Seeking Alpha
We cover over 5K calls/quarter
Profile| Send Message|
( followers)  

Intel Corporation (NASDAQ:INTC)

Citi 2013 Global Technology Conference

September 4, 2013, 9:00 AM ET

Executives

Stacy Smith - Executive Vice President and Chief Financial Officer

Analysts

Glen Yeung - Citi

Glen Yeung - Citi

Good morning, everybody. Welcome to Citi's Global Technology Conference. I actually had to look and read that because I can't remember where we are. My name is Glen Yeung, I am Citi's Global Head of Technology, and we're very happy this morning to be speaking to Intel, a company as so you all know very well and you also know the gentlemen to my left is Stacy Smith, the CFO of Intel.

This will be in the format of a Fireside Chat, we will leave a few minutes towards the end of our discussion for any Q&A from the audience, so if you have any burning questions, start thinking of them, now.

Question-and-Answer Session

Glen Yeung - Citi

Stacy, we'll just start right into it, and we're asking at the topic the couple of questions, as to all the companies. The answer can be brief. Would you characterize your overall visibility as better or worse as compared to a quarter ago or a year ago?

Stacy Smith

Does the answer have to be brief?

Glen Yeung - Citi

It depends.

Stacy Smith

Okay, the brief answer is relative to a quarter ago, I think it's better. And it's better because the macroeconomic environment seems a little bit more certain, and certainly if you go back 12 months, and if you think about last summer, we had concerns about what was going on in Europe, concerns about China slowing. So I think from that perspective its a little bit better.

I think the more interesting trend is what's happened over the longer period of time, and that's really what's driving the industry in is -- the technology has become mobile, it's become personal and the trends are being dominated by the consumer. In that consumer domination of the buying pattern means that I think demand is inherently more difficult to call.

Its spread across different form factors and different devices and so just generally it's an environment where it's hard to get your arms around demand. And for company like Intel what that means is we have to be more responsive, we have to have low inventory levels, we have to work closely with the supply chain to be as responsive as we can and have world that's just more uncertain.

Glen Yeung - Citi

Has the shift in geographic dominance of demands, for example, now it's much more Asia-centric than it has been in the past, different holiday schedules for example, has that also altered the seasonality and to some extent sort of made it harder to predict?

Stacy Smith

It has altered the seasonality some, but I don't think that in itself makes it harder to predict. I really think it's just the fact that the markets are dominated by the consumer buying pattern tends to lead to spikes and it tends to make it difficult to call between the form factors.

In a very specific, in Intel for us, a year ago, 18 months ago, our product line was really only in PC and we didn't have products that played very extensively in tablet, so first it was almost digital. If we were wrong between the mix between PCs and tablets, we had a lot of variability. Today because we have a product line that expands across both, now it's really a question that we're going to sell more tablets or sell more PCs, and so it was digital and it's more of a mix question for us.

Glen Yeung - Citi

Second, sort of generic question at the top is, would you characterize your opportunities for products to regain within Intel as better or worse than a year ago or a quarter ago?

Stacy Smith

Well, the thing that's good, opportunity for productivity growth and that -- we very consciously invested ahead of the curve, in terms of building a product line in tablet, investing in some SoC capability that will be targeted at tablets and phones. Software, some of the stuff we've been doing in process technology.

Our spending as a percent of revenue is high. And as you saw on our last earnings, we're going to bring it down some over the back half of this year. And the expectation that Brian and I have is that we can bring it down more over time. And so I think that the most macro level spending as a percent of revenue is a measure of productivity and we'll be able to improve from where we are today.

Glen Yeung - Citi

Brian, for those who don't know is, Brian Krzanich is the new CEO of Intel, and only the sixth CEO in the history of the company. How do we think about the implications of having Brian in? What's he going to do in terms of strategy? What will be different under Brian?

Stacy Smith

I think he is building on the strong leadership we had under Paul, where Brian is different. Paul was very much a market guy, Brian is a deep technologist. And so I really see Brian focusing and pushing our strategy forward relative to where we were is on how do we take advantage of our manufacturing lead to have the best performing lowest power, most cost efficient product, but with a very specific emphasis around SoCs that are going into lower-cost lower-powered devices.

And using our manufacturing advantage to do more integration of the faster cadence, which we think gives benefit to the people who buy from us and also as a place where it becomes hard for our competitors to follow. So he's really focused on that SoC execution and integration.

Glen Yeung - Citi

And about Renee's role, now that we're talking about new management, what's her role going to be in?

Stacy Smith

Yes, so you have the CEO and the President's role, she is directly responsible for the manufacturing organization and the software organization and then where I see her playing a pretty deep role is around kind of the strategy of the company and some of the commercial deals that we want to go do and just think of it is a force multiplier given the complexity of the environment that we're in.

Glen Yeung - Citi

Okay, we've now got Haswell in the market, it's the greatest change in power efficiency that we've seen in an Intel-architecture, is this going to drive an upgrade cycle in PCs?

Stacy Smith

So first talk a little about Haswell, it's a phenomenal product. As you said it enables the huge gain in energy efficiency. It's on the order of a 50% improvement in battery life, and so -- and it's got a thermal envelope, it's a very small form factor package and so what this enables at a physical level is, it is very thin and light PCs with stunning performance and battery life -- extra battery life that will be an excess of eight hours.

So think about Core i5, i7 level of performance with the graphics to go along with that and eight hours of active battery life and a great standby battery life as well. It's not just about Haswell, so now what you're seeing is that's unleashing a lot of innovation within our customer base to bring two-in-ones to the market places, so touch-enabled devices, detachables, convertibles, where you get best of a PC and a tablet in one.

I think the combination of all of that, certainly sets the stage for a lot of volume growth in the two-in-one category. We have other products so it can enable lower price point. So there are segments of the markets where I think we're likely to see good volume growth, but also want to -- so that's against the backdrop of a forecast for us that is flat revenue year-on-year, so we hope to have more growth in the future, but right now it's certain categories where I think we'll see a lot of growth and other categories like the traditional PC segment that we think declined this year.

Glen Yeung - Citi

In terms of the designs that you're actually seeing coming to market for second half, you know, Haswell is going to give them as you say the opportunities for new things like the convertibles, are you kind of happy with the level of design activity that you're seeing? Are you seeing the kinds of systems that you envisioned with Haswell?

Stacy Smith

Yes, the stunning designs. We did a show-and-tell for the board not too long ago, where we did -- there is dozens of designs coming to market and when you just kind of go OEM-by-OEM, design-by-design and you're looking what's coming to the marketplace, it's unlike anything I've seen in my 25 years. I think the question as we talked about it in earnings is what price points will those enable on shelf and how aggressive will the customer base be in terms of getting the assortments and SKUs out there and I think that story plays out over our September and October where we'll really start to see that.

Glen Yeung - Citi

Was there a learning curve that the OEMs needed to have with respect to Haswell and what they can do with it? I know that you guys had some interesting designs yourself that you -- reference designs that you could show that were pretty cool?

Stacy Smith

Yes, absolutely. And one of the reasons that we picked up ultrabooks with Ivy Bridge, knowing that Ivy Bridge wasn't the perfect product to go into an ultrabook form factor, just to enable it to an essence, pave that road around the design work that we wanted to partner with our customers to do, so absolutely. And it's not just from our customers it's from our standpoint as well. We wanted to have that year of runway in anticipation of Haswell to work on how do you get the BoM cost down, how do you get the touch to work really well, how do you work with some of the mechanical designs when you're dealing with the two-in-one, and so I think that's held us and our customers in good stead.

Glen Yeung - Citi

I wanted to talk about Bay Trail, but just one last thought, which is let's not use the term ultrabook, but just that ultra thin and light form factor, do you anticipate that that will be the vast majority of what we're selling a year or two years from now making the bricks that we're dominant a year or two years ago, basically obsolete?

Stacy Smith

Yes, I think if you just watch the trends, and I said this in my first question, I think that the trends that are dominating the industry right now are shift towards mobility, the shift towards personalization of technology and consumer. And you just look at how the traditional notebook computer has been reinvented to be this very thin and light, detachable or some other mechanical way to get the best of the tablet and the best of the PC in one, I think that trend just dominates well in time, whether it's branding an ultrabook or it's just a thin and light notebook with touch, I don't know, I don't really care, but that's where the industry is going.

Glen Yeung - Citi

So Bay Trail then gives you the opportunity to address different type segments in the markets, maybe it's just talk about that, what does Bay Trail allow Intel to do in terms of where you can play?

Stacy Smith

Yes. So Bay Trail, I'll backup a little bit, because Intel typically has been one core for the PC market. We now have multiple cores that are -- and there is overlap between them. So it's worthwhile to just define Bay Trail. Haswell, as we've talked about is a core, very strong performance, very strong graphics performance, a lot of features that's going into the higher end of the PC segment.

And Bay Trail is an Atom-based core. It's derived from the Silvermont architecture, so it's 22-nanometer part. And Bay Trail is targeted at tablets and then it will also serve the low-end of the PC segment. And so think of it is as having spectacular battery performance and really good performance, but not nearly at the level of where Haswell is. It will enable in the PC segment of the market. So think of a lower-end clamshell, thin and light notebook with touch. It will enable those to be in the high $200 by this Christmas selling.

Glen Yeung - Citi

With touch?

Stacy Smith

With touch. And in the tablet segment, it will be in the high $100 and it kind of coming down from there and so what it does for us, is that it enables us to start hitting these consumer price points in a way that we really haven't had the cost structure to hit in the past, and that does it with great battery life and a good amount of performance.

Glen Yeung - Citi

In some way, it's addressing the market that you have not previously focused on. It was either AMDs market or in the last two or three years, it was really left to his really tablet market.

Stacy Smith

Yes.

Glen Yeung - Citi

I mean is that the way to think about it, I mean if you didn't do it somebody else was going to do it, so you better focus.

Stacy Smith

Somebody else is doing it, right. There is a lot of tablets that are selling in the world today. Bay Trail is going to be a great solution there. And so, yes, we're aggressively going after that segment of the market.

Glen Yeung - Citi

And Bay Trail like Atom, designed for cost.

Stacy Smith

It's the cost and -- so when you -- with our process technology, we can have performance leadership, we can have power leadership and we can have cost leadership. Typically, when you're thinking of a specific product, you can optimize two out of three of those. And so with Bay Trail, we're optimizing power efficiency and cost. With Haswell, we are optimizing power efficiency and performance.

Glen Yeung - Citi

So when we talk about these products now we're sort of notionally focusing on the consumer applications for them. And at the top you mentioned that that sort of where the market is going. But what's the possibility that the enterprise market, which is something we don't always talk about, and it's definitely something where we are certain to see some marginal improvements, what's the possibility that that market actually starts to comeback and how might that change the mix, if it does?

Stacy Smith

Well, I think it's inevitable that that it will comeback. And I think the interesting question is when and I'm pleased to hear you, you know say it's now, that would be great. But the PC is a productivity device that hundreds of millions of people around the world use and value and cherish. There will be an upgrade cycle at some point.

And I think, for us, our job is to make sure that we've got great technology that gives the CIO the reason to say, hey, now is the time to go do it. We think Haswell actually brings that technology to the marketplace. There is other dynamics at work in terms of budgets and other priorities, but we think the technology is there.

Glen Yeung - Citi

So it lends itself to the question of whether or not the shift we're seeing in price points today, with some of these tablets would be just above $100, if you go to China today. Is that a permanent shift or is that really a function of the tablet, the very fast tablet evolution in the market and when that sort of stabilizes, we may actually see pricing creep back up?

Stacy Smith

I think that the -- if you think about the world from the standpoint of averages, you miss the important trends of what's going on. The way I view the market dynamics is I think we're heading into a world where you have a lot of segmentation of technology into different markets.

And so if you think of the classic PC segment of the market, where two-in-one's play, and I think that's a segment of the market again, there will be some growth there, it's likely to be relatively low growth compared to some of the other segments, but you'll see some growth there overtime, not this year, but I think overtime you will.

You're going to see other segments of the market like tablets where you see this explosive growth. And our job is to make sure that we've got products that have the right cost structure and the right set of features such that we can play in the different segments and then clearly differentiate, hey, when you're buying a $99 tablet this is what you get, if you buy a $200 tablet, this is what you get and if you're spending $300 on the Bay Trail clamshells, this is what you get and if you're spending $500 on a Haswell two-in-one, well, this is what you get and have those -- that segmentation playout.

Glen Yeung - Citi

So if I am spending $300 on a Bay Trail clamshell, what am I getting? The reviews of Bay Trail have -- I don't want -- they are little bit mixed, but relative to history, your historical processors, it looks great. So how do we think about that? You know, I read a review that said, Bay Trail of $300 is going to be like a Microsoft -- oh, sorry, a MacBook Air of 2010, it's that good. Is that that what you think about, what Bay Trail can do?

Stacy Smith

Yes, I don't know that I can compare it to specific product, but what you'll get at that kind of a price point is you'll get a touch-enabled thin and light computer that has great battery life and good performance. And then there will be a significant market below that and there is a significant sort of features above that, so if you go from the Bay Trail to the Haswell, then one of the questions is how do you create that segmentation. With Haswell, you'll get less battery performance, but you'll get 50% or more and more performance, you would get spectacular graphics performance because we put a lot more graphics cores on something like a Haswell.

And then you get a lot of features that you won't see with Bay Trail, so think of Turbo and Wide Eye and some of the virtualization technology that we would put in, and so that allows us to create a segmentation where enterprise likely you just going to want to buy a core, because they tends to be the use model and it gives some of the tools they need to manage the fleet of computers. Consumer will decide based on the performance level that they need between Bay Trail and Haswell and will upgrade products on both.

Glen Yeung - Citi

I wonder if you could talk a little bit about the roadmap of Atom. This is a design that came out a few years ago and serviced the netbook market initially, where can we expect Atom to go as we progress beyond Silvermont.

Stacy Smith

So I'll start with Silvermont and then we go to the next generation. So Silvermont sell in marketplace today. Silvermont is the core and there is lots of derivatives to Silvermont, so it's worthwhile to do a little bit of a tutorial. Starting in the server segments, so we'll have derivatives of Silvermont Atom going into the service segment, actually it's being launched today in San Francisco, called Avoton. That enables Microserver segment with the core there.

It's going to be great performance, it's industry-leading performance in that segment of the market, it will be the --by far the best performance per watt and we have leadership on performance per watt per dollar, so we're trying to be -- and we have server features on that segment, like 64 bit, some of the other RAS features that enterprise cares about. And so we're trying to be really aggressive of taking our own low power cores into that segment of the market, get there, years before the competition moves into that space.

Then you have Silvermont that's moving into the tablet and low-end of the PC market, that's what we've been talking about with Bay Trail. And later this year, you'll see Silvermont cores that are targeted at phone market that's Merrifield and that will be shipping later this year. Those are all 22 nanometer products. Then the 40-nanometer version of Atom, which is called the [indiscernible] is the root core, will start shipping next year.

And so you can start to see this acceleration of our Atom roadmap intercepting our process technology, and so think about the world where we're shipping 40-nanometer versions of Atom next year into the different segments of the market, in a rollout over time. But we think we're years ahead of their competition. And again that manufacturing will gives us the performance and power efficiency that others can't match.

Glen Yeung - Citi

Will it be a 10-nanometer where the core product and the Atom product will launch on leading edge and effectively at the same time?

Stacy Smith

Yes, it's definitely the places where Brian is pushing, if I look at where it is really pushing on the technology side, it's getting -- Atom is close to that leading edge node, as we can get it, and it's pushing our integration of technologies to be faster and more comprehensive than what the prior roadmap is. I think we get pretty close to the leading edge node 14. I think we're going to be right there at 10, but it's one of the places where Brian is pushing to see that if we can go even faster.

Glen Yeung - Citi

And can I go back to something, you said about Microserver, just some of the aspects that you said, is that relative to other x86 processors or are you including ARM cores when you make those ARMs?

Stacy Smith

The server market is an interesting market, you see the traditional competitors disinvesting, you see a lot of people that are targeting at least at Microserver segment of the market, which is 5-ish percent of the market, you see a lot of people coming after that. And our goal is to get there first and get there with such a great solution that it doesn't open up as seam.

Glen Yeung - Citi

And when we think about Microserver, it's beyond just process or technology at that point, then used to be phablet technology embedded in that, which you guys bought a couple of years ago, ostensibly, one would prefer to an SoC to a processor and chipset, all those things are in place, you feel comfortable with the technologies you have there?

Stacy Smith

Yes, yes. If anything I think that that place to our strength of -- we have decades of experience in the server market, so I talked about things like some of the RAS features, I talked about 64-bit, some of that connectivity, some of the phablet. We have all of those tools and we have deep engagements across the customer base where we can help them from the standpoint of how the datacenters gets architected, where we can provide real value to them, where they can build on the building block technology that we give them and implement some things that's quite a bit cheaper and more efficient.

Glen Yeung - Citi

And there is no need to recompile any software to go from a core type solution to Atom.

Stacy Smith

No.

Glen Yeung - Citi

Right, which we'd need to do. All right, let's think a little bit about the mobile, the handsets strategy or tablets strategy. And first of all, off the top, articulate that strategy exactly. What is Intel's -- what are they striving to do in mobile and tablet?

Stacy Smith

Win share.

Glen Yeung - Citi

Is there a target there -- is there a number that you need to be 10% market share in order for this to be a successful endeavor for you or some number?

Stacy Smith

Yes. There is. No, we don't enter a market to be small. We're a $55 billion company. We enter our markets to be a big player. We think that our process technology lead brings real advantage to our customers. And today our share in phones is close to zero. Our share in tablets is probably unfortunately 4%, 5% based on some of the recent designs, we've won like the Samsung GALAXY, but lots of room to grow there.

But I would expect that out in time, we grow the significant share and we do it because we can provide more performance, we can provide integration and it's interesting, kind of different from our historical position in servers or whatever. The world really wants the second source. And so what we're seeing from our customer base is a lot of interest in pulling us in and as fast as possible into the marketplace, because they want to have multiple suppliers there. And we plan to be one of those too.

Glen Yeung - Citi

One of the things we hear often from the OEMs is that -- the handset OEMs is that cell phones are increasingly becoming commoditized even at the high-end. And requirement from the OEM therefore is to work with guys you can provide a platform solution. And there's typically connectivity, AP and baseband.

Stacy Smith

Yes.

Glen Yeung - Citi

Are you sort of satisfied or confident with the offerings you have there, are the things you think you may need to acquire to broaden the portfolio?

Stacy Smith

No, I think we have -- with the exception of the LTE, which I will talk about separately, I think we have what we need. We're the number one player in RF. We're the number two player in basebands, we have a great suite of connectivity assets. I mean there is always things that you're looking to fill in, but those are more kind of tuck-in technologies that are not going to be big on a M&A scale or anything like that or that we can just develop organically.

So I think we have the pieces that we need. The place where we've been deficient is in LTE. And we're chipping multimode LTE today. We'll be shipping multimode LTE with voice by the end of this year and we'll see products on next year. So we're kind of right up to the finish line where that's kind of the final asset that we need.

Glen Yeung - Citi

Yes.

Stacy Smith

And then I think that I guess back then into our long-term strategy here is we take these assets and increase the pace that which we're integrating the technologies where we think we can give a form factors advantage to our customers or cost advantage or power advantage and it's something where our competitors will struggle to follow us. So that really is the strategy.

I'd also just say, I think the question is a profound question, because if you don't have all of those assets, it's hard to envision how you are competitive three and four years from now. I think you really do need to have all of those assets and be able to do integrations to have competitive products.

Glen Yeung - Citi

Well, so first of all, I'd like to think all my questions are profound. But let me just sort of extend that thought, which is -- okay, so you have the portfolio of technologies. You have the ability to integrate, one could argue Broadcom has that, QUALCOMM has that, if I stretch little bit, I could say, Marvell has that. I mean there are people out there who have it. But you then can add a manufacturing technology that you need. How important will that sort of fourth element be in terms of competitive advantage?

Stacy Smith

I think game changing. I think there are -- if you want to just crisply fare our strategy over the next five years, it's used the combination of our architecture and all of the IP blocks in our manufacturing advantage to create a set of products that allow us to win significant share across a really broad range of devices that are competing and that manufacturing advantage is the key.

Glen Yeung - Citi

If I look at a cost per transistor chart for Intel, I know you guys don't put a lots of details on the axes, but it doesn't look like a straight line when I go from 22 to 14, it looks like the improvement in cost per transistor is a little bit better at 14 than what I would expect on a linear trend? One, am I looking at that chart right; and two, what explains that?

Stacy Smith

I wouldn't articulate 14 as being better than the trend, but it's on the trend line. And I think that is a fundamental point, because I think there is one thing that's different for Intel than the people that we compete with, and the foundries that provide product for the people we compete, I think they're off of that historical trend.

And so just to backup a little bit, there's a cost per transistor curve, that really is the heart of Moore's law and it's that the cost per transistor comes down generation-to-generation, there's a couple of elements to go into that. One is the capital intensity. Capital intensity is measured by capital dollars per wafer per square inch of silicon, absolutely it's going up, because of multi patterning. But Intel for a variety of reasons, we're able to increase our scaling, so we get a better than average reduction in the transistor side.

And when you net that out, we stay on that historical cost per transistor curve. What we've seen in publicly released data and what we're hearing from some of the customers of the foundries is that they are off of their historical curve. What that means out in time again is that we should be able to do products that are higher performance, lower power and lower cost than what others can do, because we can keep driving that cost per transistor to historical trend.

Glen Yeung - Citi

Part of the reason -- I guess, I'm assuming here, part of the reason that I think your competitors are off the traditional curve, it's because they are playing a little bit of catch up, not on the shrink which they can sort of time with you, but on the sort of ancillary technologies that you'll implement in this particular FinFET. One, do you think that's a fair characterization; and then two, what is Intel's cadence with respect to changes in material science?

Stacy Smith

Another profound question.

Glen Yeung - Citi

Thank you.

Stacy Smith

I think there is two things going on; one is, I do think this is partially due to us being an integrated design and manufacturing company, where for others that manufacturing is separated from the design. And so keep in mind, we all go by the same equipment set, we all buy from ASML or Nikon's, so we buy the same lithography equipment. We all go buy from a Applied Materials or LAM, so we're buying the same route equipment set.

I think one of the things that has allowed Intel to extend our lead over the rest of the industry is, we can very early engage the process manufacturing guys, the manufacturing guys and the design teams together to go through and say, okay, with this equipment set how do we create an optimized set of trade-offs such that we can take that equipment set and we can actually improve the scaling of a product or create a paths that gives us that historical cost per transistor curve. And so I think we can engage in those problems a couple of years before our competitors have, and that runway is quite beneficial to us.

And a second thing that I think has separated us out over the last several generations is more on the material science side. And you mentioned high-K metal or you mention FinFET, that you can go back with sort of thing with high-K metal gate, you go before that,the same thing was strained silicon. It's a place where we've had a lot of focus -- we spend a lot of our R&D path finding, thinking through the material science side of this. And I think it has differentiated us out in time and it seems to be one of the places where we intercept the material science at the right time, and we do it in a way, where we can quickly get into a high volume manufacturing.

Glen Yeung - Citi

But you tend to begin the initial work on them. I mean if I think about FinFET, it was at least 10 years prior to the introduction.

Stacy Smith

Yes. 10 years is about right. So high-K metal gate was probably the same, where we were looking at high-K metal gate on the horizon 10 years before we implemented it. Somewhere in that five year horizon we've whittled it down to two or there choices. At the three-year horizon, we're locked and loaded on one, and then we tell you about it kind of six months before it comes to market.

Glen Yeung - Citi

So let's assume we're standing kind of a regular cadence, one might anticipate a material science, something new there in 2015, 2016 something like that, right, would sort of fit the four year cadence that you're on. So I could ostensibly go back to 2005 or 2006 and look at what you might have been working on at that time to give us a clue, things like, I don't know, nano -- carbon nanotubes or indium phosphite or something like that.

Stacy Smith

Yes. If we had told you in 2006, the things we were working on you could probably anticipate.

Glen Yeung - Citi

That's the big secret.

Stacy Smith

Yes, this is one of the things where we almost never talk publicly about the R&D horizon out that far for exactly this reason.

Glen Yeung - Citi

I wasn't trying to ask.

Stacy Smith

And in many cases they don't even tell me.

Glen Yeung - Citi

Right, and there is multiple things to it.

Stacy Smith

And there is multiple things and that's smart enough to guess which one is going to be the one that's manufacturing really.

Glen Yeung - Citi

So we make the case that for your own processor baseband connectivity, you will be leading edge technology, what about the idea that you can provide that technology to someone else via foundry. How do we think about foundry as a business for Intel, is this meant to be a core business or just a fab filler?

Stacy Smith

Yes. I mean first is to debunk the myth. We don't need foundry to fill fabs. We have the responsiveness that we can tune our capacity, our demand and so it's not like we're sitting there with excess capacity, and we say, let's go pull in a foundry partner to play. And we can typically balance capacity within about six months and keep them out from engaging where the foundry customer to actually manufacturing that product. That's a two-year cycle.

So it's a myth that it's a fab filler. We do it, simply put, because we make money at it. It's a good business and it's very consistent with our strategy. If you think about that strategy of, use the combination of our IP and our manufacturing might pull in as much of the world's volume into our factories as possible over the next five years. Foundries are nice adjunct to that. And I think the recent deal that we did with Altera, it's a great example. They came to us because they believe that landing their products on our process technology give them a significant competitive advantage out in the marketplace. So they can just have better products related to what other people can do. They're going to have great products. They're going to pay us a nice margin for the manufacturing. So it's good for them and it's good for us.

Glen Yeung - Citi

How do we think about where foundries fits on the node curve in the sense of, if I think about what Altera is intending to do, like start a process today, but it set revenues for two years, it's end technology. But by the time they get to revenues, it's actually N minus one because your goal to keep the foundry cost, to partner typically N minus one or is it just at particular instance here with Altera?

Stacy Smith

No. For somebody like an Altera where there is not any competitive considerations, we'll let them be as close to in as make sense for them and us. Now keep in mind being right at that leading edge nose, being right at the nose of a leading edge process technology is a pretty chaotic place to be. There is a lot of risk associated with it, that's one of the reasons that we've kept our Atom roadmap a little bit off of the nose, because it just adds variability on something that is already a pretty tough engineering challenge to pull off. So most people don't want to do right at the nose, but they can be as close behind that as they want to be.

Glen Yeung - Citi

Now if I look back at the history of Altera and Taiwan Semi, for example, they would often work together with Taiwan Semi to be at the nose when the nose is ready.

Stacy Smith

Yeah.

Glen Yeung - Citi

But what's different obviously for you is you've got a processor business that can do that. Do you want to work with foundry partners to help move things along or do not need them to do that.

Stacy Smith

We don't need it. But to the extent, it's valuable to them we're certainly willing to do it. But no, we don't -- we have our own products that we can use to kind of have our cadence of process technology roadmap. So we don't need to work with the foundry partner to do it. But if they get benefits of being close to that leading edge, we're happy to work with them to get them close.

Glen Yeung - Citi

Let's sort of think next about capital spending. And you make the point that we're at pretty high levels today. But we also have technologies in front of us, which aren't going to be cheap. And what's different then kind of 15 years ago is fewer companies to spread that technology cost over, but how do we think about, therefore, CapEx over the next, call it, three to five years?

Stacy Smith

CapEx, the historical trend of CapEx scales with volume and features, I think holds for us. And it comes back to the math that I was talking about earlier. Capital cost per wafer, there is no questions going on, then it's going out for the simple reason of multi patterning, now it's happening on lithography, so capital cost per square inch of silicon goes up.

We're able to offset that by getting greater scaling in our technology and so we're staying at the historical cost per transistor curve. And what that means, and this is really important is our capital cost per transistor is not going up from the historical curve. Again, we think that's different from what others are seeing. So we do think they are seeing real increases in capital intensity. We're seeing increases in capital intensity per wafer, but not per transistor.

And so now, it's a question for us of our capital spending over the time will be driven by how much capacity we're putting in place for volume growth. It will be driven by one of the features that we're putting on a dice. So we're adding more features and more graphics capability and capital intensity goes up, you saw that with the Sandy Bridge generation.

For example, if we're driving a product line where we're trying to take that transistor benefit in cost, then the capital intensity will go down. And the good news for us is that those kinds of drivers of capital -- when we invest in the capital, we tend to get a really good return on that investment.

Now, the other thing that will drive CapEx over the next several years is going to be the 450 millimeter transition, as we go to the larger wafer size that absolutely will drive an increase in capital intensity for a period of time, but that's also a very high net per present value to us to make that investment, because once you get there, you get a much better answer as you have the larger wafer size and you get to spread the capital cost across a lot more day.

Glen Yeung - Citi

And that was proven at 300 millimeter when you did the same.

Stacy Smith

Every generation, 200, 300 and we test the parameters at 450 and they still sold.

Glen Yeung - Citi

Given Brian's prior role at Intel, do we think about manufacturing strength, the capacity as more of a competitive advantage or is it really just the same thing, it's always been there?

Stacy Smith

I don't think it's a change and we don't really think about the work from the standpoint of putting the capacity in as a competitive advantage. We think about the world of -- we have this manufacturing leadership and how do we best manifest that. But we really are investing in capital consistent with how we see unit growth, how we see the features that we need and how we see the technologies of all.

Glen Yeung - Citi

And then just sort of one last thought on CapEx, which is TSMC is now openly talking about the idea that in the future nodes, cost are going to go up, right, and they have to pass this on to their customers. Is your capacity as a -- do you see just to think about, we really got to go aggressive now, because we can really push the lead here and make life a lot easier in year or two or three years from now.

Stacy Smith

I think we may get to the same answer, but we don't approach it from a standpoint of putting the capacity because of that. I think that the way this will play out, in particular because of I think now the responsiveness of our planning system is, if and as they fall off of the historical cost per transistor curve, we should have products in the marketplace that have benefit related to the competition or we're engaging with foundry partners, because they're benefiting from our process technology. That would then us to invest more in CapEx. But we weren't investing in CapEx on the anticipation of them falling off the curve we will do it because we're seeing progress in segments of the market where we want to put in the CapEx to satisfy that volume.

Glen Yeung - Citi

One other quick question on foundry that comes to mind which is, you have the potential to be a foundry for someone who could be a competitor to you, for example in handset, is that a problem?

Stacy Smith

Not for us.

Glen Yeung - Citi

You'll be willing to do that?

Stacy Smith

Well, this is where I threw in on your earlier question, somebody like Altera that doesn't compete with us, we would put them as close to that leading edge noses as they want to be in as they are willing to pay for it. No, you get into much more interesting strategic considerations. You can take the extreme example of QUALCOMM. If you give QUALCOMM leading edge process technology at a foundry margin, the answer is probably not, because it would make their product so much better. And so now you're taking a foundry margin on one side, but you're creating a much more significant competitor on the other. So those become much more nuance kinds of considerations and conversation.

Glen Yeung - Citi

Let me at this point open it up to the audience. We could have a couple of minutes left. Does anybody have any questions out in the field? I see one over here.

Unidentified Analyst

I have two questions. One is, now that's obviously a meaningful inflection point in the universe. You cost per wafer is going up in the first time in history and you guys -- you will have stay on that curve of keeping cost per transistors lower. How much longer do you think you can do that? And how -- at what point does -- did laws of physics as well as the laws of economics become an impediment? And then I have one quick one after.

Stacy Smith

So actually just -- the cost per wafer has been going up forever. I showed these curves, if you go back to the Analyst Day from last year. But in rough numbers, you can think of that the capital cost per wafer kind of goes up at 10% to 15% per year. What's happening now is that curve is accelerating. So it's going up at a faster pace than historical. And then I guess offset by efficiencies and it gets offset by the fact that we get the scaling benefit. And so even though, the dollars per wafer go up, the cost per transistor comes down because there's other benefit to offset it.

What we're seeing right now is a kink in both of those curves, the cost per wafer for us is going up at a faster pace because of multi-patterning. We're finding opportunities to improve our scaling at a faster than historical rate and that keeps us on that historical cost per transistor curve. We showed a graph that showed a solid line to 14-nanometer where we felt really confident. And then the dotted line to 10-nanometer, where we're still solving some of the problems, but we're signaling that we were pretty confident, we can keep on that curve at least through 10-nanometer. We haven't shown further out in that.

In terms of how long does Moore's Law continue, and Gordon Moore back when he postulated it, he was the first to say, no logarithmic curve continues forever. So there will be some into it. I'm 25 years at Intel. There is probably 24 of this 25 years. There has been some hypothesis that we're coming to the end of Moore's Law. We tend to have to the prior conversation about R&D, we tend to have a horizon looks how 10 years and some idea of how we solve problems of physics. I think if you talk to our technologists, he'd say, they can still see out that 10-year horizon and then beyond that it gets murky, which is kind of the same as it's historically been.

I will put one asterisk by that, though, which is we know there is one transition that the world has to go through sometime in the next three, four or five, six years, which is the transition to EUV, that's what gets us out of this multi-patterning situation that we're in. That's a really important transition. And it's not so much about the physics of Moore's Law, it the economics of Moore's Law. You do reach a point that as you go to double-patterning and then quadruple-patterning and then potentially more than that, that is just is no longer economical to push the canes at the same pace. So that's why that EUV transition becomes so important and that's why you've seen our partnership with ASML really jump to the top of our strategic list is to help get that done.

Unidentified Analyst

And then second question is, your imagine that new inflection with actually weak demand in your core segments and so do you think that's a structural shift in PC demand that you're dealing with or is this more of an economic cyclical shift that we expect to bounce back and go back to it's historic demand curves? And then if you having any thoughts on a 100-gig InfiniBand, I'd love to hear.

Stacy Smith

I'm sorry, what was the last part?

Unidentified Analyst

100-gig InfiniBand.

Stacy Smith

The 100-gig InfiniBand, yes, so know on the last question, to the trends in the marketplace, it's interesting. If you look at the market for compute devices, so you kind of take tablets and notebooks and two-in-ones and netbooks and all of those devices, but what you've seen is robust growth. And it's not terribly different from the historical curve. The issue is that there has been big mix shift within that. And so I think the good news for us is that the market for devices that compute and connect to internet is huge and it's growing fast.

The bad news for us over the last few years is we had products that were fantastic products at the high end of the market, but that segment of the market was declining. We didn't have products that were in tablets in some of the more consumer-oriented price points. We're there today, and so when I look at that overall market, I would expect that we're going to see unit growth over the next several years, because I think the market for things like compute and connect is growing fast.

Glen Yeung - Citi

Unfortunately, we have actually used up our allotted time. So Stacy thanks for being here. We really appreciate.

Stacy Smith

Thanks for having me. I Appreciate it.

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!

Source: Intel's Management Presents at Citi 2013 Global Technology Conference (Transcript)
This Transcript
All Transcripts