Advanced Micro Devices, Inc. (AMD) Q4 2013 Earnings Report, Transcript and Summary

Good afternoon. My name is Marvin, and I will be your conference operator. I would like to welcome everyone to the Xilinx Fourth Quarter Fiscal Year 2013 Earnings Release Conference Call. [Operator Instructions] I would now like to turn the call over to Rick Muscha. Thank you. Mr. Muscha, 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 March quarter and then 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 documents the company files with the SEC, including our 10-Ks, 10-Qs and 8-Ks. These documents contain and identify important risk 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. Fiscal year 2013 was highlighted by sales growth from our new products, which increased more than 80% from the prior year, driving PLD share gains for the year. Although muted macroeconomic conditions contributed to an overall sales decline of 3%, Xilinx achieved a record gross margin in fiscal 2013. Gross margin of 66% was an increase from 65% in the prior fiscal year, a direct result of our continued focus on margin expansion projects across our product portfolio. In the March quarter, Xilinx sales were $532 million, up 4% sequentially and at the midpoint of the guidance we've provided. Sales from all geographies, with the exception of Japan, increased sequentially. Gross margin was 66.1%. Operating expenses of $204 million including amortization were $4 million lower than expected, due primarily to lower discretionary, sales-related and litigation expense. New product sales increased 13% sequentially, with particularly strong growth from Kintex-7 and Virtex-7 FPGAs. 28-nanometer product sales exceeded $40 million in the quarter, surpassing our previously stated goal of greater than $30 million. Let me now turn to a discussion of end markets. Sales from communications and data center decreased 1% sequentially. This was slightly weaker than we had forecasted. We had expected this sector to be up slightly, driven primarily by wired communications. The actual results were driven by a decline in wired customers, primarily on older products. New products, however, were up sequentially. The decline in wired was partially offset by wireless customers shipping both 3G and LTE technologies on a global basis. We continue to expect TD-LTE business to be a significant sales contributor in the second half of the calendar year. Industrial, aerospace and defense sales increased 7% sequentially, with strengths from all 3 subcategories. Broadcast, consumer and automotive was stronger than anticipated, with double-digit percentage…

Thank you, Jon, and good afternoon to you all. The 4% sequential sales growth in the March quarter was driven by double-digit increases on 6 of our secondary end markets. This broad-based strength highlights recovery in our business. Revenue was primarily driven by exceptional growth from our 28-nanometer product family, which exceeded $40 million, significantly surpassing our expectations. 28-nanometer sales were driven by both our midrange Kintex-7 and our high-performance Virtex-7 product families. We are now converting technology leadership to revenue growth. 28-nanometer Xilinx is a generation ahead of our competition in 4 areas. We have the broadest portfolio with our all programmable FPGAs, SoCs and 3D ICs. Second, we have the best products with an extra node of performance, lower power, high levels of integration and connectivity with the industry's absolutely best service. Third, and per our customers, critically more importantly, as our average design complexity has skyrocketed, we enable the highest productivity with Vivado, the industry's first SoC-strength tool suite. This enables delivering unmatched time to integration and implementation and, by far, the best quality of results, i.e. higher density, higher performance and lower power. Lastly, we enabled smarter systems with the unique combination of SmartCORE IP, C-based design tools and libraries, embedded software running on our market-defining ARM-based solutions. Leveraging all 4 advantages, revenue growth is already being driven by continued PLD share gains, considerable displacement of ASICs and ASSPs, aggressive market penetration and core target growth driver applications. Smarter 100-gig wired networks, wireless HetNets, data centers, retail and vision-based systems. In the March quarter, we announced multiple industry first, the 20-nanometer, clearly demonstrating between our -- this node to staying a generation ahead. Vivado Tool Suite now supports 20-nanometer designs. We are currently working with numerous Early Access customers. We'll be taping out our first 20-nanometer device this quarter as predicted. The rollout of our 7 Series and Zynq All Programmable SoC product generation is nearly complete with all 5 families in production. Fiscal year 2013, 28-nanometer sales comfortably surpassed $100 million. Over 500 customers receiving shipments in the March quarter alone in a broad base of applications including communications, industrial, defense and broadcast. In the June quarter, we expect our 20-nanometer shipment -- 28-nanometer shipments to, again, exceed $14 million in sales. As we discussed at our Analyst Day last month, we are very confident 28-nanometer sales will grow to more than $250 million in fiscal year 2014. With an improving demand environment, industry's broadest and, by far, best 28-nanometer portfolio, focusing penetration into the highest growth applications and significant progress in displacing ASICs and ASSPs, the new fiscal year promises to be an exciting year for Xilinx, here, where we will clearly connect technology leadership to revenue and demonstrate we're a generation ahead for accelerated growth. Let me now turn the call back to the operator to open it up for the Q&A session.

[Operator Instructions] Our first question comes from the line of John Pitzer with Credit Suisse. John W. Pitzer - Crédit Suisse AG, Research Division: Jon, in your prepared comments, you talked about a meaningful uptick expected in the back half of the calendar year for TD-LTE. The LTE cycle has been one that investors have been waiting for, for a while. I guess what gives you the confidence in the TD-LTE market? And if you could also just discuss the kind of the LTE build-out here in the developed countries. We saw AT&T kind of lower their CapEx for the out years. When do we start to see kind of the content pickup story for you guys within that spend?

Sure, John. So let's just talk about the TD-LTE portion first and I'll go to the second half after that. So we get -- the information we get is primarily from our customers. And the customers in China who are providing, or are the likely candidates to provide that equipment have given us forecasts that indicate in the second half we're going to have a -- the beginning of the ramp for that design. We have already sold into equipment where they're kind of doing their booked prototyping and their first-generation of testing of that capital equipment. And so I'm just putting these 2 things together. They're forecast for the second half of the year now, of course, these aren't binding forecasts, these aren't orders that are booked and all is subject to change. But usually what we've seen is the beginning of the prototype build followed within a quarter or 2 of the ramp. So the combination of their forecast, which means there must be some confidence that they want us to build enough parts for that as well as the beginning of prototypes is where I'm getting that information. The second question or maybe Moshe, do you have any additional color on the LTE before I...

Well, China is expected to happen in the second half, and all of the signals highlight that, that is the plan and is ongoing.

Okay. The AT&T announcement is kind of interesting. We obviously are pretty aware of that, of what was said. But when you peel back the onion of what's inside there and we have talked to the carrier about this, their program is associated with wireless rollout of LTE. Actually, it's preserved in terms of the kind of ramp they've been talking about in terms of capital spend and deployment across the country for the rest of this year and into next calendar year. So it's a very specific name to that program and we've gone and talked to them and checked with that. And our best information is, is that the wireless LTE portion of this is not substantially impacted by the statements that they made. John W. Pitzer - Crédit Suisse AG, Research Division: And maybe as my follow-up, guys, just quickly, another good quarter on the gross margin line, without yet seeing, I think, the benefit of some of the share gains you're expecting in 28-nanometer. I guess, given your leadership position is at 28, as that becomes a bigger mix of the overall revenue stream, is there opportunities to kind of get the margins above the higher end of your target, given that there's still a gap between yourselves and your closest competitor?

Yes. So there's no question we are continuing to do everything we possibly can to move the margin number up. The fact that we haven't changed our business model, I kind of anticipate the question on that as well. The fact that we haven't changed our business model is not really important relative to our view in the short term that we do think margins are going to continue to be healthy. I think one of the things we've talked about at the Investor Day was around the increasing importance of Zynq against ASSPs. And we're still working through those kinds of competitive situations. And in terms of changing our business model, I think we're just being cautious. But there's no doubt in the short run, as evidenced by our forecast for the June quarter, we are confident that margins have drifted up and continue to drift up a little bit if you go back over the last 6 quarters and plot it.

Our next question comes from the line of Vivek Arya with Bank of America.

One more on gross margin. One thing that Altera has done, had these products customized for different applications, i.e. with cost optimal structures, some of it in high-k metal gate, some without. And they have often explained that, that's the reason that their gross margins have tended to be higher. Is that a valid reason? Is that the kind of gap that you can close through your own design activity? Or do you think there is that structural gap that will remain?

So we have, over the past 5 years since the new management team has been in place, we've focused on improving the gross margins. And we believe that with the unquestionably superior portfolio we have and the broader portfolio we have and the deeper portfolio we have, we will get -- we audit for that and there is now the operational focus on driving defect densities and making sure that the overall portfolio has the right mix. So there's various ways of addressing it, and we like the way we've addressed it. It gives us technology leadership, market leadership and, unquestionably, a strong position. So I wouldn't say there's only one way to address it and the results speak for themselves. Ours have floated up. And despite the fact that we're now shipping a lot of 28-nanometer, the margins have gone higher rather than lower, which typically in the past it would go the other way when we are introducing a new generation of product. So -- but fundamentally, I believe that our overall portfolio strategy at 28 is absolutely winning and superior at this point.

And as a follow-up, there's some frequent concerns that the large Chinese equipment vendor has been trying to actively use more of their internal ASICs. I know it's not a new concern but it does come up frequently, and the related concern to that is that maybe for this initial prototyping part of PBLT, they might be using more PLDs. But as they go forward, that they might try to use more of an internal ASIC solution. Can you put that concern to bed once and for all, because I think that has -- it just comes up so frequently?

Well, we are confident that we have tailored our Kintex product to be absolutely a perfect solution for those wireless requirements. And what does happen is for the first few years, things change quite a bit, and the flexibility inherent in the FPGA is fundamentally a big feature that enables them to address that. And so between the flexibility they need and the very strong match that we have with our product offering to their requirements, and Kintex was targeted specifically at that, then I think we have a very strong position. Now any of our customers at any point in time can decide to do this. It typically is a risky, lengthy and complex thing to do, and it does happen from time to time. It just is becoming rarer and rarer. And so the number of companies that can do it has dropped. And even at the biggest names, and you've alluded to one of them, where they do undoubtedly have capacity, they tend to want to use it strategically where it has the biggest impact, and that biggest impact actually tends to be more on the consumer side where they have huge volumes and very short product life where they can exploit that most effectively. So I would never say never. I'm just saying it's more difficult and we're intent on making it more and more challenging as we move forward to advance nodes at the fastest pace possible.

And we're seeing no signs of this activity. So I mean, again, it could always happen in the future but I wouldn't think if -- you had specifically talked about, maybe right after prototyping. No way, I don't believe that's happening at all just for all the reasons Moshe said, in terms of the change that's still going on in the overall environment and et cetera. So I don't think that's going to happen.

Our next question comes from the line of Ambrish Srivastava with BMO.

First question is just getting back to the end markets and the guide. Jon, I think you mentioned that in the coming quarter, wireless will be down. But then I'm trying to reconcile that with TD-LTE builds happening. And since you did some prototyping, if they're happening in the back half and assuming they're strong, then what is weaker in wireless? So that's my first question. The second question is back to your business model through the cycle -- through the cycle, through all the top line volatility, the franchise has become more profitable. So could you please remind us again, and I won't ever fault you for not returning enough to shareholders, but the business is more profitable and you churn out a boatload of cash. So just remind us on the priorities on buyback versus dividend.

Ambrish, wow, you've covered a whole lot of stuff in that question. From an end market perspective, yes, I can sense why there's may be some dissidence or differences in those 2 statements that I made that you mentioned. But a lot of what's going on is just what our customer patterns of individual projects. So nowhere near the entirety of our wireless business is dedicated to TD-LTE. In fact, it's obviously very small at this point in time. And again, prototypes tend to be built in batches, and then there's a deployment and then there's contracts-led and then there's approval for it. So it's natural to have kind of a buildup in one quarter, a pause and then a kind of an initialization of ramp when you're talking about a new technology rolled out. So that's some of what you're seeing in the June forecast. And then very broadly, the rest of our business, which is related to both 3G and LTE rollouts elsewhere in the world, it's just the normal bumpiness. I mean, we're not going way down or way up. It's just down a little bit in wireless and then up on the wired section. So it's not really particularly very pronounced for either one of those 2 subsegments in the direction that we've forecasted. So hopefully that was helpful. From a business model approach, yes, we are very profitable. I will remind you that we did return $430 million of our $660 million of operating cash this year to shareholders, so a very high percentage of what we generated this year. And we are confident about growing profitability in the company, and we will make adjustments to our return -- our cash return to shareholders as those opportunities present themselves relative to growing cash balances and decisions that we make about how to do that. So just to refresh your memory on what our strategy is, the first preference is dividend, and we've raised our dividend every year since inception, and we raised it again here in the last quarter by a double-digit percentage. And the buyback is at least dilution or in the range of $150 million -- $100 million to $150 million every year plus opportunistic on top of that, which is really dependent on the stock price and the cycles that are going on. And so we have an algorithm that we use as a guideline, but not as an absolute, and we implement that as it hits our stock prices.

And our next question comes from the line of Jim Schneider with Goldman Sachs.

Just returning to the carrier CapEx commentary again. If you set China aside for a second and look at the broader market outside of China, if you total up what all the carriers have said about their spending plans for the year and look at what they spent so far, you mathematically conclude that there is going to be a big recovery in the back half of the year. So my question is, based on your customer forecast, do they basically support that snap back in the back half of the year or are you more cautious on whether that can really happen?

Jim, when we put together our forecast for the year this year, and we try to add this thinking into what we talked about in the Investor Day in March is I agree with you. It looks a heck a lot more bullish if you add up all the capital spending that's been published, and we've indicated a very strong snap back in the second half, and we haven't actually -- we haven't modeled anything near that strong. I view it -- I think it's going to be more of a steady growth situation, not some sort of a step function up for a couple of quarters. So I think we've been a lot more cautious.

Got it. That's very helpful. And then as a follow-up, you have noted a really large increase sequentially in Europe. I think Europe was up 14% sequentially, and obviously, that's way outside the bounds of what the broader market in something like analog is doing. So can you help us understand what's going on with the European sales? Is that all being driven up by step-up in your large European telecom customer? And what do you expect from that region going forward?

Actually, the growth in Europe was very broad-based, and the industrial improvement that we had forecasted in Europe came through. We had double-digit growth, I would say, in the mid to small customer range, just as we had forecasted. And it's nothing -- if anything, by the way, it was a little weaker than we thought in North America in that category, but in Europe, everything came out just right. So we had, in addition to this broader base industrial growth, which came through distribution support primarily, we also had strength in automotive this quarter that was focused a lot on that European market. So very strong business there. We had a few isolated aerospace and defense increases in Europe, greater than on a growth basis as well. Moshe talked about having multiple sub end markets growing. I mean, it really, this "recovery" that we had forecasted really was broad-based and it kind of got us back, I'd say, to the normal supply-demand kind of thing. I haven't been able to reconcile what other semiconductor companies have said in this, because I know there's been a lot of negativity going even into the March quarter around strength in industrial. We absolutely saw strength & industrial and in test & measurement and in broadcast and in automotive. And it wasn't just a large communications customer in Europe.

Our next question comes from the line of Glen Yeung with Citi.

This is Delos on behalf of Glen. I wanted to ask really quickly, if you can disclose it. How much of your cash balance is offshore and how much of your cash flow would you say is offshore as well?

Yes, I'm always hesitant to give you -- to give precise numbers on the offshore/onshore cash split because we do have ways of bringing cash back to the U.S. cash efficiently -- cash tax efficiently. Obviously, not all of it, some of it would get a 35% haircut. So I always bristle when someone or an investor says, "I'm just taking your estimate of your offshore cash and giving you a 35% haircut." That really isn't right -- the right way to look at it. In the neighborhood of 1/3 or so is onshore and 2/3 is offshore. But again, I'm cautioning you some of the offshore can be brought back cash efficiently. From a cash generation perspective, I think we've told you, definitely have talked about this before. We're comfortable with our U.S. cash generation, able to absorb dividend -- our dividend payment plus increases in the future and dividend along with supporting all our U.S. needs for cash. So we're in a pretty good position in cash generation in the U.S. and also pretty good position in terms of our overall cash balances.

And then regarding OpEx, it seems to be coming in a bit lower than expected. Are you being conservative there? Or -- and when does spending for 20-nanometer kick in?

Yes, so I think we're pretty much on our thought process in terms of our June forecast for OpEx. And it does grow throughout the year, so the second half of the year is greater because of tape-outs primarily. And there will be additional tape-outs for the next generation. I mean, there's still some tape-outs going out but it's on a relatively low level of finishing off the 7 Series and Zynq parts. And in the second half, we will get some uptick in the second half, nowhere near the kind of uptick we experienced in the last fiscal year, where we had our huge crescendo. It's a little flatter starting, if you will. So second half up from first half.

Great. And if I can just ask one more. 28-nanometer color, where is it doing well in the short run and where do you see it most successful in the long run?

Well, it's, by far, our broadest product offering. And if you look at the portfolio, it goes all the way from the low end, the Altix family, through Kintex, which is a wireless-oriented family, to Virtex and then there's the high-end of Virtex, which is primarily implemented with the 3D technology, where we're the only ones who are shipping that in production today. If you look at the breadth, it addresses all of the markets we're in. So there actually isn't a market that we play in which will not be served very well by 28-nanometer. If you look at timeframe, these markets tend to happen at a different rate. So wireless, the high end of test & measurement actually happened to occur quickly. Emulation, of course, happens quickly, too. So those are the 3 faster ones. And then the other markets tend to happen at their own distinct rates, where typically some elements of the aerospace and defense take the longest, but not all of them because even though they are markets that move faster than the average. But I would say it's a continuum, where the 3 that I mentioned are the fastest coming out. But based on the fact we've shipped to 500 customers in all of the markets we serve, it's coming out very broadly and strongly across the entire portfolio. And we expect it to manifest itself in being by far the strongest node we ever have had. And the early signs and the $100 million worth of revenue that was shipped this past fiscal year and the $250 million, which we feel confident in, are all manifestations of that.

Our next question comes from the line of Anil Doradla with William Blair. Anil K. Doradla - William Blair & Company L.L.C., Research Division: Yes, Moshe, I had a question on your Zynq product line. Clearly, there's a trend of ASSP replacement and Zynq is benefiting from that. Can you walk us through how you expect that to play out over the next couple of years? And do you think Zynq could be a 10% revenue generator for you guys?

Well, we -- it's a groundbreaking product and it addresses a lot of markets, which, again, happens at different rates. But for sure, it will be 10% or higher of the 28-nanometer generation. And because at any point in time, we're shipping numerous generations of product in parallel, it isn't likely to become 10% of our overall revenue anytime soon. But then there's the follow-on product of Zynq which will build upon that. So for 28-nanometer, for sure, it should be in excess of 10% above that. The way these markets work and the target markets to Zynq, some of them, for example, automotive and industrial, take a little longer to manifest themselves. So that will -- won't happen in the short term, but 3 to 4 years out, I would expect it to be well in excess of 10% of the overall revenue of the 28-nanometer shipments. Anil K. Doradla - William Blair & Company L.L.C., Research Division: So is it fair to say that initial deployments are on high end, and as you get out new iterations, you're going to work your way towards the mid and low end?

I wouldn't use high end and mid and low end here. I would say that it's market dependent. But Zynq is a mid and low end product by definition. It's not an ultra-high end product. It's basically focused on Kintex and Altix type sort of markets. But I would say within those, it's at the high end of those markets. That's what we've targeted Zynq at. And it's -- the design wins are phenomenal, and this year, it will start to be visible. The year we're in, the fiscal year we're in, it will be visible on the shipments. It just won't be at the 10% level yet by the end of this year.

Our next question comes from the line of Joe Moore with Morgan Stanley.

Can you give us a similar discussion of the 3D business today? Can you help us now that you're over $40 million a quarter, can you help us decide how big that exposure might be and how big it might get?

Well okay, let Jon -- I'll let Jon answer the first part and then I'll talk about where I think it could go.

Yes. So I mean, the 3D, which implementation generally describes our largest devices, and the largest devices go into a combination of end markets relative to test & measurement and emulation and prototyping, and then also high end communications for high-speed -- for lots of [indiscernible] and then also aerospace & defense and other in communications, where you can take out much larger ASICs. So those are all descriptions of businesses that many of them don't take off like right away. But yet the emulation and prototyping do take off relatively fast. And vis-à-vis the strength of our $40 million are definitely was a significant contribution there. But by and large, as Moshe talked about how strong wireless and other end markets have been, our midrange product has taken off much, much faster from the revenue perspective. So it is extremely important to us, that technology, and it has been a very strong contributor to the $100 million level, but it's actually not the biggest component of what we've been shipping up to now.

Kintex by far has been the biggest component, and there's been huge take-up of Kintex shipments in the wireless market, so that's the biggest. And then the high end of Virtex is the second, and then it's the normal Virtex product followed by Altix and Zynq are still a very small percentage point. Over time, we'd expect Virtex and Kintex to probably be neck and neck, the mainstream Virtex and the main stream Kintex to be neck and neck in terms of being the largest and then the high end Virtex to follow that.

Our next question comes from the line of Chris Danely with JPMorgan. Christopher B. Danely - JP Morgan Chase & Co, Research Division: With the 20-nanometer business being a little bit better than expected, can you just talk about your expected ramp for that in the June quarter and then the second half of the calendar year? And then, did you guys mention what that 40-nanometer/45-nanometer revenue did during the quarter? If you could talk about that as well.

Sure, Chris. The 28-nanometer ramp number next quarter, we said it would exceed $40 million again. So we exceeded $40 million in the March quarter. And in June, we said we would exceed $40 million again. And then just to remind you, Moshe said we would exceed $250 million for the year, which is consistent with what we said in New York at Investor Day. So that would lead a person to believe that there will be a quite a healthy ramp through September, December and March in order to achieve this greater than $250 million. So we do expect September to be approaching 10% of our revenue and continue to grow from that. We have -- again, these are predicated on lots of forecasts from customers but we have a very, I'd say, very detailed tracking around and monitoring of -- say, monetizing 28-nanometer at this company may be different than we've approached things in the past. And so we have a pretty good handle on what we think designs are ramping from our customers with respect to that technology. So we feel, I think, every bit as confident as we did in New York on this particular area. We didn't specifically say anything about the 6 Series. If you do the arithmetic on new products, quarter-on-quarter, new products is really made up of 28-nanometer and the 40-nanometer node. You would see that the 6 Series was in the neighborhood of flat quarter-on-quarter. So we do expect the 6 Series to grow next quarter, in the June quarter, and it starts to get flattened out per quarter and then it will start growing. Christopher B. Danely - JP Morgan Chase & Co, Research Division: Great. And as my follow-up, I guess just one more question on capital allocation. So if I look at your share count from what's basically fiscal '07, '08 until fiscal '11, it went down at a pretty healthy rate. However, now that fiscal '13 is in the books, your share count's actually gone up a little bit each of the last couple of years, so along with your cash balance, too. So I was just wondering like how you guys look at that? What are the plans to deal with sort of the share count creep going forward here?

Yes, the biggest dropper of our share count increase is the dilution effects of the convertible that the accounting world says we have to take but the call option we have to protect us on the strike price can't get recognized in the P&L, so on terms of the diluted share count. So it's actually a little bit of tyranny of accounting. If you net all that out, you would've seen a slightly declining share count for us. But because our actual stock price has been increasing, which is a good thing, over that same time period. You got to give some credit for stock price going up during that time. Now that doesn't necessarily wipe away your base question. I'm just trying to make sure that you understood what the mechanics were relative to the share count. So absolutely, we see our cash balances creeping up slightly, not to a great degree, and we have had kind of over 2 years of flat return of cash to shareholders in the neighborhood of 425 to 450 or 420 to 450 kind of a range and so we're quite aware of that. I'll also point out that we distributed in the neighborhood of 2/3 of our operating cash for the year, which is actually a very high payout ratio. So we do look at a combination of payout ratios and our cash forecast in order to make that determination. And we're very aware of it and we have continuous conversations with the board about it and, again, we are continuing to review those particular alternatives. I'm not going to give a statement today that we're going to change what we're doing right now.

The next question comes from the line of William Stein with SunTrust.

From your comments earlier in the call, it sounds like the booking pattern during the quarter improved, do I have that right? And then I have a follow-up.

I think it kind of came out normal, kind of, I'll say, a normal quarter. We had a strong beginning and an okay third month of the quarter. And the second month is always a little lighter, vis-a-vis the first month of the quarter. So I wouldn't say it was exceptionally strong at the end. I think it was kind of average.

Great. And then to a degree there's normal with regard to seasonality anymore, can you remind us what you'd view as kind of a typical September quarter change sequentially?

Yes, September is -- has been -- has a lot of -- has had a lot of different patterns for us in the past. We always have this effect of the industrial and smaller customers, particularly in Europe, getting a lot softer for us in September, in the September quarter. So June has been up. And I don't know if you look at history and seasonal, we're in the range of where we've been. And September sometimes has been down. But I'm not making any statement about September. Again, with all the CapEx spending growth that we believe is going to happen, it's way too early to know the timing of some of those programs to know about September.

Our next question comes from the line of David Wu with Indaba Global Research.

I've got one on Zynq because that's a new business for you and you're replacing existing processes. How should we think about the profitability of that business versus the FPGA business?

At this point in time, what we've seen is it's pretty -- right now, it's pretty close to what -- from an end market by end market basis, to what the average is for each of our end markets. Now acknowledging that the first wave of business came from existing FPGA customers who are now substituting an FPGA and a processor for a Zynq product. So it's, I'll say there's some cannibalization going on relative to our FPGA business, but it is value-additive, because now we're getting the dollars for the processor as well. And so by and large, we aren't sensing any sort of downward pressure or difference. We have some designs where the value proposition is so high, we've been able to get a premium for it and in some situations where it's a more price sensitive.

Yes. And just to complete the thought there, the processor business is huge, and it goes all the way from the dirt-cheap microcontrollers to the ultra high-end, which obviously has very handsome margin. But we've targeted the middle of the market and so we have very intentionally stayed away from the profitless prosperity of the ultra low end, and we're making sure that the product, in of itself, targets markets where we believe that it's generally in line with our FPGA business. We're not targeting, at this point, neither do I expect to see in the future to target those ultra low end, dirt-cheap markets where we can sell billions of these but it's unlikely to enable us to sustain our model. So if you look at the way the product is designed, if you look at the power of processor there, which is actually a very formidable dual processor core, you can see why we're pretty confident in our ability to do that. And that's why we're not saying this is a couture product, it isn't. It's targeted for markets which are more in line with our typical markets.

I see. Can I ask a quick follow-on. Do you treat that, the process CPU market, as a separate line item in your own internal business? Or do you see that as an ASP enrichment? So really it's, P&L-wise, it's still folded into your FPGA business.

Well, we look at it both ways that when you introduce a new product, which really is revolutionary and we were the first to do this way ahead of anyone with regards to product of this sort, and we now have been followed happily into this. But quite fundamentally, at this point, when you have a new product, you want to isolate it so you can accelerate your learning and it is looked at separately at this point. I do believe that over time, these markets do tend to converge. And if you look at other technologies we've introduced in the past, where they were totally separate or they address a small niche of our applications, over time, those technologies have become widely used. And so DSP cores are now widely used, SerDes are widely used. So over time, I do expect it to converge and become closer to our core markets.

Just one quick follow-up. Is there a need for you to expand into the 64-bit addressing that the ARM camp have just recently announced and does it help you in your customer base?

More technology is always better, right? So I won't say it doesn't help us. But if you look at the product we have, it really has very broad applicability. And there is an emerging 64-bit market. And as this progresses, I wouldn't be surprised if, over time, we look at ways to address that. But this isn't a short-term need for our core market. Our core markets are more than adequately serviced by the existing architecture and its capabilities.

The next question comes from the line of Ian Ing with Lazard Capital Markets.

This is actually Tyler Radke on behalf of Ian Ing. My first question, as it relates to TSMC's 16-nanometer pull in. I'm just trying to understand how that changes your strategy with respect to 20-nanometers, and can you just kind of walk us through the puts and takes of that.

Okay. Why don't I answer that and if you could kindly raise your voice when you ask the second question, because you're very barely audible to us so we can hardly make it out. But we are very pleased with our success with 28-nanometer. We were first in the market. We have the broadest portfolio by far, we have the deepest portfolio by far, and we've demonstrated that even with a totally new foundry, we can come out and demonstrate leadership. Similarly on 20-nanometer, we've already announced some elements with regards to tape-out, where we believe we're going to be the first FPGA company to tape-out a 20-nanometer software. The software is there already. It's being used broadly by customers and a whole host of other things, which gives me tremendous confidence that in 20-nanometer we're going to be able to continue and extend that leadership and keep our generation ahead. And we have not announced anything yet on our plans beyond that, but rest assured that we are not resting on our laurels here, and we expect to continue to deliver leadership to our customers, and we will communicate that and our strategy when it behooves us. At this point in time, most of the design wins or a large plurality is in 28-nanometers. We'll continue to be there in that for some years going forward and then immediately after that, there's the 20 and we will make sure that we have a response for anything beyond that at the appropriate time.

Okay. So my second question, my follow-up, I don't want to beat the dead horse here but just as it relates to wired and carrier CapEx, obviously, I think you guys said wired was down more than expected but you expected it to be up next quarter. And as you look at obviously, AT&T and Verizon CapEx coming in a bit lighter than they'd hoped, do you think this is a function of late releases of their budgets and this stuff getting pushed out into June or is there something else, your customers are saying that we should be thinking about?

Our specific method is and then our confidence about the next quarter was directly related to only a couple of customers' programs, that the timing was off from what we had thought. So I want to make another statement in a second, but I was going to make sure that it was clear not to read too much into my next statement, because there really is a macroeconomic overhang where people -- where the confidence in spending kind of ebbs and flows almost on a weekly or monthly basis. And that does end up with caution in our customers' ordering patterns and they always say they want to expedite and do something. And we are finding it harder and harder to nail down forecasts and trends on key programs just because of that secondary factor. But again, the fact that we were down, lower than expectations, and we're going -- when are we going to come back up some in wired is really related to a couple of specific programs at a couple of specific customers.

And with regards to LTE deployment, one way of looking at it, and probably the best way is, as a customer, regardless of who's winning and who's not winning on the smartphone world, that world just continues to go through the roof. It generates huge demands on the infrastructure. Those infrastructural demands can only be addressed by, over time, moving to the newest generation technology. And to a large extent, we're seeing a commitment, a strategic commitment to do that and it's in each country, region. It happens at a different rate, but it definitely is happening. And to some extent, when you have a major -- in all of these companies, they have numerous carriers. At any point in time when one pulls ahead, it generates tremendous demand for the others to catch up. Otherwise, their service levels become unacceptable. And we're seeing that happen. So even if from time to time, they do need to watch their deployments, we see more of those dollars being directed at new technologies as opposed to upgrading old ones. So given our very strong 28-nanometer position and the larger percentage that we get, that tends to help us. So I am very bullish on that and I don't see that trend reversing itself anytime soon.

Our last question comes from the line of Srini Pajjuri with CLSA. Ryan Goodman - Credit Agricole Securities (USA) Inc., Research Division: This is Ryan Goodman for Srini. Just as a follow-on to that last question on wired, so it sounds like the macro is weighing a bit on maybe the enterprise demand -- enterprise demand within wired. Could you give an update on how the data center demand looks? And also last quarter, I think you had mentioned a large data center high-speed connectivity win that was supposed to ramp in the second half. Any update on that as well?

Yes, so I think you characterized the wired accurately there. And relative to data center, our data center market is a new area for us, and it will take a little bit of time before it grows at a more accelerated rate. But when it does, some of the design wins we won already and are about to win is going to start to take off at a more accelerated rate, but that's going to be in the back half of the year and into next year. We have won the design -- any of the designs that we've talked about in Investor Day in New York, we're still either on track to win or have won because things are going very well, yes, and it is around high-speed connectivity as our biggest value add there. Now we have been in solid-state drives fairly broadly as controller units. That business continues to be very good for us across a variety of customers. But it ebbs and flows on individual customers' shipments, et cetera. But by and large, we're where we wanted to be on data center, but it's not the big part of our communications business at this point in time. Ryan Goodman - Credit Agricole Securities (USA) Inc., Research Division: Okay, great. And then just for another area, in industrial, looked pretty strong across the board. Just curious, how much of that do you think is coming from inventory replenishment versus sell-through demand? And how do feel about customer inventory levels now? Are you comfortable where they are going forward?

Yes, I think given the fact that we're maybe a little bit of an outlier of calling the industrial recovery and we had it, I think quite a bit of maybe in the neighborhood of 2/3 was inventory replenishment. So we are seeing strong -- signs of stronger continued ordering patterns from some key customers there, which do lead me to believe there is sell-through going on in that particular end market category. I do not believe there is much inventory in the channel in that end market at all at this point. No build up.

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 website. Our next earnings release date for the first quarter of fiscal year '14 will be Wednesday, July 17, after the market close. This quarter, we'll be presenting at the 2013 Barclays Telecom and Media Conference in New York City on May 22; the Bank of America Merrill Lynch Global Technology Conference in San Francisco on June 5; and lastly, at the William Blair Annual Growth Stock Conference in Chicago on June 11. This completes our call. Thank you very much for your participation.

Ladies and gentlemen, this concludes today's conference call. You may now disconnect.