructure business from what was roughly 60%, 70% of the size of broadband now to almost the same size of broadband and the strength of that business has improved the gross margin. As we move down the process nodes and as we optimize between 40 and 28 and then 28 and other nodes, I suspect there will be similar other opportunities. And I think you can see that in the gross margin today in terms of the work we've done. I can't forecast exactly what the savings will be in the future, but I do believe there are opportunities to enhance the gross margin and certainly offset the ramp of some of the lower gross margin business we could see in mobile and wireless.

Chris Zegarelli

There's a question over here on the front, left corner.

Robert Sales

Bob Sales, LMK Capital. Two questions on wireless side. First of all, can you give a little more detail about how you're merging the 2 baseband groups because now you -- it seems that you have a bunch of really smart people that did the original design there, it's unclear what they're doing now. And then a bunch of smart people in Europe that you bought, and it would be interesting to see how you're merging those 2 groups. And the second question is, what you have offered as your position on the -- on your position in the handset for Wi-Fi seems counterintuitive to what Broadcom's strengths have been. In other words, once we get past 802.11ac, it seems you will be vulnerable towards more Wi-Fi functionality being sucked into the baseband when you're seem to be professing a discrete solution that may become vulnerable, maybe not in 2014 but towards the end of 2014.

Henry Samueli

Okay. The first question is, how did we integrate the teams? Did we have 2 modem teams? The answer is no. As I think Eric articulated when we -- after we did the acquisition, we did do a headcount reduction. And we kept part of our original team to work on things like TD-SCDMA and Advanced Envelope Tracking Technology. So the places where they were not overlap was the places that we kept our teams. And this is just on the modem team. Of course, cellular team is made up of the processor team, the image signal processing IP generation team, the graphics team. No changes to any of that. So, really, just impacted of the modem IP team. I think we did a very smart integration in terms of what we kept and what we actually jettisoned off. And like I said, I think we made a smart pragmatic decision to just move to IP base that was more advanced in terms of 4G LTE. So I hope I've answered that question. So I don't think there's very little overlap between the 2 teams now since we were working on TD-SCDMA for quite some time before the Renesas mobile team.

With respect to the Wi-Fi comments that you made, first of all, I want to make it clear that on our complete solutions, we are not being out-integrated. Even though we haven't gone an integration of Wi-Fi onto our cellular baseband yet, we have the same number of chips in our platform as any other of our leading competitors, okay. Because we have more conductivity integration on the combo island, we have the same number of devices as any of the other companies that you're thinking about, okay? And what we've been saying for a number of years is that the cadence of connectivity is moving different than cellular. Especially -- and in each segment of the market, it will move at a different cadence. So I'm not saying we're not going to integrate connectivity for certain segments of the market, but for the high end, we're going to keep driving connectivity features that are going to enable new use cases, and there are new Wi-Fi standards coming. The Wi-Fi technology is not going to end with 802.11ac. There's always a new standard. We're driving those standards in places like IEEE and the Wi-Fi Alliance, okay? So there's more stuff coming, and that will also dictate when and how we do the integration between Wi-Fi and cellular SoCs. And again, there's new technology coming even beyond that, right? I mean it's not just Wi-Fi and Bluetooth. There's NFC, the GPS technology, I mean, there's a lot of different pieces that have to be integrated at the right time.

Chris Zegarelli

There's a question up here in the front, on the far left. Anil?

Anil K. Doradla - William Blair & Company L.L.C., Research Division

Yes, this is Anil Doradla from William Blair. Question is to Dan. Dan, if you look at the cable industry, obviously, last full months have been pretty exciting. We've seen many upgrades going on. You talked about Rogers, I think they're even doing 1 gigabit or something like that. But can you help us understand what are the perhaps technological limitations for the cable industry? And we're going from 8 channels to 16 to 24, do we stop there before we go to the fiber? Or do you think, I mean, this trend continues on over the next couple of years because these guys have to catch up with the fiber guys, right?

Daniel A. Marotta

Well, I think when you go to DOCSIS 3.1, which is a follow-on to 3.0, that there's virtually no limitation. I mean, the 3.1 physical layer is so powerful. I mean, here you got white space elimination, you got huge bands that just have a staggering throughput. So I think with that standard -- and of course, that just overlays on top of standard piece of coax. It's a staggard amount of bandwidth and they'll compete fine with fiber. Remember, I mean fiber guys can't just give away their bandwidth and an unlimited...

Henry Samueli

DOCSIS 3.1 can accommodate between 5 and 10 gigabits per second over the full coax bandwidth available. So definitely multi-gigabit per second services will be available with DOCSIS 3.1.

Robert Americo Rango

Right. So, it's huge. And I think the phone guys are going to probably go hybrid because they have a little bit more bandwidth challenge, and so, if you use G.Fast in combination with many fiber nodes, so that will be -- that's actually more interesting for us because you get multiple modems in the path. And of course, we know the fiber story. So I think it's very, very good trend. And this is happening next year, right, so technology will be introduced next year. And then, of course it takes time to propagate. But I think the communication engineering people are doing our part.

Chris Zegarelli

There's a question right in the back row.

Unknown Analyst

[indiscernible] with USAA. So you talked about the significant reductions in stock-based compensation since 2008, I'm curious what do you think is a reasonable level and how much more runway is there on that front?

Robert Americo Rango

So if you look at an average of the peer set, they're probably running high 3s, give or take, as a percentage of revenue. That's probably not a reasonable assumption. We'll probably be slightly above that. Because I think we're more equity-centric, but we're not certainly going to be anywhere near where we were in the 2007, 2008 timeframe. And you can see, if you adjust it for NetLogic, we're actually getting into the 5. So we're getting to the flat part of the curve.

Chris Zegarelli

And we have time for 1 more question before we move on to the reception. Let's end it in the back row. I believe it's John Pitzer with a question.

John W. Pitzer - Crédit Suisse AG, Research Division

John Pitzer with Credit Suisse. I want to go back to Henry's presentation about the growing importance of scale, the amount of IP reuse you have between divisions. If there comes a point in time for either technical reasons or economic reasons that you feel like the baseband market's not a good market to pursue, can you have a viable connectivity business without baseband, and without connectivity, can you have a viable broadband and ING business, i.e. the scale imperative, does it means that you just have to get the baseband right or there are other offsets like do you run out of things that will drive scale and allow you to continue to move down the argument that Henry made in his presentation?

Henry Samueli

Well, certainly, there is a lot of leverage between the groups as you saw in all of the presentations. Dan has Wi-Fi being integrated in its gateways and enterprise Wi-Fi, et cetera. So clearly, you want to have that, which is why we're driving so hard to achieve success in the business. But worst case scenario is if you have to extract businesses out Broadcom, I mean, it's doable, it's a very difficult task and not a desirable task, but nothing is impossible to do. But it's clearly not our objective. We want to drive to success so we can use that leverage across our businesses and create a Broadcom platform that's very powerful. So we don't want to have to plan for our worst case scenario. We want plan for success and success means driving that leverage across all of our businesses.

Chris Zegarelli

And with that, thank you, all, very much for coming today. And I'd like to leave you with a few thoughts. We have some milestones going into next year. And in the home, I think some signposts you should look for are, do we continue to grow in these emerging markets? That's going to be one of the biggest growth drivers for us in that space. Do we ramp these new technologies like HEVC, vectoring, UltraHD and so forth? So are those things that we drive? So look for that going forward. In the end, I think it's very clear we want to accelerate our connectivity business with innovation, okay, and continuing to drive penetration of some of the new features like 5G WiFi. And then of course, material LTE revenue, you should look for there and customer rollouts. In infrastructure, driving innovation in the data center and roll out of the service providers and accelerating some of the new market opportunities that Rajiv talked about. So those are the signposts for us next year. With that, please join us at the reception. Thank you very much. Have a wonderful holiday and a prosperous new year.

Copyright policy: All transcripts on this site are the copyright of Seeking Alpha. However, we view them as an important resource for bloggers and journalists, and are excited to contribute to the democratization of financial information on the Internet. (Until now investors have had to pay thousands of dollars in subscription fees for transcripts.) So our reproduction policy is as follows: You may quote up to 400 words of any transcript on the condition that you attribute the transcript to Seeking Alpha and either link to the original transcript or to www.SeekingAlpha.com. All other use is prohibited.

THE INFORMATION CONTAINED HERE IS A TEXTUAL REPRESENTATION OF THE APPLICABLE COMPANY'S CONFERENCE CALL, CONFERENCE PRESENTATION OR OTHER AUDIO PRESENTATION, AND WHILE EFFORTS ARE MADE TO PROVIDE AN ACCURATE TRANSCRIPTION, THERE MAY BE MATERIAL ERRORS, OMISSIONS, OR INACCURACIES IN THE REPORTING OF THE SUBSTANCE OF THE AUDIO PRESENTATIONS. IN NO WAY DOES SEEKING ALPHA ASSUME ANY RESPONSIBILITY FOR ANY INVESTMENT OR OTHER DECISIONS MADE BASED UPON THE INFORMATION PROVIDED ON THIS WEB SITE OR IN ANY TRANSCRIPT. USERS ARE ADVISED TO REVIEW THE APPLICABLE COMPANY'S AUDIO PRESENTATION ITSELF AND THE APPLICABLE COMPANY'S SEC FILINGS BEFORE MAKING ANY INVESTMENT OR OTHER DECISIONS.

If you have any additional questions about our online transcripts, please contact us at: transcripts@seekingalpha.com. Thank you!