Advanced Micro Devices, Inc. (AMD) Q3 2011 Earnings Report, Transcript and Summary

Good afternoon. My name is Kristen, and I will be your conference operator. I would like to welcome everyone to the Xilinx Third Quarter Fiscal Year 2011 Earnings Release Conference Call. [Operator Instructions] I would now like to turn the call over to Mr. Rick Muscha. Thank you. You may begin your conference.

Thank you, and good afternoon. With me are Moshe Gavrielov, CEO; and Jon Olson, CFO. We will provide a financial and business review of the December quarter, and then we'll open the call for questions. Let me remind everyone that during our conference call today, we may make projections or other forward-looking statements regarding future events or the future financial performance of the company. We wish to caution you that such statements are predictions based on information that is currently available and that actual results may differ materially. We refer you to the documents the company files with the SEC, including our 10-Ks, 10-Qs and 8-Ks. These documents contain and identify important factors that could cause the actual results to differ materially from those contained in our projections or forward-looking statements. This conference call is open to all and is being webcast live. It can be accessed from our Xilinx Investor Relations website. Let me now turn the call over to Jon Olson.

Thank you, Rick. During today's commentary, I will review our December quarter business results. I will conclude my remarks by providing guidance for the March quarter. December quarter sales decreased 8% sequentially to $567 million, in line with the revised guidance we published on December 21. Turns business for the quarter was 44%, down from 48% in the prior quarter. Gross margin of 65.7% was slightly higher than guided, primarily due to proactive cost reduction efforts at the company as well as more favorable product and customer mix. This is up from 64.1% in the same quarter of the prior year and represents the fifth consecutive quarter of improvement. Operating expenses of $189 million were inclusive of $4 million in restructuring charges and were slightly lower than guided due primarily to lower variable expense associated with lower sales. Operating income was $184 million or 32.4% of sales in the December quarter. This is up from 26.6% in the same quarter of the prior year. This operating income performance represents a year-over-year increase of 34%, over 3x the rate of sales increase during the same period. This speaks to the aggressive cost reduction efforts and spending discipline undertaken by the company over the past year. New product sales decreased 10% sequentially in the December quarter to 43% of sales, driven primarily by Wireless Communications sales declines for Virtex-5. Sales from mainstream and base products both declined sequentially. Sales from all geographies declined sequentially during the quarter. European sales were down 22% sequentially, driven primarily by declines from Wireless Communications. But Wired Communications, audio/video broadcast and test and measurement were also weaker than expected. North American sales declined 7% due to declines in most end markets. Asia-Pacific sales were down 1% sequentially as Wireless sales decreases were partially offset by increases in…

Thank you, Jon, and good afternoon to you all. Declining sales in the December quarter were primarily attributable to a large customer rescheduling in the Wireless Communications end market. While we did not achieve a fifth consecutive quarter of record sales, my conviction in the programmable imperative remains strong. Calendar 2010 marked a tremendous year of growth for Xilinx with sales increasing 36% over 2009. According to iSuppli, ASICs and ASSPs grew 7% and 25%, respectively, during this time period. This means that Xilinx grew more than triple the rate of ASICs and easily surpassed the ASSP growth rate as well. Increasingly, FPGAs are becoming the default solution in next-generation customer systems where the need for low cost and flexibility are making FPGAs a preferred alternative to ASICs and ASSPs. Combined sales of our 40-nanometer Virtex-6 and 45-nanometer Spartan-6 product easily exceeded $100 million in accumulated sales in the December quarter, one quarter earlier than we had anticipated. Sales from our 40-nanometer, 11.3-gig Virtex-6 HXT devices increased significantly in the quarter. The customer traction we're gaining from this product family, coupled with having the industry's only 45-nanometer high-volume FPGA family with embedded transceivers are significant drivers of the leadership we have in transceiver shipments. During the December quarter, we made two key announcements, which are integral to our 28-nanometer strategy. We announced the industry's first stacked silicon interconnect technology, which will allow Xilinx to deliver breakthrough capacity, bandwidth and power savings using multiple FPGA die in a single package. We expect this innovative platform approach to enable Xilinx to overcome the technical boundaries of Moore's Law and to offer customers unparalleled power, bandwidth and density optimization for the large-scale integration of their systems. We also announced our Virtex-7 HT FPGAs and demonstrated 28 gigabit per second serial transceiver performance required…

[Operator Instructions] And your first question is from Uche Orji with UBS.

So let me just ask you quickly. The $35 million that was lowered during the last quarter preannouncement, is that now fully incorporated in the March quarter guidance?

So yes, I mean, we were -- from the midpoint of our original guidance to the actual numbers, it was more like $40 million, right? But sure, I mean, our next quarter's guidance is reflective of what we view next quarter to be at inclusive of a recovery of our Wireless business.

Just sort of quickly, I mean, I just want to understand how you think the competitive dynamics has changed in the industry since you rolled out your 40-nanometer products. And also part of that, if you can also talk about the recent comment about pushouts or delays in India, whether that had any impact on you and whether we've started to see that come back through either now or when we think that can come back in 2012.

There's a series of questions there. So with regards to our competitive position, it is clear that we were later than our major competitor on 40-nanometer, and we also actually had been suffering from skipping a generation on the high-volume 65-nanometer. That's why we brought out both the 40 and 45, and at this point, the 45-nanometer has had unprecedented success. And so that should enable us over time to recoup the market share on the high-volume side. When you add that to the fact that we have the broadest product offering and we are committed to being out there as early as possible with that product offering at 28-nanometer, we expect to shift the market share back to where it has historically been. So combination of all of these elements, the 45 high volume, the 28 with its full breadth, and then that gets built upon with the ARM solution, which we are well on the way to implementation, we've already shared with the customers, that will enable Xilinx to regain market share.

And in your Wireless question, we don't really have a great picture by sub-segment right now of the next year given our fiscal year's the end of March. But I've seen the same things about India and the concerns over the absorption of the kind of price that it would take for the average India cell phone user to absorb 3G, and there's some speculation about all that. But quite frankly, most of our shorter term demand, meaning the next quarter or two, is coming from very strong LTE shipments and expansion of U.S. networks as well as the European networks and then followed by continued growth in China. And I think one of the things that also adds on to the first part of your question is very apparent also in the Wireless segment is the penetration of PLDs into that segment continues to be more and more. So even if CapEx is down year-on-year, we could very well have flat to up revenue in this segment because of the penetration. And this is just validation of our technology overall. But specifically about India, we're really not counting on a, I would say, very large number in the coming quarters for India. The big wireless uptick is really around the LTE generation starting to take foothold and the expansion of the AT&T environment. Also, over the last couple weeks, we've seen that AT&T has started to change the names of things to 4G, and they're going to pull in their real 4G deployment by a year. So all those things are very positive for us.

Your next question is from the line of Tim Luke with Barclays Capital.

You described how at the higher end of 40-nanometer you'd seen pressure, but you feel that you're seeing traction in high-volume 40-nanometer. And then you're suggesting that your 28-nanometer should help you. Do you think that your share will start to stabilize such that your revenue growth will be similar to your peers by the middle of your coming fiscal year? Or when should we think about you returning to sort of a more stabilized share position?

I think we'll talk a little bit more about that in March. But I believe from a high-volume product family perspective that the kind of traction that we're seeing with our Spartan-6 45-nanometer product is that by the end of the calendar year, we should see stabilization of share in the high-volume segment and then an increase, starting to see an increase in share. The high-performance end of things is a little fuzzier because we're bifurcating the product family to some extent. And as Moshe pointed out, we were late to the party on 40-nanometer, so it may take a little longer for that to move the other direction. But the way we're coming out, the speed at which we're coming out with technology of 28-nanometer, we're pretty confident that that's going to start around in '12.

Also, just, Jon, could you give us some feel for how you think the year, the calendar year or the fiscal year is going to play in terms of which segments you think may grow and which may grow more? And whether you think, now that your lead times have normalized, et cetera, that you're going to see fairly traditional seasonality in the June quarter and onward?

Well, I would expect that we'd start to return to overall more seasonal patterns just in general just because of that because we're at a level of business where still can be obviously still be volatility in certain end markets. But the economies are at least seemingly stable for now, and all that looks pretty good. We really aren't -- I'm not going to venture our forecast on which end markets are going to be better next fiscal year or not. We'll do that in March.

Your next question is from Ruben Roy with Pacific Crest Securities.

I'd like to follow up with you, Jon, on the commentary around the LTE. And if you look at kind of the near-term strengths that you're seeing in Con and Wireless related to European OEMs perhaps and then U.S. rollouts of LTE, would you say that, at least the initial designs and maybe through the end of the year, the design activity for LTE is behind us? And following on, when is the next real design win opportunity as it relates to perhaps next-generation LTE? And is that what you guys are talking about when you talk about 28-nanometer and potential market share gains at that node as you look out to maybe the middle of this year?

Yes, Ruben, yes. This is all kind of confusing to some extent if you start to equate LTE to 4G and 3G kinds of activities and then the providers changing their names of what their networks really are providing, et cetera. What's really happened here is then a consolidation of technologies on to FPGA so it handles all the protocols and waveforms in one line card. So while they may be selling into an HSDPA-plus kind of a network, it also can support LTE going forward. So most of the base stations that we sell into have some level of LTE capability so that providers don't have to replace everything all the time. That being said, at least from what I've read and certainly what we know, is most of the capabilities of the base stations are at the full range of what the LTE spec says. And what that means is that the bandwidth that they're providing under the name of LTE now will increase over time as they do improvements in terms of the base stations and et cetera. And those designs, I would say, would be kind of the next generation. Those designs need to be won in 2011, and those products will start rolling out in 2012 and 2013. And that's why we're pretty confident about our new 28-nanometer product family and having designed products specifically to win those designs.

And your next question is from Glen Yeung with Citi.

If we were to take out or that move from last quarter into this quarter, technically you're guiding for declines in the March quarter. And I looked back at your history. In 16 years, you've only had three March quarter where revenues declined. Can you give us a sense as to what you specifically think is going on in the first quarter? Is it your customers working down inventory? Is there something there that looks particularly weak?

Well, I think we did note that even in the December quarter, there was weakness more than we had anticipated. But one of the phenomenon -- and we specifically said that in the audio/video broadcast area was a little bit weaker than we thought, for example, and we think that's bouncing back. And some of that, specifically on AVB, was inventory correction at the customers we serve. So there has been a broader based inventory balancing going on, which I think started in the September quarter and continued on in the December quarter in markets other than communications. Now with respect to the Wireless business and what's going on there, China, while it was still significant amount of business for us, but the fourth phase of the TD-SCDMA is now wrapping up. So from a geographic perspective, in Q1, there's actually some declines going on in that part of the Wireless business, while the other parts that we serve primarily through the larger European customer are growing extremely rapidly. So there's a lot of gives and takes going on. And while the overall demand is still really, really strong for us in the Wireless, we had this giant lump, a negative lump that we had in the December quarter.

And your next question is from the line of James Schneider with Goldman Sachs.

On the inventory side, even if I strip out the increase that you attribute to the legacy products and the shutdown of the line there, I'll have to look back until almost 2001 to see an inventory absolute dollar number that was this high as it is today. Well, so I guess my point is, what's your level of confidence in terms of the strategy of maintaining that high an inventory level going forward even if excluding the legacy products? And how do you make sure that that doesn't stay or really get out of control?

Jimmy, that's a good question because we are taking a little more risk than we normally would have here. But I will also point out that we're operating at a much higher revenue level than we were back in that time period. So you do need the inventory in order to support that. But that being said, we've also gone through a couple of periods the last two to three years of not being able to deliver to our customers. And that's not lost on us in terms of a customer service level perspective. And then since we're taking this position that we feel that things still could be tight at foundries, we wanted to be cautious about how we adjust our capacity. So we've taken this position to add a little more inventory. But we are going to be very watchful of what's going on in the revenue line so we can manage this on a go-forward basis. We have no intention of getting ourselves out of line here in a big way. It is a kind of part by part we're managing this, and we have a monthly review meeting where we, in tremendous amount of detail, going through this to make decisions on which wafers we start out in time.

Your next question is from the line of Patrick Newton with Stifel, Nicolaus.

I wanted to dig a little bit into your product results. And if I heard you correctly, you said that Virtex-5 sales declined sequentially. I was wondering if you can discuss the magnitude of the decline, and perhaps your thoughts on the potential that Virtex-5 revenue has peaked as a family?

Yes. So we did talk about new products being down about 10% and Virtex-5 is the largest revenue contributor of the company. So therefore, it would be the largest percentage of the new products. So you could look at that 10% decline and use that as a surrogate for the number you ask. I mean, it's not precise, but it's close. And the additional question?

Well, has it peaked?

Oh, has it peaked? No, we don't believe it has peaked yet. Even though it went down this quarter because of the situation with the one Wireless customer, we don't believe it's peaked yet.

Your next question is from Ambrish Srivastava with BMO.

Just a question on the architectural differences. Timing seems to be -- so both you and Altera are on the same time for 28-nanometer. Just help us understand what are you guys targeting versus Altera, whether it's end market, and then what are the architectural differences that you have confidence and that will enable you to regain some market share on the 40-nanometer side?

Well, if you look at our product offering, we use -- there are a couple of key elements. First and foremost is we are targeting the HPL process at TSMC, which is a process that we co-developed with them and has since been adopted by several additional customers. Basically, that's a process that enables you to achieve very high performance at a lower power point. And as a result of having that, generally speaking, we can use a unified architecture to implement all of the three major product families. So it's the Virtex, Kintex and Artix. So it's the high end to mid-range and the low-end. That's unique. That will enable us to roll out products at a much faster rate than we ever have before. It will enable our customers to scale the designs, which is something they've asked for across the product families because they do need that very frequently. It will provide us with lower power points overall. And actually, having a lower power product will enable us to achieve at the high end a higher performance product because actually, it's the power capacity which is the big issue and we can achieve high capacity. So if you look at our product offering at the high end, it's actually significantly higher than everyone else, and one of the elements is the lower power process. The other element that enables us to do that is the stacked silicon interconnect, the SSIT, and that's another area that we're pioneering. So even though in terms of rollouts there's no doubt that generally, first availability of silicon is more or less equivalent to all of the companies that care to be at the front end, and we do care, it's the speed at which we'll be able to roll out the family, the breadth of the family and the transition between the different elements of the family which should be significant differentiators. The other element is we've been working on the ARM-based solution for a long period of time. We actually have emulation boards we provide our customers, and we believe that, that is a breakthrough product. Integration of the most popular microprocessor family in the world with the most powerful FPGA fabric will enable us to break into additional markets. So when you look at all of those elements, I'm extremely confident in our product position at 28.

Your next question is from the line of Shawn Webster with Macquarie.

So other logic vendors in the market have been talking about accelerating their operating expense growth. I was wondering if you could share your view on calendar '11 or fiscal '12 OpEx trends for Xilinx. And then I was wondering if you could quantify also what your lead times are currently. You mentioned they were nine weeks roughly for September.

Yes, so we're not forecasting our spending level for next fiscal year at this time. We'll do that in March. But we will have an increase because of the very rapid rollout of 28-nanometer, and those mass are quite expensive. So we will have an increase, and we'll share more about that in March. We are not planning, I would say, any large-scale headcount increases in the next fiscal year. So the dollars will be driven -- increase will be driven heavily by the mask and wafer increase and, to a much smaller degree, by any headcount and other kinds of spending at the company. And we'll be more specific about that in 60 days, or a little less than 60 days actually. On the lead time perspective, we have this metric that's the percentage of our product lines that are four weeks or less in terms of lead time, and the target we have is 85%. And in this current quarter that we're in, we're back into that milestone of being greater than 85% of our products are less than four weeks.

Your next question is from Mahesh Sanganeria with RBC Capital Markets.

Moshe, if you can just talk a little bit about your product with embedded processor. I know that you are ahead of your competitor in that area. If you could just talk about what significant applications you are targeting and what is additional market you can address and something on the timing of when does that becomes significant.

So this family was announced quite some time ago, and we haven't yet disclosed the details about the family members. But that program is being executed in parallel to the 28-nanometer mainstream, and it will definitely tape out this year. And we expect to provide samples to customers this year, too. It's a unique product in that it, for the first time, addresses the low and medium range of the market with a very powerful embedded microprocessor. And when you look at the embedded market, there is quite a few applications that are ideal for that. To mention just a few, if you look at Industrial, Scientific and Medical, if you look at Automotive, if you look at some elements on the Communications side, they actually are ideal for a product of this sort where the integration, the higher performance and the lower power are very important. So this is a breakthrough product. We have had embedded microprocessors in the past, but they were more targeted at the high end of the midrange and the high end of the market. This sort of goes to another place and should enable us to expand our footprint very significantly. We're having tremendously positive customer traction on this product. We have emulation platforms we've given to our customers so they can actually start developing software way ahead of the silicon availability. And it's probably one of the most exciting products to come out in the recent years in this logic space.

Your next question is from Sukhi Nagesh with Deutsche Bank.

Jon, can you just clarify a little bit on the inventory side again? Did you say that roughly $30 million of the buildup was due to the closure of the fab line, which should indicate roughly $40 million? The incremental $40 million was due to the higher node product, is that correct? Am I reading that right?

So the $29 million that I talked about is part of our $240 million of inventory. So the amount of that ending balance for the quarter in that was $29 million of this build ahead, if you will. And that $29 million number is going to increase over the next several quarters as well, and we'll talk about how much as we go through because it's about a timing with the line closure and the amount that we have to build ahead in order to intersect with our customers' need. I didn't really try to reconcile every bit of the inventory growth number, of how much was associated with the revenue drop. The $40 million was the decline in revenue from the midpoint that we originally gave in guidance. So the midpoint of our original range versus our actual results was approximately $40 million.

Your next question is from Srini Pajjuri with CLS Securities. Srini Pajjuri - Credit Agricole Securities (USA) Inc.: Jon, just wondering why the Turns business declined in the December quarter and why do you think it'll come back up in the March quarter. And also, one more on the gross margin side. Just wondering, you said that you achieved some cost savings in the quarter. Wondering how sustainable those cost savings are going forward.

Yes, the first one on the Turns decline. Because the downward impact of our -- or the revenue decline was at a customer that we have on a hub situation where they did demand pull essentially when they need it. The turns for us, it's not that the POs are in our backlog, all right? So all that meant was is that the revenue got lower by this $40 million. That didn't show up as a turn. And therefore our turns -- it was anticipated to. And therefore, our turns percentage was lower. So it really wasn't much more than that. And that's kind of the same story going forward as we have higher turns. The reason we're confident is we have a tremendous amount of our -- a large portion of our revenue increase associated with customers that have this kind of vendor hub situation. And therefore, as it becomes a turn as they pull it, there isn't a PO in our backlog. And that's really what's kind of making this turns percentage kind of a more volatile number than we might have experienced under more stable times in our history. Relative to the gross margin cost savings, I've said this before, we have a variety of projects at the company that are always trying to contribute. There were a lot of defect density improvement, price stabilizations, et cetera. Most of our cost reduction efforts are sustainable over some time period, but it's also -- the mix of our customers and the prices, depending on which those customers are up or down, does have an impact as well. So if I ever look at the two things that have the biggest impact on our gross margin, it is typically our customers/end market mix and then whether we've been able to sustain a cost reduction through, for example, sustained defect density improvement, et cetera. In this case, I think we're able to sustain most of this cost improvement.

And your next question is from Chris Danely with J.P. Morgan. Christopher Danely - JP Morgan Chase & Co: I guess can you just talk about your relative growth expectations for the four end markets you have for the rest of the year? And what gives you confidence that we're not going to have anymore, I guess, bugaboos in Con end market?

Yes, so Chris, we're really not going to forecast long term our end market at this point. We'll do that in March. I think overall, we do feel good about the Communications business, but there's a lot of other end markets that we need to look up at a little bit harder. So yes, a lot of bogeymen and bad things are going to happen to us on a macro basis. But we react to what our customers are giving us as long-term forecasts. And even though there's been some volatility in some places like Network Enterprise business and things like that and then this big Wireless thing, at least for the next couple of quarters we don't feel like there's anything going on out there. Obviously, macroeconomic issues have an impact, but we're getting very strong signals from our top customers. And they're very strong signals not just about their confidence in being able to sustain revenue levels, but also there's signals that they're adopting FPGAs more and more. And we're winning more and more complex designs and supplanting other technologies. As Moshe pointed out, some of the percentages in his remarks about the design starts and design wins and growth in areas, that we're outpacing ASSPs and ASICs.

Your next question is from Sumit Dhanda with Citadel Securities.

First, the restructuring charge that you're taking, can you talk about the timing and the extent of the benefit that you will see based on those charges in future quarters?

Sure. So the restructuring charge, all in for this last quarter of actuals in our forecast, is $10 million. And if you go back to the previous quarter's call, I think I indicated that we thought the total amount for the March quarter would decline over December, but it actually increased. And the biggest reason that it increased is that we are closing our software development operation in France. And I think many of you are aware how costly it is to extract yourself from a situation in that particular country, and the costs rose more than we had anticipated. So that isn't the only thing going on in the restructuring area, but we do expect a benefit in the neighborhood of $4 million, I guess, on an annual basis from that overall $10 million expense number. Now at this point in time, we don't anticipate any further restructuring charges in the following quarter. We don't have anything that we have. This is kind of the December and March quarters. The $10 million is it for our current plans.

Your next question is from Brendan Furlong with Miller Tabak. Brendan Furlong - Miller Tabak & Co., LLC: A question on Wireline side, just sort of something different. Seems to be a lot of buzz out there about the build out of 40 and 100 gig in the optical space. And I'm just wondering how you think you're positioned or the potential opportunity for that over the next year to 18 months.

That's an excellent observation. And there is this transition and it will happen. And the high end of the Virtex-7 product offering and the fact that it has the lowest power, the highest service, the highest bandwidth and it actually has the highest capacity in our mind puts us in a incredibly strong position to win the plurality of the business in that area. And the product was defined exactly to address the requirements and its own schedule. It will be rolling out this year, and we're in great position there.

And your next question is from Sandeep Shyamsukha with Auriga USA.

Actually, my first question was regarding if you could give me more color on the Wireline side of the business, how it fared in the December quarter, how do you expect it to do in the March quarter and the inventory situation at your largest Wireline customer, if that's impacting your business there?

So the Wireline business overall was pretty flat for us in the December quarter relative to September. And there were some adjustments going on in the network area. The telecom side was pretty flat to maybe slightly up. But generally, this kind of hit a flat spot with us. The uptake of enterprise, the network enterprise piece in terms of routers and data switches are still very strong for us. And I think you were probably alluding to the Cisco cautiousness that they've put out relative to their business, but at this point, we don't see a great deal of disruption in our business for, I would say, our core network business there. And we do think Q1 overall from an IT spending perspective, meaning Wireline, Cons, things that we sell into and, to some extent, Data Processing that we sell into still looks pretty good. And from a seasonal perspective, we still think IT spending overall is okay, which will contribute for Q1, meaning calendar Q1, the March quarter, and we should see good network business. So all their Data Processing segment.

Your next question is from John Pitzer with Crédit Suisse. John Pitzer - Crédit Suisse AG: Jon, just a clarification. I think I heard you say that the $40 million miss in the December quarter from original guidance was all mostly due to one customer in Europe. When you look at the March quarter guidance, do you expect to recoup all of that within the March quarter?

So John, all of the $40 million wasn't attributable to that. The largest individual contributor was that. But there was also some softness in China Wireless, and then I've pointed out audio/video broadcast was also softer than we had. So there were several contributors, but there was one large blip that made it such a large number for us. And I'm really not going to talk about specifics of our revenue forecast or any specific customer, but the forecast is quite strong for the Wireless business and we had forecasted up in Europe, and I think that's enough said about that particular area. So we're pretty bullish in that particular segment for the March quarter.

Your next question is from Hans Mosesmann with Raymond James. Hans Mosesmann - Raymond James & Associates: The dynamics on the foundry side of things that you're seeing tightness, is that at most of the nodes that you buy? And is that with multiple foundries or just one foundry?

I don't want to be a soothsayer for the foundry business too much in this particular comment. I'd like you to think about it as we got a little burned in not getting what we needed for our customers over the last couple of years. And so we're taking, I would say, a higher risk position on inventory to make sure that it doesn't happen again. And so there would be an uptick in the next year. But back to your point question, we are seeing full capacity in most technologies at most of our suppliers. And while you can still sit here and forecast and say, well, two quarters from now or three quarters from now the industry is not going to have that situation, we don't know that. And we've been surprised in the past, so we don't want to be surprised again.

Your next question is from Ian Ing with Gleacher & Company. Ian Ing - Gleacher & Company, Inc.: But for Moshe, first, can you talk more about the silicon stacking applications? Do you have to go really on the higher density of multiple FPGAs or would you consider combinations with other silicon vendors, I mean, vendors perhaps?

Well, the current usage and what we've announced enables us to have the largest FPGAs in the market, and that's because FPGAs are inherently scalable. This is true that if you take this to the next level, you can do additional things. And that is not lost upon us, but you know that it's not -- we can go into a long discussion. This is a very compelling topic and everyone has opinions. But at the minimum, this enables us to bring FPGAs out, which are larger, it enables us to reduce the complexity. It actually should enable us to impact defect density at the very low FPGAs. Just with the technologies that we control directly, it gives us a lot of interesting opportunities. Integrating other semiconductor elements, there's a whole host of issues related to that. But this is the first step that enables going in that direction.

Your next question is from Adam Benjamin with Jefferies. Adam Benjamin - Jefferies & Company, Inc.: I was just curious if we can just go back to the share discussion. You talked a little bit about the share stabilizing by the end of 2011. If you can clarify, are you talking in terms of share in revenue or in designs?

So yes, so let me -- I think you asked about my specific response. I was talking about share stabilization by the end of 2011 in the high-volume category. And I was talking about revenue, not designs. So some of the share loss that we're experiencing now is because of the high-volume base product family. And we have the only game in town, quite frankly, at the 45-nanometer, and our design win rate has been quite significant. And we start shipping in more volumes. The primary competitors, older technology will have peaked, and we'll be growing . And therefore, the share loss should stop and then we should be gaining share back in 2012 in the high-volume category.

Your next question is a follow-up from Uche Orji with UBS.

Just to understand your optimism around this product, how do you -- you're probably the first company we have heard talk about this a bit more aggressively. What gives you the confidence that this can be effectively manufactured?

Well, you're talking about the stacked silicon interconnect? Was that the question?

Yes.

This is a technology we've developed for several years, and we've worked with the entire supply chain. Now we've prototyped it. We're moving into -- we have the plan to move into volume production. The technology risks are very well contained, and we're doing this in a phased approach. And we're highly confident that those risks can be addressed with the approaches we've taken, and we have the entire supply chain in place in order to address it. So this is obviously a big step forward. That's what leaders do. And we're very confident that we will deliver there and facilitate the higher capacity products that our customers yearn. That enables us to even more rapidly grab market share away from ASICs because one of the few areas that ASICs have an advantage is at the very high end because it is a denser technology. But this way, we can actually compensate for that, and that enables us to eat into that market at an accelerated rate. So very confident that we've done everything that needs to be done in order to move that into the volume arena.

Operator, do we have time for one more question?

Yes, sir. You do have a follow-up from Sumit Dhanda with Citadel Securities.

Jon, the excess wafers you're running both at the leading-edge node and then for the end-of-life products, is there any benefit from small fixed cost absorption? I know you're a fabless operation. That's showing up in gross margins, which will be a headwind when you're done with the extra inventory build.

Nothing significant. We've done a lot of cost reduction over the last couple of years and particularly in that area. And as this kind of a throughput goes through, it's a pretty diminishing amount of overhead impact from a fixed cost perspective. So no, there's not a significant benefit. And then kind of a snapback, that's really not a material number at all.

Thanks for joining us today. We have a playback of this call beginning at 5 p.m. Pacific Time, 8 p.m. Eastern Time today. For a copy of our earnings release, please visit our IR earnings website. Our next earnings release date for the fourth quarter of fiscal year '11 will be Wednesday, April 27, after the market close. This quarter, we'll be presenting at the Goldman Sachs Technology & Internet Conference in San Francisco on February 15 and at the Raymond James 32nd Annual Institutional Investors Conference in Orlando on March 7. Lastly, we'll be holding our Analyst Meeting in New York City on March 14. This completes our call. Thank you very much for your participation.

Thank you. You may now disconnect.