Advanced Micro Devices, Inc. (AMD) Q1 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 First 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'll provide a financial and business review of the June quarter and then we'll open the call for questions. This quarter, we changed our revenue by product category classification. We make changes to these classifications periodically as products mature and new products are introduced, enabling the product categories to be more meaningful to investors. For comparative purposes, we have provided a supplemental schedule on our website at www.investor.xilinx.com that presents results based on both the new and old category classifications. 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 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. Xilinx's sales were $583 million in the June quarter, up 4% sequentially slightly better than the midpoint of our guidance. The quarter was characterized by strength in Asia-Pacific, particularly in communications applications, marginally better than anticipated shipments of last time buy inventory and expected weakness in North America driven by defense and industrial, scientific and medical. Turns were 55% for the quarter and relatively linear on a monthly basis. We continue to get the capacity we need from our foundry partners, and our lead times with few exceptions are within normal ranges. Gross margin was 66%, at the high end of our forecast due primarily to continued new product yield improvement. Operating expenses of $220 million were in line with guidance resulting in an operating margin of 28.2%. New product sales were up 31% sequentially, with strong growth from our 28-nanometer, 40-nanometer and 45-nanometer families. Sales of our 28-nanometer products increased significantly during the quarter surpassing our target of $10 million. On a year-over-year basis, new products are up nearly 80% with most of this growth coming from our 40- and 45-nanometer product portfolios, which continue to gain momentum as designs ramp into volume production. Mainstream products increased 4% during the quarter, but are down 16% versus the same quarter of the prior fiscal year and base products declined 5% during the quarter, but are down 10% on a year-over-year basis. Let me now turn to a discussion of end markets. During the quarter, we modified our end market categories in 2 ways. First, we moved all data center customers into the communications category. And second, we renamed the categories to be more descriptive of the applications contained within each category, as well as more reflective of how we run the business. The communications category will now become communications and data center. Sales from this category were particularly strong during the quarter increasing 8% sequentially driven by strong wired communication sales and double-digit sales to wireless applications. The category that used to be called industrial and other will be now called industrial and aerospace and defense. This category decreased 5% sequentially with decreases from industrial, scientific and medical and aerospace and defense. The consumer and automotive category has been renamed broadcast, consumer and automotive. Sales from this category increased 12% sequentially during the quarter driven by strength from audio/video broadcast and automotive, while pure-play consumer applications were essentially flat. Lastly, the data processing category, which has been a relatively small business for Xilinx over the last several years, was renamed other. This category will continue to contain applications such as high-performance computing, server and computer peripherals. Sales from this category increased 11% sequentially. Net income for the quarter was $130 million or $0.47 per diluted share. Other income and expense, with a net expense of $9.7 million, higher than anticipated due primarily to lower than anticipated income from our investment portfolio and slightly higher interest expense associated with the accounting treatment of the convertible debt instruments. Operating cash flow for the June quarter was $163 million before $8 million in CapEx. We paid $58 million in cash dividends during the quarter and repurchased 3.1 million shares for $99 million. The tax rate in the June quarter was 16%. Diluted shares for the quarter were 274 million shares. There was a 6 million share dilutive effect from our convertible notes. For questions relating to the dilution associated with the convertibles, please visit our Investor Relations website at www.investor.xilinx.com. Let me now comment on the balance sheet. Cash and investments increased $16 million to approximately $3.1 billion. We have approximately $1.3 billion in convertible debt, and our net cash position is approximately $1.8 billion. Days sales outstanding increased 5 days in the June quarter to 40 days. Inventory dollars at Xilinx declined by $12 million sequentially during the quarter. Combined inventory days at Xilinx and distribution together were at 99 days, down from 110 days in the prior quarter. We are now within our target of 90 to 100 days, but expect days to increase slightly in the September quarter. Let me now turn to a discussion of guidance for the September quarter at Slide 13. Our backlog heading into the quarter is down sequentially, due in part to the completion of our last time buy program, which we expect to negatively impact us by approximately $25 million. This is a somewhat larger number than we had originally anticipated given greater than expected last time buy sales in the June quarter. Additionally, we believe macroeconomic uncertainty may result in unpredictable customer ordering patterns. That being said, we are expecting continuing growth from our new products as we continue to enjoy strong demand for products in all technology shipping in this category. From an end market perspective, we expect communications to be down sequentially. Within communications, there will be certain pockets of growth such as wireless in North America and most data center applications. However, we expect these areas of growth to be offset by declines isolated to a couple of major end customers as they adjust internal inventories. Within industrial and automotive -- excuse me, industrial and aerospace and defense, we expect defense and test and measurement to decline with ISM to be relatively stable. Lastly, we expect broadcast, consumer and automotive to decline driven by consumer and automotive, with sales from broadcast expecting to increase. As a result, we're expecting total sales to be down 4% to down 8% sequentially with sales from all geographies decreasing. The midpoint of our sales guidance is predicated on a turns rate of approximately 57%. Gross margin is expected to be approximately 66% consistent with the June quarter. We continue to make progress on margin improvement projects across our product portfolio with particular emphasis on new product margin improvement. Operating expenses in the September quarter expected to be approximately flat at $220 million, including $2 million of amortization of acquisition related intangibles. Other income and expense is expected to be a net expense of approximately $8 million. The share count is expected to be approximately 274 million shares. The tax rate is expected to be approximately 16%. Let me now turn the call over to Moshe.

Thank you, Jon, and good afternoon to you all. While there continues to be uncertainty in the current economic environment, I'm very pleased with the sales increase of 4% in the June quarter. This growth was driven by double-digit increases in half of our secondary end markets. New product sales were exceptional during the quarter, increasing more than 30% sequentially. This growth was driven by across-the-board strength from both our 28-nanometer and 40-nanometer, 45-nanometer product families indicating widespread customer adoption in all application segments. Particular sales from our 28-nanometer product family grew significantly in the June quarter exceeding our $10 million milestone. There is no question that we are realizing the advantages of our highly differentiated breakout portfolio that includes not only FPGAs, but also our pioneering all programmable 3D ICs and SoCs. The growth in June was driven by significant increases in revenue shipments of Virtex-7 and Kintex-7 FPGAs, Virtex-7 3D ICs and Zynq SoC products. Our 28-nanometer product family rollout continues at an accelerated pace. Not only did we achieve a significant revenue milestone for the family through the announcement of the first shipments of our Artix-7 family FPGAs, we are now shipping members of the entire 28-nanometer family. The Artix-7 devices extend the reach of 28-nanometer FPGA technology to applications requiring low power, low-cost programmable devices with the highest performance in the industry by far for this category of device implemented in a small form factor. Artix-7 devices are securing design wins in a large number of applications, including portable, medical, handheld radio, small cellular base stations. This past quarter, we also began shipping for revenue the Virtex-7 H580T FPGA. This is the industry's first 3D heterogeneous all programmable product. Virtex-7 HT devices use our stacked silicon interconnect technology to deliver the industry's highest bandwidth programmable devices, featuring up to 16 28-gigabit per second transceivers and 72 13.1-gigabit per second transceivers, making them the only single-chip solutions for addressing key 100-gig and 400-gig line card applications and functions. With the April announcement of our Vivado Design Suite, we now have all 4 key elements of about 28-nanometer portfolio in the market. All programmable FPGAs, SoCs and 3D ICs coupled with the next-generation design environment enable these next generation devices. The customer acceptance of Vivado has been outstanding, with more than 200 beta customers designing with this new software suite, which provides up to a 4x productivity advantage and an improvement of 15% in terms of quality of results over previous generation development environments. 28-nanometer design win momentum continues at an unprecedented pace. We continue to see customers move away from ASICs through programmable imperative accelerated this node. The same time, we're integrating and displacing evermore ASSPs, as mid-tiered vendors struggle to fund and deliver on their next-generation devices, while our 28-nanometer portfolio is enabling a whole totally new level of programmable systems integration coupled with game changing price-performance per watt parameters. Additionally, with our breakout 3D Virtex-7 ICs and our Zynq-7000 SoC products, we're penetrating markets and applications that traditionally were not served by FPGAs, and there is no PLD competition there yet. We have demonstrated indisputable leadership with our 28-nanometer portfolio. We have now shipped parts from all 5 of the 28-nanometer product families to hundreds of customers further demonstrating the very broad acceptance of the family. Having achieved the $10 million sales milestone in the June quarter, with sales expected to exceed $20 million in the September quarter, I'm confident that this 28-nanometer node will, by a very significant margin, be our most successful node ever. Despite some near-term concern in the macroeconomic environment, our clear 28-nanometer leadership puts us in a phenomenal competitive position that will enable continued share gains, against both ASICs, ASSPs and traditional PLD competitors. 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 Ian Ing with Lazard Capital Markets.

You talked about 5 out of the 10 submarkets being up this quarter. Can you talk about which end markets are sort of most or least below their prior peak in demand and given the guidance, are we done with inventory replenishment and restocking at this point?

Yes, I'd say from a peak perspective Ian, this is Jon, I think from a peak perspective, wired is a little bit below the peak and ISM is a little bit below the peak. I think those are a couple that are, I would say, marginally softer, but slightly softer. And then the computing aspect of things has been down a little bit from its peak, but generally others have been up. I'd say so wired and wireless down a little bit, I didn't mention wireless, wireless has been down a little bit because obviously there was some bigger build-up when both China and U.S. we're going at the same time. So it comes down a little bit and then pure industrial down a little bit. And did you have a second part of that question, Ian, I'm sorry, I missed that.

Just your sense as -- is inventory replenishment and restocking over at this point or if there's still...

Yes. I may make some general comments about our posture and positioning here. And I think from an economic perspective, we really are seeing -- we aren't seeing anything really negative out in the market. I mean, orders have slowed, but yet cancellations and push outs have not increased. And my view here is there isn't any -- there is not significant build up of inventory in the overall in any of our end markets, with some exceptions in the communications segment. There are a couple of customers that I think have been procuring inventory a little bit ahead, and they're going to work through that in the next quarter or 2. I think that's the only place where we see anything. And that really isn't across the board, that's just at a couple of selected customers.

Our next question comes from the line of Brendan Furlong with Miller Tabak. Brendan Oliver Furlong - Miller Tabak + Co., LLC, Research Division: Just a follow up on that question, I'm wondering if you could kind of breakout, is it on the wireless or wireline side of this inventory piece, a few major customers that you have are building some inventory?

Yes, it's wireline, primarily. Sometimes it's hard to tell because we do ship to some customers that have both wireless and wireline, but it is a little bit more wireline than it is wired. Brendan Oliver Furlong - Miller Tabak + Co., LLC, Research Division: And then I guess the second question would be on the industrial side, which had a -- we just rebounded in the March quarter, I guess in the June quarter with some inventory refresh or replenishment to the cycle there, do you think that's now kind of a little bit completed or do you think there's any legs left in that part of the business?

I think it's completed, and I think everyone's being really cautious about ordering. I wanted to emphasize our -- the fact that our lead times are in great shape on a relative basis so there -- my belief is that people are being a lot more cautious about stocking inventory, and they know they can get it in a relatively short amount of time. So if I go back to look at some of the other cycles we've gone into, we've seen push outs and cancellations increase ahead of the revenue drop, and we really aren't seeing that. So I think it's possible we've been cautious on our guidance for the coming quarter, but I do think we're going to be going through -- living through periods of up and down here as people make a decision on whether the economy is slowing down or not. Yet we still see CapEx spending very strong in the wireless area and even some wired initiatives going on in China. So it doesn't feel unhealthy to me. It's just feels like a lot of caution. Brendan Oliver Furlong - Miller Tabak + Co., LLC, Research Division: And I'll sneak one last one if I could on the P&L. R&D was supposed to peak, I guess, in the September quarter, is that still the expectation, I mean, in the back half of the fiscal year we see some decline in the overall OpEx rate?

Yes, so total OpEx I believe will still be lower in the second half of our fiscal year than in the first half of the fiscal year. That being said, I think OpEx will move up a little bit from the $220 million in the December quarter, but then decline in December. So we're going to have a little more flat, instead of having a more pronounced peak like peak, it's going to be a little more flat than a decline.

Our next question comes from the line of Chris Danely with JPMorgan. Christopher B. Danely - JP Morgan Chase & Co, Research Division: So Moshe put out some 28-nanometer revenue targets, can you just give us a sense of how you expect 28-nanometer to ramp after this quarter and then if you could provide what the 40-, 45-nanometer revs were in the previous quarter and how you expect that to ramp as well?

Yes, well, we aren't making any specific forecasts for this year, Chris, on 28-nanometer, but we do expect every quarter to increase on a going forward basis throughout the rest of this fiscal year. We're seeing very strong uptake from our customers even for samples, which is an extremely positive view relative to history on 28-nanometer. So we would expect increase. I know we're forecasting going from exceeding $10 million to exceeding $20 million and on a percentage basis, that's pretty good. I doubt that, that percentage increase will continue, but I do expect significant growth on a quarter-by-quarter basis for the rest of the year. And our 40 and 45, the whole new product, the new product category now is defined by being 28-nanometer and 40, 45 technology solely. So there's nothing else in there. So we grew overall new products over 30%, so 31%. And so I think you can probably do the math since we gave you the 28-nanometer number, you can figure it out pretty well. Christopher B. Danely - JP Morgan Chase & Co, Research Division: And for my follow up, I guess a bit of a longer-term question, so if we look at your guidance for the September quarter and compare it with a year ago, revenues have bounced around, but it's roughly flat. And yet OpEx is up, I want to say like almost 15%. So can you just talk about your longer-term plans on how to get a little more leverage out of the model and to get the OpEx down towards your target and how do you feel about -- like the growth of the PLD industry and the growth of Xilinx these days?

Yes, so there's no question that we are spending ahead of the revenue growth level at this point in time and that's really primarily to leverage our leadership at 28-nanometer to make sure we get everything out as fast as we probably can and that speed and then the expense of the technology are driving that. And we've talked about in our long term revenue growth model of 8% to 12% growth. I mean, we need -- in order to get ourselves into model, we do need to grow at least at the 8% growth, and certainly the midpoint of that 8% to 10%, 5-year CAGR would put us in a position where we would comfortably be back in the model in my opinion. So we do need that, the top line growth in order to make that happen and to the extent that it happens, then our model holds, to the extent it doesn't happen, then we'd have to reevaluate, but we're not in that situation of thinking about reevaluation at this time.

And we're confident that the strong position on 28-nanometer is what will drive that growth, hence the ongoing investment to make sure that we don't compromise that.

Our next question comes from the line of Romit Shah with Nomura.

Jon, can you just explain this last time buy issue demand? I'm not familiar with it. It looks like it's having a meaningful impact to your guidance in the quarter. Is it a problem that goes away after the September period?

Yes, well, this is just the obsolescence of old technologies that after 12 to 14 years, we tell our customers that since volumes are relatively low we're not going to continue to ship small quantities and hold the manufacturing capability in our foundries beyond that, and we'd give them a long notice to say, better to tell us all of the units that you need and then take them, and we give you a last date that you have to actually receive them all. And so, this is basically a 2-year cycle that goes on, so every 2 years, we have this kind of process going on. And in this particular cycle from that technology, yes, we have had some elevation in the normal run rate for those products, which we've characterized as approximately $25 million impact in the September quarter. We will be shipping a very limited number of those products from those technologies on a go-forward basis. So yes, after -- starting in September, we won't be shipping very many of them.

And then on North America infrastructure, it was an area of strength for you last quarter, and I think you guys are saying it's going to be up again this quarter. Considering that this market tends to be lumpy, are you at all concerned that maybe it fades as the year progresses?

Yes, so we're still very bullish about wireless. Our customer's customer serving North America, and we believe that even though if you just look at the CapEx spending between Verizon, AT&T, Sprint and T-Mobile, at least Verizon and AT&T on a year-over-year basis, you said, while the CapEx spending really isn't going up that much why would Xilinx improve? Well, if you look inside of where they're spending their CapEx, they aren't spending on wired. They're spending more of it on wireless this year than they did last year. And so we're benefiting from their increased focus on 4G rollout in the U.S. And Sprint continues to be deploying as well and then T-Mobile in the second half of this year starts to kick in, in North America. So we are -- even though many of those units are shipped at different geographies for build and then shipped back into North America, we are still very positive on the rest of this year's impact with respect to our business.

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

Jon and Moshe, first, just a little bit more clarification on what I'm hearing from you guys versus what the customers are saying because on the wireline side it seems we have seen a parade of misses from the com equipment guys, and they have specifically called out North America spending slowing down, so you are saying that North America wireless is trying to think. Ericsson also talked about LTE becoming stronger going forward, but just reconciling that with what you were saying and more importantly, you mentioned that no "alarm bells" are going off and maybe you didn't use that word, but what then sets the alarm bells for you guys when you look at the order patterns, kind of what are the metrics that you look at and say, "Uh oh, this is not what we were thinking," i.e. it's much worse than what we were initially thinking. And then my quick follow up was rather tied to what Chris was asking, interestingly enough, when you go down another node and we heard about multiple patterning and how that entails a higher cost, would that keep your OpEx higher as you go to the next node of Virtex?

Okay, so back on the comment on infrastructure business and particularly in North America that wire -- actually, wireline is a little softer. It will go down in the September quarter we believe and is wireless and that is the biggest issue positive issue for us. Wireline overall is, I'd say, stable in most places, down a little bit in North America and then down a little bit in Asia-Pacific. So it is kind of a mixed bag for us, it's a net down. But wireless, while wireless is -- went up quite a bit in the June quarter, a very healthy double-digit percentage, we think we'll be able to kind of hold that level, if you will, and there are these additional things that are happening with other carriers that we believe is going to give us some strength. So I know it's a little confusing because it's up some places, down some places, but we think we're in a really good position in this cycle with both, 65-, 40- and 28-nanometer technologies across the board winning designs at the key suppliers supporting LTE technology. Then your ordering pattern question. Yes, I know, there's all these things written in the paper, there's been a couple of companies that have come out and say things are softer in industrial, and they might miss their numbers and those kinds of things, and I respect that and that's why I think we've had a relatively cautious forecast, meaning down because we do think some things in industrial are going to be much more challenging. The data points that we look at are the rate of cancellations and push outs, the rate of orders, particularly when you're talking about industrial, how it shapes up from our distribution channel and what their forecasts are and what they're seeing, and we look at that by geography so we get a sense as to whether it's one geography or broad-based, a broad-based scenario. And it's entirely possible that we aren't seeing it yet and it's yet to come, but I also don't -- we talk to customers, we aren't getting the feeling that they're canceling their orders. And our backlog's down some, it's not down gigantic amount. So there's a lot there that could push out and cancel if they wanted too. So we look at all those data points, we review all that weekly, and we've taken our best shot of what we think the quarter is even though I know there's all this macro uncertainty that you read about. And I try not to read the papers too much because it can get you depressed and try to use the data to guide the business. Then the third question's around node...

Yes, why don't I take that? So definitely, 20-nanometer is going to be more difficult than 28 and similarly 28 was technically more challenging and more expensive on a per device basis than the 40-nanometer. And the way we are addressing it is we are being ever more diligent in making sure that the return on our engineering investment is there to support the level of investment. So generally speaking, I think that it should more or less remain at the same level, but we will be more careful with regards to tape-outs and we'll make sure that everyone counts as they become more expensive and more difficult and the design investment, it's incumbent upon us to make sure that the returns are there to support it. So generally speaking, our product planning is more meticulous at the 20-nanometer node than it has been at the 28-nanometer node and similarly at the 28-nanometer node was a lot more disciplined than it was of the previous node. It's one of the reasons that we believe we have actually hit the ball out of the park on the 28. The products were designed very well, and they're more or less getting that response from the customers. They're getting the right level of performance and power for each of the target markets, and we just need to be more diligent because you can't, if the design costs go up 2x or the market costs go up 2x, then you definitely can't afford to have the same number of tape-outs and hence we are being more careful and more meticulous about the planning process.

Our next question comes from the line of John Pitzer with Crédit Suisse. Ryan Carver - Crédit Suisse AG, Research Division: This is Ryan Carver in for John Pitzer. Just following up on that last comment, given the focus on the return of engineering investment on the allocation of engineering resources for future nodes, how does that coincide with your stated intent to be pursuing ASIC and ASSP sockets, which traditionally have a lower ASP? I mean, how do you balance this resourcing for future nodes towards the right efforts with also pursuing traditionally lower ASP sockets?

So that's a good observation, one doesn't preclude the other. We just need to be -- to the extent that we're targeting or replacing ASICs and ASSPs, which is happening en masse basically. Clearly, much faster on the ASIC front than on the ASSP front, but its happening on both of those areas. You just need to make sure that the devices you do come out with are targeted to compete effectively and have the right performance and power and set of features that the customers need. So it's incumbent upon us to do that. Having said all of that, you need to realize that these same parameters apply to the ASSP and the very, very, very few companies that are still pursuing ASICs and typically only for extremely high volume consumer-type application. And their design costs are going up, so they can't continue to blanket the earth and start designs willy nilly, and they're not. And what we're seeing is an ongoing accelerated transition where less and less companies are pursuing those angles on the ASICs front and on the ASSP front to be very well documented that the companies that are moving forward, are moving forward very aggressively, and we are one of those who strive to be at the leading edge, but more and more are holding back and saying, "Well, we'll design at the new nodes 2 to 5 years from now because we can't afford it anymore, and at that point in time the market costs will go down and the yields will be at their peak, et cetera, and we'll be able to make the financials work." But at that point in time, we will, if we continue to move forward, we'll have a big advantage versus those companies that are holding back and that is part of the programmable imperative and it continues at least at this point basically unimpaired and full steam ahead. Ryan Carver - Crédit Suisse AG, Research Division: Great. And then just sort of going back into the coms business. If you look back to the Q1 timeframe, I mean, you guys were up sequentially in coms, where your comp was sort of down 20%, and it seemed to be a pretty big gap in similar sectors, albeit, different geographical mixes and now you guys are talking about sort of a down September quarter on large customer inventory burn down. So I'm just curious, how much of the strength in the Q1 and the Q2 timeframe calendar-wise was attributable to perhaps pull-ins versus sort of organic demand?

Yes, I don't think there's any impact on any pull-ins to any large degree. We did call out, there was a couple of customers that built -- that has some inventory issues that we believe they built ahead. But on a broad-base, there really wasn't an exacerbated calendar Q1 or calendar Q2 impact from any of that. We were participating in a recovery in the March quarter in a very large way as customers very broadly across all of our end markets replenished their inventory, and I think we've had more, say, typical growth out of an inventory cycle. I'd say it's very typical for us that you'd have to ask a competitor about their particular profile and issues. Ryan Carver - Crédit Suisse AG, Research Division: Great. One last if I could, on 28-nanometer relative to March quarter expectations for the growth, I mean, how does that growth trajectory look? And I guess what I'm thinking about is, given early in the year commentary that 28-nanometer would be sold out through the balance of the year, how has your 28-nanometer trended relative to what you had perhaps told your foundry partners at the beginning of the year and are there any issues in terms of needing more product that they be can potentially supply or being forced to take more product than you actually need?

Well, we projected some time ago that when we'd hit $10 million this quarter, and we're delighted to have done that, and that's in line with our plans. For several quarters, we expected this to be a $10 million quarter. And similarly, we're looking at the $20 million mark and for the quarter we are in, the September quarter and as Jon alluded to, we expect growth, and it could be significant growth through the rest of our fiscal year because this is a rocket ship, and it's one of those things which when it takes off, given the nature of our business, it doesn't turn on immediately, it starts layering. And we're seeing an acceleration, which is the fastest we've seen, which we strongly believe is a function of the technology position we have and the breadth and depth of our product offering, plus the very significant advantage this gives over the previous generation of product in terms of much lower power and much higher level of -- much higher density or gate count. So fundamentally, the product offering is very strong. We're seeing that translate into design wins. Now design wins, those are very big numbers, they're in the billions of dollars at this point, but what really matters is revenue and that's the number we're fixated on, and that's the number that really matters. And we feel very comfortable in our fab partner, TSMC, at 28-nanometer being able to cover for us, and you need to keep in mind that these are great numbers for us, but these are still small numbers in terms of wafer volume for a large fab like TSMC. So they have the capacity to address this, and we do not see any issues either this year or to be honest next year in terms of meeting our demands.

Nor do we have any deals, any relationship that forces us to take a certain amount of wafers out. You made some comment that I didn't really -- maybe I had misinterpreted, but that's not how our relationship works where we're required to take it. So I wasn't 100% sure where you were going. First it was sold out. You said it was sold out, implying that we didn't -- we couldn't get capacity. Well, we have what we need to meet our revenue needs, and I guess that's all I'll say about it.

Our next question comes from the line of Uche Orji with UBS.

Moshe, first of all, congratulations on the progress you've made on 28-nanometers. Can you give me, is it possible for you to provide some color as to the relative strength of this business across the 3 segments you mentioned, the stacked silicon and the high end product content so that I can just kind of get a sense of where the strength is coming from, why is it different across the board.

Okay. So no, we have 5 general product families and 2 of the 5 are unique, and they are product families that we pioneered, and we believe we have huge as in a year or more and potentially could be significantly more lead over the competition. Those 2 product families are SSIT, the 3D stacked silicon; and the Zynq product line. The Zynq product line is still, in terms of number of units shipped, it's quite significant, but it's a small ASP. The stacked silicon tends to be a smaller number of units, but it's quite significant for devices of that complexity, and it is a significant part of the revenue that we have shipped. In addition, when you look at the 3 other product families which are the mainstream ones, which is the mainstream Virtex-7. That is basically doing extremely well, is ahead of plan and given that it has a significant power advantage over the competing products, plus we actually have higher overall bandwidth going in and out of those devices. We're seeing a tremendous level of design wins, which is starting to translate to revenues. But that's not yet huge as part of the $10 million number. So the largest components of the number are the SSIT and actually the Kintex product offering. But very shortly, we expect the Virtex to translate into a significant number and actually if you look at the overall Virtex family it will be the majority of the revenue at the node will be the Virtex and the SSIT, which is the high end of the Virtex. And then the Zynq is sort of an expansion play, which is on schedule and we already have in our mind a year at a minimum of lead over the competition and that is starting to move into volume production in the foreseeable future. So it will be a major part of our revenue coming soon, but it will be less than either Virtex or Kintex as a stand-alone. Artix, we have sampled the product and that tends to be the low end so in terms of the overall percentage of the revenue that will likely be the smallest of the --

Of the mix?

Of them all. Hopefully, that answers your question.

It does. It does. Let me just ask you a difficult question, part of the strategy to drive growth is to retake share in the 40-nanometer category, and you've grown that over just -- in the quarter that just ended. Any way for us to calibrate your progress and if we have any sense as to where your share was and where you like it'd to catch so that we know how to calibrate that, that would be helpful.

Jon could just...

Let me try that. So quarterly share is really dependent on customer mix and things like that relative to us versus our competitor, and so when I did the Analyst Day, as you may recall I talked about the fact that in calendar year '11, we had 35% share at that node, and we felt on a run-rate basis over the next several years it was going to move up to 50%. Last quarter, I'll say the March quarter, make sure we're in the right quarter, March quarter we estimated to be in the low to mid-40% because we shipped quite a bit, our competitor didn't ship as much. This quarter, I don't know what this quarter is going to be, but we had very significant growth, as we said, we had over 30% growth of new products and the overweighting dollar value of our new products happen to be in 40-, 45-nanometer because we're still very fresh with the 28-nanometer, $10 million or so. And so we had a very, very strong growth of 40-, 45-nanometer so we're very happy with where we are in terms of these key designs that are moving into production and starting to take volume. And I think because we've modified our new products to be only 2 technologies now, it's going to be easier for you guys to figure out what the growth numbers are, good, bad or different at any given quarter as we provide that data and information.

Sure. That's helpful. And just one last question please, if I may. Moshe, can you just provide a bit more insight as to what's going on within China, ZTE, which is one of your customers, preannounced recently I think on Sunday night and that if I think about the various phases of events within China, it's something a little difficult to kind of get a clear view. So any sense as to how you see the China coms business developing and the road you expect to play in that will just be my last question.

Well, China's been a focus area for us, and it clearly was a smaller business for us than it was for our primary competitor, and I think part of that was due to a geographic focus and part of it was due to product focus. And I have no doubt that the Chinese companies and the 2 biggest ones are very well known are going to do very well in the communications business, and we feel that we have by far, the best product offering which addresses their requirements. So we don't know who the winners are going to be, but in terms of percentage of business for us in China, it should be much higher than it has been in the past and that will be driven by 28-nanometer. And it's already manifesting itself in terms of design wins and actually significant parts of the $10 million was shipped into China. So best we can tell that's going in the right direction for us.

Our next question comes from the line of Tristan Gerra with Baird. Tristan Gerra - Robert W. Baird & Co. Incorporated, Research Division: Could you talk about the -- what you see in terms of small cell in terms of FPGA content? How meaningful that could be later in the year and whether we should be concerned on second generation of those small cell that there could be some migration back into ASIC or you think that FPGA has a very solid position in those new products or new class of products?

So what is becoming clearer is that they are going to be heterogeneous networks, which have a mix of macro cells and small cells and when we talk to our key customers in this area, we're hearing more and more that they expect that to be the norm, and it isn't going to be one replacing the other, but both of those will need to coexist, and they expect their end customers to deploy networks, which has a mix of those. And so as such, we are continuing to see a big investment in wireless base station infrastructure buildup which uses FPGAs. In those cases we actually see that the FPGA is becoming a larger portion of the bill of materials as our customers painstakingly point out to us every time there's a negotiation. And so generally speaking, we're addressing what is a larger market. Portions of it will definitely be addressed by small cells and those might be either done as ASICs or there's a whole host of ASSP companies that are probably overpopulating that arena and hoping that it will grow at a large rate. We expect those to coexist and we actually play in some of those markets with our products, but it's not in the high-volume consumerish places. It's in -- for a larger number of users that's where FPGAs are used to implement that. So this announcement on the death of macro cells is, and I'm paraphrasing, very premature at this point in time and everything we hear from our customers implies that they're going to be deploying massive amounts of macro cells in the future, both short, medium and long term, and it's not a one technology obviates the need for the other, they both will coexist.

Yes, we're actively designing for some concept with FPGAs as a player in certain categories, in certain numbers of users of the cells -- of small cells, which if you get more educated, everybody's got their own version of a small, medium and large, but we believe we can participate in the medium-sized and the larger sized small cells if you want to get to be real basic in that. So we do think that we are -- we do believe it's going to grow and that category is going to grow, and we think we have a play in them as well.

Our next question comes from the line of James Schneider with Goldman Sachs.

I was wondering if you could talk about on the wireless side, some of the geographical trends you're seeing and you expect over the next couple of quarters there's been a lot of mixed data points about what's happening in India and China and in other geographies including Europe. Can you maybe talk about what kind of deployments you're expecting? Do you think those are slowing down or you see them as largely unchanged in the end of the year?

Yes, so let me take that one, Jim. Over the next 2 to 3 quarters, I think there's a bigger opportunity for carrier spending first in North America then in Europe and then maybe third in China, and then India would be last. And I know Europe might sound odd, but some of our data points seem to indicate that there is a need to have more CapEx and to start to upgrade the network, which is starting not to keep up in some of the suppliers in particularly they have not even really touched the LTE technology at this point in time and there are carriers that are now making the call to invest even in the face of -- for their biggest macroeconomic uncertainty is, which is in Europe. So if those actually -- if they really go forward with that, I think that will be a very positive thing for us as well. I've talked a lot about North America and how strong it is with all the major carriers here making investments. Even Russia, for example, we believe there's a big play going on in Russia to start to roll LTE and again that's a very positive thing relative to the customers where we are. China seems to be the biggest potential consumer, but not a whole lot of movement going on about when things really happen, and when things are going to take off. And then India, I have no new information on India, just seems to still be uncertain of when it's going to happen.

That's helpful color. And then maybe as a follow up, normally in the December quarter, your defense business is up rather strongly based on the visibility you see in products today and how you think that's going to shape up, do you still think that's going to be the case this December quarter or do you think there is a potential that the fiscal issues could -- and the budgetary issues could have an impact on that this year?

Yes, there's always an issue. There's always a chance that the fiscal issues could have some impact. But typically, it's more the volume or our revenue is typically projects that are well in flight somewhere and sometimes that can be impacted by the short term freezes and things like that, in other times not. I would still expect a better December in aerospace and defense than in September. That's my expectation, but I have to admit I have not looked at it on a customer-by-customer basis yet. We're saying we're down at September versus June, and I would -- my expectation is that we would be up again in December in that category.

Our next question comes from the line of Ross Seymore with Deutsche Bank.

Just a question on the magnitude of your guidance by segment, about 4 out of your down 6 points in your overall guidance is coming from the last time buys, if we take that out of the equation, you talked about your main end market categories, can you give us an idea of the magnitude of how you believe those would perform? Kind of what would be best, what would be worse, et cetera?

Parsing out pieces of our revenue and not having a way to explain it to you on an equal basis is actually a little dangerous thing for me to do. I don't think I'm going to give non-GAAP information about our end markets that we publish on our filings and things like that by taking certain things out. I will -- I can tell you that, that last time buy has impacted several -- we ship things to several end markets, but industrial, automotive, aerospace and defense and some communications and kind of in that order. Those are the ones that were impacted the most by it. So you might imply from that, that oh, since industrial is going to be kind of flat, not counting aerospace and defense, then gee if you didn't have last time buy maybe it would have been up a little bit. So I'm not trying to be difficult, but I don't really think I want to get into non-GAAP analysis.

Got you. And then one that might be a little bit straightforward, I guess, just a clarification of an earlier comment you made, Jon, on the OpEx trajectory for the rest of the fiscal year. Just to make sure I have the fiscal and the calendar quarters right, did you say that the December quarter OpEx would actually be up a little and then the bigger fall off to get the back half down year over or half over half would be in March or was it actually going to be down in the December quarter as well?

It's what you said first, it will be up, it's going to be up a little bit in December quarter and then down in March.

Our last question comes from the line of C.J. Muse with Barclays.

I guess first question just a point to clarify. In terms of end of life, what was the sequential downtick in September that's negatively impacting your guidance?

$25 million. Okay, we'll ship $25 million less of that category of inventory in September then we did in June.

And that will fully be resolved end of September and should not impact revenues going into the December quarter or will there be a sale?

Correct. The tail is really, really small. So essentially it's -- that's it. Now we're done with that, with that technology.

Excellent. And then I guess a question on 28 and thank you for the great color there in terms of what's driving the growth in the mix there. And I guess my question is on the stacked silicon side, can you talk about what the kind of growth trajectory looks like there, my understanding is that emulation is the biggest driver for you guys, and I guess curious what the sustainability of growth looks like and how we should think about that part of the business exiting calendar '12 and moving into calendar '13?

So you're right in highlighting that emulation is a driver, but it's not the only driver and so if you look at who would use stacked silicon, you need to look at the -- what it provides customers and basically provides the customers with higher capacity and higher bandwidth than they can achieve any other way. And to some extent, it generates a market in of itself because the only other way to achieve this would be to design a hideously expensive ASIC and in some cases, even if they went to the most advanced process node then they would need to design for these to achieve these high frequencies, which is in of itself a very significant challenge. So clearly, the people who use it first are ASIC prototyping, but actually then it moves into other areas, which are wired communications, test and measurement and actually A&D, and it's all in the high end of those markets. So this is not a one quarter wonder and then the whole thing goes away. It does not really go away until there's an alternative, which if you use traditional forms at the very earliest, is available at the next process node, and it's available at the next process node only when the largest chips are available at the next process node, which typically take some time to get there. So you can count on this having a significant vibrant life and emulation are the early swallows, but then full-fledged spring starts with all of the other applications that use it, and it should have a 3-year life cycle just driven by the fact there is no other alternative. And as such we are very pleased with the opportunities it opens us.

That's very helpful. If I could just sneak one last question in. Jon, what is -- and I know it's hard to calculate because nothing's normal these days, but what is typical seasonality for Q3?

Q3 calendar, Q3 fiscal, trying to make sure it's September.

Calendar.

Yes, September quarter. As you look back at our history, it's been up a little. So pretty flat and of course the standard deviation is actually pretty broad. We've had some times where we've been down a whole lot and up a whole lot, but if you look at averages over the last 5 years or 10 years, it's really in the neighborhood of kind of a plus or minus 2% range.

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 the IR website. Our next earnings release date for the Second Quarter Fiscal Year '13 will be Wednesday, October 17 after the market close. This quarter, we'll be presenting at the 2012 Citi Technology Conference in New York City on September 5. This completes our call. Thank you very much for your participation.

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