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

Altera Corporation (NASDAQ:ALTR)

Raymond James Systems, Semiconductors, Software & Supply Chain Conference

December 10, 2013 13:20 ET

Executives

Scott Wylie - Vice President Investor Relations

Analysts

Hans Mosesmann - Raymond James

Hans Mosesmann - Raymond James

Go ahead and get started. Well, hopefully everybody has enjoyed the lunch. We have an Altera presentation that will be done by Scott Wylie, VP, Investor Relations. He'll go through a fresh slide deck that I think is relatively new and for about 5 or 10 minutes and then we will go into fireside chat. And Scott, thank you very much.

Scott Wylie

Great. Thank you, Hans. And thank you all for coming. Let me first of all cover this obligatory ground with you. This presentation contains forward-looking statements, which are made pursuant to the Safe Harbor provisions of the Private Securities Litigation Reform Act of 1995. Forward-looking information generally refers to any information relevant to a future time period. Investors are cautioned that actual results may differ materially from these forward-looking statements and that these statements must be considered in conjunction with the cautionary warnings that appear on our SEC filings. They're available from the company without charge. Further, this discussion today may repeat elements of our guidance and in so doing, the company has neither reaffirming nor modifying their prior guidance.

So with that said, I want to spend a couple of minutes with you at the outset talking about technology. And I have multiple reasons for wanting to do so. One, and this is perhaps stated in obvious, the technology side of Altera is really that which creates growth and creates success for us. It’s the way in which we win businesses, it’s the way in which we created value for our customers and of its core obviously what the customer does. And a healthy record of innovation in that regard and some of that which I want to touch on at the moment.

Secondly, we're at a point in time, where we're sort of on the cusp, if you will, of rolling out the next generation of products. The pace and cadence for us has been every two years, give or take. We're at one of those moments to where things are starting to get to find things or starting to happen, things are starting to arrive and you could being to get an impression of what we're up to. So this is a very ripe time to talk about technology.

And finally, the lay of the land in terms of what we're doing in the next generation is quite a bit different than the way in which we prosecuted the prior generation. Generically, yes, you have recognized some very clear sign, but the nature of the innovation that is being implemented is substantial. And so it does have a different feel to it in terms of what at a similar point of time we might have been talking about with the prior generations some two plus years ago. So that’s a sort of my motivation, if you will, to introduce this topic to you a little more deeply.

Now, just by way of reference, I find this slide is useful because what it does is it takes our product offerings and apportions them into the high-end, the mid-range and the low cost end. And as you can see in green, the high-end has traditional been, what we expect this will continue to be the case, will be the largest revenue generator for us as we go forward. So, its obviously a very important product range for us.

The mid-range now that we and our main competitor are both prosecuting mid-range products is arguably about a quarter of the revenue pie, the low cost the remainder. I've also indicated on the slide, the general relationships between any one of these three categories and the end markets that we serve.

Now this isn’t a one for one, meaning in this case, its not that the Stratix series sells exclusively to computer and storage, its kind of a central Tennessee, if you will. That tends to be the biggest place of demand if you will for that category parts and so on. So we do sell all of our parts across all of our verticals, but there is a sort of a natural relationship between the high, mid and low cost area in a variety of end markets.

Stratix series is obviously is the performance leader, its going into the most demanding from a performance standpoint applications. And that’s not to say that there isn’t interesting things to talk about in the rest of the business or indeed is this good healthy growth drivers. But there is a natural kind of affinity if you will between the high, mid and low cost and in a series of end markets. So keep this in mind as we talk going forward.

If you look at what we are about in terms of this next generation, we're refreshing product across all three of those categories. MAX 10 which is focused with the low cost end, I will not spend much time on today. It will be manufactured on TSMC's 55-nanometer Flash technology. It is a non-volatile memory product, CPLD like in that regard if you're familiar with a programmable logic world. And there is an ongoing demand for FPGAs that have that non-volatile memory inside them.

This is after some years a very notable refresh in terms of this product category as we go forward. There will be more information in terms of the details on this product family as we get into next year. It will be available to our customers next year. And it’s an important piece of our overall product portfolio.

But turning to the middle, Arria 10, which is the mid-range product will use TSMC's 20-nanometer planar technology and it really is geared as you might expect by being the mid-range going for the suite spot in the mid-range market. And we'll see in the following slide here there's really quite a evolution with respect to what Arria is going to be able to deliver to customers.

And then finishing out the portfolio, if you will, the high-end, this is the Stratix 10 part will be sourced from Intel, which is the first time for us in many, many years that we've had Intel as a supply partner. We will be implementing their 14-nanometer Tri-Gate FinFET technology. We have unique access to that technology and that will pass on some related benefits to us in terms of what their high-end part can do.

Now, let me dive into the mid-range here a little bit. And I will confess to you, out of the public’s interest in the high-end and FinFET and Intel and what not, I some times think that the mid-range doesn’t get us a fair sake. There is a lot going on in the mid-range and it’s an important segment for us in terms of what we offer overall.

And if you look at just very simply in terms of what next generation mid-range can do for you? First of all, performance, speed, a significant step-up in performance but here we're measuring against our previous high-end part. So its not previous mid-range, it’s a previous high-end. And for that matter, it exceeds the performance of our competitors' high-end for whatever that’s worth as well. But, this is the mid-range, it is going into territory where previously you will be forced to have dealt with the high-end part. So a significant step-up in performance generation on generation. At the same time, have done some work with the product itself to where you can accomplish it to work at notably lower power consumption. Here, we're comparing against the previous Arria family, so this is the hardest compare, if you will.

And as a consequence of the work that we have done in this mid-range, we think we got ourselves a very competitive, very appealing to end customer product at the next generation.

Now, where do we stand in terms of timing? This product has been available on software for a while. It will tape out, pardon me -- it has tapped out already, it will be available in terms of first silicon, first quarter of next year. So if you put these pieces together what’s notable is, at this moment in time, we're the only FPGA supplier with broadly available software. And we're close to introducing a product. This is a much more competitive position than we were in at a similar point in time in terms of the prior node.

So it’s quite a bit different in terms of performance, in terms of entry into the market and so on. And you will hear more about this obviously as the year progresses next year but customers are now designing with the product. That’s all they need a software, that’s the important point of maturation in terms of the stability of the process, it’s publicly available software.

Now, change to the high-end, you see something is pretty remarkable here. Here we're looking at performance. This is speed. And this is over time and what you get and this maybe a little difficult and that’s okay. You get the prior generations and you can how the prior generation steps up in terms of performance generation on generation about maybe a 20% bump, right.

What you are experiencing here is the inherent limitations of the planar transistor technology. It’s getting harder and harder to be able to get higher levels of performance at acceptable levels of power. And so you end up compromising to be able to have acceptable power profile. As a consequence the performance bumps happened historically about 20% of generation.

The next generation Stratix 10 part, will make a significant bump over the prior generations as you can see effectively it’s doubling the speed of the part. And that’s a combination of several factors simultaneously. One, it is a derivative of what you can do given that you got a 14-nanometer FinFET process. Secondly, it is a clear result of some significant architectural innovations that give the part that kind of characteristics that I have just described.

And its also fair to say the part of the reason why you see the significant jump in terms of performance is we're going from 28, which is the last prior node to 14. It’s a two node jump, okay. So if you're talking about two nodes, if you can look at fairly significant bumps in performance but truly it’s a combination of all three of those things are creating that what you see.

Now, we're the only major FPGA company with access to the Intel Tri-Gate technology a 14-nanometer. So this gives us the open door to some rather unique performance which is obviously advantageous competitively, it is also equally clear in terms of market expansion potential because often times we cannot compete for parts because we aren’t fast enough. This will make a major inroad and I have to mention as well.

There is more things going on under the hood, a real step up measured in terms of performance per watt. There is a new offering at the high-end that is it will happen in embedded processor. In this case, it’s a pretty muscular Quad-Core 64-bit ARM A53 processor and talks with our customers as we're making up our mind in terms of what embedded processor are offered in the high-end customer reaction to this particular processor family was very, very supportive. So we think this is going to be a clear win for us. So this now means, that you can get embedded processor capability all up and down our FPGA product offering from high, mid, to the low cost Cyclone family.

Step up in terms of DSP capabilities, obviously, there is more and more intensive DSP like work going on with our customers. And this will be the highest density chip and it will be monolithically delivered meaning it will be able to top any other FPGA offering because of the advantages you have at being at the 14 FinFET level. You can put more transistors in that chip therefore create logic density.

And that advantage will accrue to our benefit as this part begins to roll out. One would think, therefore, that the early stage prototyping and emulation market swings back into the Altera column as a consequence. And we do have available to us the 3D packaging capabilities of Intel as well. So it’s a very interesting combination of attributes.

Where are we in terms of timing? The first test chip for this family that’s an R&D chip is in-fab right now with Intel. The development process has gone smoothly. We're right exactly where we wanted to be. We go typically through a series of test chips while we're on the path to ultimately commercializing the technology. We will get in due course the chip back in-house and will allow us to take the next stage on the R&D process itself.

Secondly, the software to design this family, the Stratix 10 family will be available in the first half of next year. Now, that’s important. This may seem a little hard because here we're a semiconductor company. The reality is for our customers, once they have access to the software, they have the ability to start to put their designs into what will ultimately be our product. They don’t need the physical product. It’s a public availability of software which is really the milestone event.

And it’s a milestone event in a couple of regards, one obviously, it means customers can start to work. But two, it also speaks where are you in terms of the development of the processor. It implies a fair amount of maturity in terms of where the silicon sits and obviously in terms of the agreed development on the software side as well. So you can’t – once you make software publicly available, you can’t drive within a totally different direction without upsetting the customer base.

Later in 2014, fourth quarter 2014, we will tape out our first 14-nanometer FinFET hard chip. It, therefore, will be available early in 2015. That’s the timeline that we're looking at. And the change from planar technology, the FinFET technology obviously is a very significant shift. And I thought it might be worth just one last look at some of the dimensions of that shift.

As you maybe aware, Intel today is the only manufacturer of FinFET-based semiconductors. They did an unusual thing. At 22-nanometer, the prior node from the one we're going to intersect, they chose to do the typical die shrink that is down a size in terms of process geometry, but they also implemented their FinFET technology at the same time. And they are today in volume for sale.

They will begin to manufacture 14-nanometer FinFET chips here in the next few months. We obviously tape out at the end of the year. We have an advantage actually in this regard because we will intersect a rather sorted out process. We're not their lead customer. The process will be well bolted down and we will get the advantage of that. So actually it is in some respects a loss risky way to transition into the FinFET world. We're going with the people who do it now; they're the only people who can. We're going to intersect a process which is going to be pretty well shaken down at the time, the time we grab it.

Now if you look at the implications of the change, this chart might be some interesting. What we're doing here is we're comparing Altera against our competitor, we're using 20-nanometer as our point of departure. Now remember, in our case, using the Intel processor, we will drop down to a true 14-nanometer Tri-Gate processor and we will get the kinds of performance that I outlined for you, lower power and dramatically lower cost all as a result of the implementation of the FinFET technology.

Our competitor, who does not have access for this technology, widely understood that their source which is TSMC and this is similar across other potential suppliers of FinFET technology. They’ve taken an approach, what we’d like to do is take our existing technology adopt it to FinFET and go forward on that basis. Don't shrink it. Take what they were doing and adopt it.

So what you see here is a 20-nanometer planar chip identical in size conceptually to a same sort of chip that we will be operating, but as they go forward, there is no scaling. This has been publicly acknowledged to be the case. And so what they would get is, they would get some step-up in performance, they would get more favorable power profile, but because the wafer that's manufacturing the chip is more expensive and there is no scaling, the cost is going to go up.

So for customers who are willing to pay up for that that’s fine. But the relationship on our side is quite a bit different because you will be able to get lower cost logic at higher performance and with lower power. This different we think is going to really manifest itself in terms of the competitive match up that we’re likely to encounter all while this next generation is rolling out. And you can now see part of the logic as to why we were so enamored of using Intel as a supply partner.

So if you take a look at it overall, one, from a product enhancement standpoint, significant movements in the mid-range and the high-end and obviously, a refresh in the MAX family as well. Two, a very timely introduction in the mid-range in terms of entering the market and three, their shift to FinFET being one that looks like we're going to be the beneficiaries of a supplier whose structurally a node ahead of anyone else in the world.

With that, let me stop here. Hans, I suspect you probably got some questions for me.

Hans Mosesmann

Yes, so great run, Scott, thank you very much.

Scott Wylie

Thank you, Hans.

Hans Mosesmann

Just for everybody to catch up, you guys have 40%, 45% exposure to the communications market?

Scott Wylie

Correct.

Hans Mosesmann

Clients want to know what’s happening with China Mobile, the base station and questions you’re going to get --

Scott Wylie

Sure.

Hans Mosesmann

Often, just a quick update there and we will go on from there.

Scott Wylie

Yes. Let me just review the recent history in that regard. If you look out our June quarter results, there was the latest phase of China 3G deployment. Believe it or not, the TD-SCDMA standard was being deployed at the China Mobile. We were the beneficiaries of that. That business did not go forward into the September quarter, that was no surprise and obviously the 3G standard is not the new about ready to be deployed standard.

What then occurred is in the September quarter, we saw the arrival of some business that was fuelling China Mobile's LTE deployment but two netted against each other. We ended up with positive compares sequentially in terms of wireless revenue and what we see going ahead is we’re going to maintain about that level of overall business going into the fourth quarter. It is devilishly difficult for us to predict what the profile looks like in quarters further out. As a consequence, I can’t offer you, here is how to think about Q1, Q2 and Q3 for instance.

Other than this generalized observation, China is a large country where in this case, China Mobile is the world’s largest wireless carrier. They are early, early, early days in terms of their overall LTE deployment. They would be joined not too far distant period in time by China Telecom and China Unicom as well. The 3G deployment, if that’s any measure, is probably a healthy five plus years from beginning to end, and so we’ve got a long, long runway ahead of us.

And being able to benchmark, exactly what the deployment pattern is going to look like is very, very hard. And it was true in the 3G era as well. I think that what’s underappreciated is the complexity of what the wireless carriers are engaged in in terms of these deployments. There is a lot of variety. There is a lot of engineering work to get accomplish, it influences the supply chain in terms of nature of the equipment that they need to deploy and it is variable. But nonetheless, it’s substantial and it would go forward. Fortunately, FPGAs play role in those systems have historically would continue to. So, it’s net of a favorable event for the FPGA industry. Everyone quite reasonably would love to know exactly what’s going to happen tomorrow. Tomorrow meaning in the next year --

Hans Mosesmann

Right.

Scott Wylie

Hard to figure.

Hans Mosesmann

The competitive dynamic in China specifically Altera versus Xilinx is going back and forth. You guys are comfortable that you have market share advantages or continued?

Scott Wylie

Yes, we’ve done well in the wireless space over time that would be a global comment. I would say, that would be the case in China as well. Remember that what’s deployed are base stations on the one hand, the remote radio heads on the other and the backhaul equipment on the third, it’s all part of a package, if you will. If you look at the balance of totality, it looks like our position going forward in China continues to be quite good.

We have for sure a viable competitor in the mix. They are doing everything they can to gain as much share as they possibly can. They, in terms of their 28-nanometer offerings, have been able to see some success in that mid range, but nonetheless when you put it all together, I think our position in terms of China wireless will turn out to be fine.

Hans Mosesmann

Any question from the audience?

Question-and-Answer Session

Unidentified Analyst

(Question Inaudible)

Scott Wylie

Yes, great question. It would be different.

Hans Mosesmann

Can you repeat the question?

Scott Wylie

Yes, the question is, if you reflect back on the 3G deployment against the 2G predecessor, and what it was doing, how does LTE versus 3G look like just sort of move the focus ahead one generation? To revisit 3G a bit, here is one piece that would probably be different. As 3G was deploying, there were still actually very healthy 2G deployment, growing, maybe growing modestly but it was by far the bulk of the spend and 3G came along and effectively it was able to add to it. Again, ultimately 2G went into decline but it was modest, it wasn’t like a fell off from cliff and just stopped.

The Chinese for instance, we’re engaged in a period of 3G deployment, they were doing healthy 2G deployments at the same time, a kind of a bifurcated strategy in terms of their market it make sense. In this case, what’s going to be the fate of 3G in comparison, obviously, where the front-end of the whole process, so, I’m not going to make up for sure prediction in terms of how its going to phase in. There are parts of the world that are still waiting for 3G in much the same way that people are waiting for 2G. But nonetheless, to think that 3G is going to be a growth driver for equipment, I think that day has gone.

Could there be some short-term noise quarter-on-quarter as a -- that is a blast of 3G some place and not a moment of LTE some place else, I mean it’s conceivable that would be the case. But, over time 3G is not destined to be the growth contributor, but in absolute dollars, it’s still quite substantial. So this thing moves ever so slowly, I’m thinking back and just don’t take this as science, trying to remember back to one to 3G equipment spend eclipsed 2G equipment spend, it was very late maybe 2006, 2007, something like that, I mean years after the onset of the 3G deployment. So I expect that kind of gradualism is going to be the case. One number is fixed in my head.

If you looked at subscribers today, granted we're not selling in the handsets and what not, worldwide subscribers who got LCD capability is a huge 2%, right I mean so we're right at the very beginning and subscriber growth is obviously one of the things that forces the carriers to spend as their initial deployment days end because they need to continue to upscale their networks. So intentionally not a direct answer to your question but I wanted to kind of get the parameters in place to frame it for you as best I can.

Unidentified Analyst

(Question Inaudible)

Scott Wylie

It’s in some sense of course remember the 28-window is still open for design competition of course. I think it’s – it is absorbable for instance, Huawei continues as our largest customer. We have been the largest FPGA supplier to them for years; I think that’s entirely likely that, that’s going to continue. Against the other people that are supplying the Chinese market Ericsson is worth noting granted it will be smaller in terms of contribution to the Chinese wireless build. But, we're right on the cost of becoming the largest FPGA supplier there. So I mean, I think when you put the pieces together is this still a very constructive environment for us. I don’t think there is any question about that.

Unidentified Analyst

(Question Inaudible)

Scott Wylie

You are saying some form of fixed-function chip alternative.

Unidentified Analyst

(Question Inaudible)

Scott Wylie

Sure. They are in a class of customer that has the capability of engaging in fixed-function chip design who will use off-the-shelf standard products and will use FPGA so all three technologies. It’s just a question of the balance between them and it will never go 100% one camp or the other. The essential ingredient for people who want to increase their fixed-function chip usage is they need to find an environment to where they can get to sufficient volume to justify the upfront development cost. And as transistor sizes gets smaller of course as upfront development cost climb and they climb spectacularly. So the economies just get more and more challenging. And so in this sense the – any of the larger customers and I would include Huawei in this regard have the ability to move product on the margin out of one zone into the other and will do so.

I think it will be governed by economics and what we’ve seen is economics has been the friend of the FPGA industry I think likely to continue. But the principal is ever present. At the very beginning when they are trying to figure out whether they want to use a flexible FPGA or they want to use a fixed-function chip it’s a question of whether they have enough lifetime volume to justify the upfront development cost. At any subsequent point if they when did the math they’ve got the same hurdle in front of them. And what we’ve seen is the passage over time is because that hurdle has risen it becomes harder and harder for them to go down that road. I won’t say never because they will find some cases to where they can get the volume. But I think it’s going to be governed by the same sort of economics that govern any other competitors.

Unidentified Analyst

Sure.

Hans Mosesmann

Okay, Scott. I noticed that with the Arria 10 slide that you are really pushing the notion that in the middle of the market you have a product that it potentially disruptive, Xilinx certainly had a great product of 28-nanometer in the middle of the market. So how does that competitive dynamic play itself out over the next year or so in terms of design win activity and then what could be your advantage and perhaps having software that’s available earlier or is that earlier?

Scott Wylie

The software is, we're as far as I know today, the only company with software publicly available. Obviously our competitor will be there nipping at our heels eventually. But the mid range if you look at the breadth of what FPGAs can cover from the low cost and to the high end it is an extraordinarily large territory. So segmenting the products into three is not unreasonable given the changes that you see in density from the bottom to the top. And the mid-range has really come of age with the 28-nanometer node, I mean its 25% of revenues of the industry, we and our competitor believe. So it’s time has arrived.

And you can see with the kind of adjustment that you can make to its deliverables that it has what can I say ongoing and very healthy future because it can stretch in the categories that previously could not have caught. At the same time, the high-end goes through its own evolutionary process and steps up its own performance which means it opens up markets previously unavailable to us. And the net of the two means that we get growth. And yes there is this ongoing evolution to where the mid range becomes more and more robust but it’s coincident with the high end equally becoming more robust so the whole area expands if you will as a consequence.

Hans Mosesmann

Okay. (Ted) you had a question, we have 30 seconds.

Unidentified Analyst

(Question Inaudible)

Scott Wylie

Correct.

Unidentified Analyst

(Question Inaudible)

Scott Wylie

Yes, short answer to your question not at all. And Bill (inaudible) whole through is the presenter in this regard, was very clear about what it was that created the delay and it was the elapsed time that it taken to get to what they believe is satisfactory yields to allow them to do their full scale development. We tape out fourth quarter of 2014, we start to see chips first quarter of 2015, I’d call it normally a year behind, right. That’s more than enough time for the process to be thoroughly shaken down and in fact the information that they presented showed significant progress already, right in terms of their ability to do it.

So that obviously, they are making the product now for sale in the first quarter so they are totally confident in terms of ability to do it. But now, it does intersect at us at all. And just to underline this point quite interesting for us it means we will intersect a 14-nanometer process which has got time and grade. It’s already been out, it’s already been sold, it’s already understood by the manufacturer. That’s very different than being the lead customer for somebody because the lead customer is inevitably a little bit of a process of mutual discovery, right. So this appears to us actually to be a much more copasetic path to go convert to that FinFET technology from where we sit.

Unidentified Analyst

(Question Inaudible)

Scott Wylie

Yes, the agreement is public record by the way. It’s attached to an SEC filing earlier this year and if you are curious go, read it. The agreement that Intel entered into with regard to 2014 is that they would manufacture chips for us for 12 years and will provide the last time buy at the end. There is a consequence, if that turns out not to be the case meaning there is a financial consequence. And the combination of those features says to us we have every reason to believe that Intel is going to be able to fully participate in the supply of a product to us.

Hans Mosesmann

Okay. And with that Scott, thank you very much.

Scott Wylie

Good. Thank you, Han.

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: Altera's Management Presents At Raymond James Systems, Semiconductors, Software & Supply Chain Conference (Transcript)
This Transcript
All Transcripts