Advanced Micro Devices, Inc. (AMD) Q1 2014 Earnings Report, Transcript and Summary

Good afternoon. My name is Jay, and I will be your conference operator today. I would like to welcome everyone to the Xilinx Q1 Fiscal Year 2014 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.

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 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. In the June quarter, Xilinx sales were $579 million, up 9% sequentially and better than forecasted. Most of the sales upside was related to better-than-expected growth in wired communication, with aerospace and defense sales also stronger than expected. Sales from all geographies were up sequentially. Gross margin was higher than guided and a record 69% for the quarter. Roughly half of the gross margin upside is related to the result of margin expansion programs, which are yielding benefits faster than previously anticipated. The other half of the upside is product mix related. Operating expenses of 261 -- excuse me, $206 million includes amortization were in line with guidance. We were able to offset higher variable sales and profit-related expense with overall spending reduction. Operating margin was 33%, the highest since Q2 of fiscal year '11. New product sales drove most of the incremental sales growth during the quarter, increasing 24% sequentially. 28-nanometer sales exceeded $50 million, again, surpassing our forecast. Sales from this family were driven by all 5 family members, with Kintex family remaining the largest contributor. Our 40-, 45-nanometer node also posted double-digit sales increase, with strong growth from both Virtex-6 and Spartan-6. Mainstream products decreased during the quarter, and base products increased as a result of strength in specific aerospace and defense programs, as well as accelerated sales associated with the previously discussed foundry line closure. Let me now turn to a discussion of end markets. Sales from communication and data center increased 8% sequentially, better than forecasted as a result of strong wired communication sales. Wireless sales were flat sequentially, slightly stronger than we had anticipated due to a pickup in North America-based LTE activity. Industrial and aerospace and defense sales increased 10% sequentially, with strengths from all 3 subcategories. Defense sales were…

Thank you, Jon, and good afternoon to you all. The 9% sequential growth in the June quarter was driven by increases in 6 of our secondary end markets. New product sales were exceptional during the quarter, increasing nearly 25% in the March quarter, 75% on a year-over-year basis. This growth was driven from both our 28-nanometer and our 40-, 45-nanometer product families, demonstrating widespread customer adoption in all application segments. 28-nanometer sales comfortably exceeded $50 million, higher -- significantly higher than our expectation. Sales were driven by growth from all 5 product families, led by our midrange Kintex-7 and all components of our high-performance Virtex-7 product families. In the September quarter, we expect 28-nanometer to continue its accelerated growth, exceeding $60 million in sales. We remain confident that 28-nanometer sales will surpass our fiscal year goal for 2014 of $250 million, a goal we had established at our Analyst Day in March. The accelerated sales ramp and tremendous customer adoption clearly demonstrate that we have established a proven technology leadership formula with our 28-nanometer portfolio. With Vivado, we've created the industry's first SoC-strength tool suite, which enables unmatched time to integration and implementation, and a very significant improvement in the positive results over previous development environments. Second, leveraging our world-class partnerships with both TSMC and ARM have created the broadest portfolio of All-Programmable FPGAs, SoCs, and 3D ICs, offering an extra node of performance, lower power, superior connectivity and an absolute breakout in programmable systems integration. Lastly, we're enabling smarter next-generation system and an even better alternatives to both ASICS and ASSPs, with a combination of SmartCORE IP, our C-based design tools and embedded software running on our industry-leading ARM-based solutions. We recently announced the tape-out of the semiconductor industry's first 20-nanometer device and the PLD industry's first 20-nanometer All-Programmable device. We also implemented the industry's first ASIC-class programmable architecture called UltraScale. UltraScale devices enable next-generation smarter networking equipment, smarter vision systems, high-performance computing and intelligent surveillance and reconnaissance systems. These devices, when coupled with our Vivado design suite, will enable Xilinx to deliver another 1.5 to 2x system-level performance and integration value advantage, well ahead of any competition. In the June quarter, we also extended what has been world-class partnership with TSMC by announcing that we're working on a program called FinFast to create the fastest time-to-market and highest performance FPGAs to be built on TSMC's 16-nanometer FinFET process. We expect to deliver test chips later this year with first product coming in 2014. Having established proven leadership formula 28-nanometer and now expanding that formula with our next-generation architecture and product milestones, we are very well positioned to drive even more share gains against both ASICs, ASSPs and traditional PLD competition. Let me now turn the call back to the operator to open us up for the Q&A session.

[Operator Instructions] Our first question comes from the line of Ambrish Srivastava with BMO.

Jon, my question, the first on gross margin. You said half is structural and half is product mix. So beyond the next few quarters that you have guided to should be -- what's the right way to think about it? 150 bp is variable, but the 150 from structural stays. And then I have a follow-up as well, please.

Yes, so we have -- we clearly have been focusing a lot on margin improvement programs across the company, as I said in my remarks, that starting -- some of them are coming a little earlier than we anticipated. And so what I was saying in the call about the gross margin forecast for the next few quarters, we do expect it to be in the neighborhood of 69% at the June quarter. And it is driven by a combination of continued structural improvements that I expect will continue to improve in -- as we go throughout the year, and then the impact from the product mix will be a little variable as you go quarter-to-quarter. But I would expect over this coming quarter, that impact starts to decline a little bit. And because you have one positive trend and one trend that's moderating a little bit, we think we can keep the margins in the neighborhood of 69%, flat for a while.

Okay, okay, that's helpful, because -- so the key takeaway is that the half from structural is not done yet. You're still working on that part.

Oh, absolutely, and I see additional benefits out in time for us there.

Okay, good. And then a follow-up for you, Moshe. As we think through, and especially in the context of what Vivado is helping you do, as we think through the next node, 28 and beyond that, where does the mix for the business going to look like? Because right now, most of the time, we spend on the comp side, but is industrial and auto and all the other segments finally getting to a point that you see them becoming a bigger proportion for the -- and then what would the mix be in 28 and beyond?

So 28-nanometer is the broadest and the deepest node we've ever had, and it benefits all of our businesses. But they tend to benefit from it at a different pace. They all benefit from it, but the ones which clearly have a lead in our communications at this point. And of course, some ASIC prototyping, which was the very early product offering, and that continues now we're sort of seeing other markets like test and measurement start to pick up, and we expect wired communications to pick up too. You are right in that if you sort of look at it from 30,000 feet, [indiscernible], then communications is the biggest benefactor at this point from 28-nanometer, and the other markets will benefit over time. And the Zynq products offering, in particular, will have a very significant impact on industrial, scientific and medical. And then some of the others like military, aero just tend to take quite a bit longer, as does medical to translate into revenue. As you move forward, the product offering tends to have a higher end component to it where communications will benefit more from that than some of the other markets. So I would say that the 28-nanometer, you've yet to see the other markets emerge in force. It's just starting. It's less than 10% of our revenue at this point in time, and it's growing rapidly. 20 will lead with communications, and it will take a little longer for the other markets to manifest themselves.

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

This is Sanjay Chaurasia for Romit Shah. Moshe, one question I have for you, if you could provide any color on the wireline side of the business. It seems like it was the driver. Any color around if it was tied to any specific deployment and to service provider in enterprise [ph] or in technology, 10G versus 40G would be great.

Sanjay, let me grab that one. So first, I would say that our wireline and data center business was very -- the increases were broadly distributed around customers. So if you look at the top 20 communications customers, of which there are quite a few wireline contributors in that more than the wireless, is pretty concentrated on a few names for us. But just the wireline, more than 75% of those -- of the customers that are in the top 20 showed an increase this quarter. So it wasn't just one technology or one application. We have made tremendous progress in 40-gig and 100-gig, since you brought those up, and particularly with our OTN-related IP is driving us to a tremendous number of design wins in that category. And some of those are starting to get into a preproduction kind of phase. But I have to say that if you look at the names of the tier 1s and even tier 2s, with communications customers in the wireline side, we saw quite a bit of growth, very broadly distributed.

The next question on the line comes from John Pitzer with Crédit Suisse. Ryan Carver - Crédit Suisse AG, Research Division: This is Ryan Carver in for John. Just a question on the linearity in the quarter. The data points seem to imply that April was a pretty good quarter and May was a pretty good quarter, but maybe there was some falloff in June. Can you talk about -- a bit about your linearity in the quarter and just kind of, if you saw any of that or something different to that?

Ryan, June was actually pretty -- the quarter was pretty linear from a historic perspective. And in fact, at times, we've seen June fall off towards the end as it goes into the summer from -- particularly from an industrial perspective and, in some respect, some of the communication equipment. And we didn't really see any kind of fallout whatsoever. So June for us was, I would say, a very, very solid quarter because -- I mean, we ended up being over the top of our guidance a few points, right? So it certainly showed strength throughout the month of June, maybe more than we had modeled as we walked into the beginning of the quarter. Ryan Carver - Crédit Suisse AG, Research Division: Okay. And just as my follow-up question, on the OpEx side, when you took up -- I mean, I think the guide for the September quarter was a bit above expectations. When you took up the full year numbers, it looks like around a lot of R&D. But even as a percentage of revenue, it looks like it came up a little bit. How should we think about sort of R&D, specifically, the percentage of revenue? And -- I mean, is this the new run rate that we should expect? Or is this sort of a peaker, a peaking time around 20-nanometer and should expect that to fall off? And I guess, how should we think about this sort of OpEx levels kind of trending over the next couple of quarters?

For the near term, our OpEx will -- the second half of the year by quarter will look pretty similar to our forecast for the September quarter. So we're forecasting $225 million. It's going to be pretty flat, maybe up just a tad in the second half on a quarterly basis but relatively flat to that. And what's happening underlying that is we start to get incrementally more mass expense in the second half of the year than the first half of the year. And the first half of the year ends up with more variable impact from profitable variable kinds of things, bonuses, et cetera, because of the -- that's directly tied to our op margins from an employee bonus perspective. And so what's actually happening in the second half is the underlying I&D -- R&D, not the variable piece, but the contribution from developing and launching products goes up somewhat and the variable expenses moderate a little bit.

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

I'm hoping you can dig into the TD-LTE opportunity for us. It seems like we didn't see the strength from that this quarter as anticipated in the next quarter, also not a big impact but something bigger, you said, starting in December. Can you help us think about the potential upside there? And as we've seen the bid for one of the big carriers, investors have contemplated competitive moves by the others to drive a more robust growth opportunity in that end market. Can you comment as to whether you've seen any of that?

Sure, let me -- we'll be happy to do that. The China TD-LTE business, we've been saying for some time that it was a second half of our fiscal year growth for us. So the December quarter and the March quarter for us was where the ramp would take place, so beginning in the September-October time frame. And at this point in time, we still have that same view that's been pretty consistent for us for the last 6 months, but that's been our view. Even though I know various other companies and other parts of Wall Street have tried to forecast acceleration, that will be pulled into the summer, et cetera, we've never really seen that. We've shipped in the March quarter and -- particularly and a little bit in the June quarter, we have shipped in with some prototype and early production for some of the early city trials that they're doing. And we certainly acknowledge the fact that the bids have now gone out at least for one of the suppliers, and so this process is moving along, I would say, pretty much as we had thought it would. And we're not -- I'm not trying to say we've got a great crystal ball and we know precisely what's going on, but it is relatively consistent with our view of the world. Now relative to the speed of the ramp, now we still believe there's the 200,000 base stations that are going to get deployed, and there's a lot of conversation around, are some of them upgradable from the old generation and the new generations? Those are very difficult things to know because it depends on which manufacturer, OEM, gets the business and which -- whether the provider decides to pursue some sort of a hybrid upgrade strategy. There's all these different articles written about how it could be done and whatever. At this point, it's not really clear how that's going to work out. We're very confident with our design win levels at all the manufacturers that we believe will be providing units. And again, we see the ramp beginning in the September, October timeframe for us.

The next question comes from Christopher Danely with JPMorgan. Sameer Kalucha - JP Morgan Chase & Co, Research Division: This is Sameer Kalucha calling in for Chris Danely. One question on the share count. It has been creeping up, so I'm wondering how much is that from the convert? And how does it -- or how do we expect it to trend forward? Are there going to be any buybacks to bring it back to the earlier levels?

Well, clearly, in the last several quarters, the impact has been a great deal from -- disproportionately from the convert versus other sources that have been driving it up. Now we do have some anti-dilutive protection in one of the 2 debt offerings that, from an accounting perspective, doesn't end up showing in the diluted shares. But that being said, it's really being driven more by the converts than anything else. And as the stock price goes up more, then more of it gets into the money in terms of the bigger way. We are still committed, as we've said consistently, to a share repurchase and a dividend program to return cash to shareholders. And we are still committed to buying back at least $150 million of stock for fiscal year, and I'll just leave it at that.

Next, we have Anil Doradla with William Blair. Anil K. Doradla - William Blair & Company L.L.C., Research Division: A couple of questions. Moshe, how would you address the concerns from some quarters, saying that the industry can skip 20 nanometers and go from 28 to the 14, 16 FinFET? And the follow-up question was, was there any end-of-life for last-time buys? During the quarter, did you benefit from that?

Okay. So no, if you look at the -- our current plans, we expect the next offering to be split over 2 nodes. It will be our UltraScale 20 and our UltraScale 16. And we are basically making sure that for the product offering, which benefits from the ultimate in terms of performance, then it's the 16 node, that, that will be a more expensive node and hence, the more cost-sensitive applications, are better serviced by 20-nanometer. And we're just splitting the family -- our next family across those 2. There'll be a very broad product offering at 20-nanometer and being first, and we believe actually having a larger lead now than we had at 28-nanometer node. We expect to have a very strong position at 20-nanometer. And we'll continue with our plans and committed to moving forward. And we, being in the lead, other companies have different strategies. But generally, I believe, for FPGAs, there's a huge benefit from -- going from 28 to 20, and that will be available -- as we tape out now, that will be available to our customers to start design, and they're already starting designs. And if they wait for nodes after that, that delays them by quite a significant period of time. So I think it's a very viable node for FPGAs. In other industries, in particular, ASSPs, I can see why for the very high performance, maybe graphics, they would want to skip straight to the 16-nanometer node. And that would make sense for them. For us, both nodes have their benefits, and we're splitting the business across them.

I'll handle the second part of the question. So Anil, the situation we talked about some time ago is that one of our boundaries is end-of-life-ing [ph] a line. We were kind of the last man standing, and we've gotten to the point now where we cannot start any more wafers in that line. And so we've accumulated a significant amount of inventory for our customers, but this isn't what I would call our typical end-of-life for products. This is something that was accelerated by the fact the foundry was -- stopped producing products for us, so we've built a lot of inventory and are getting forecast and commitments from our customers, and we'll be delivering those products over an extended amount of time. So I'm talking about many, many quarters out in time. So you're not going to see, at least certainly anywhere in the near term, some giant pop and then a reversal of fortunes downstream. If you -- a good way to look at this is to look at our base product family. And if you look at the base products, which is where -- which is part of the base product family, is this particular set of products, and you'll see that the base products went up about $20 million quarter-on-quarter. Of that, about $10 million of that was related to an acceleration, if you will, over a normal run rate that certain customers decided to take part ahead of price increases. But the other $10 million was driven by strength in aerospace and defense business and the specific A&D program that continue to buy our legacy parts in order to ship into some of the larger programs that we've won historically, which will continue again for many years, because these are very large programs that the government is going to continue. So the real short of the answer -- to answer the question, there was some impact to that to both revenue and gross margin, but we don't expect this to be some sort of a leading, large pop that's here and then goes away in the following quarter.

Next, we have the line of Sumit Dhanda with ISI Group.

This is Ian in for Sumit. Did you guys say how much Q2 was helped by the specific government programs?

We did not say that. I did just comment on the answer to the previous question that there was about $10 million in our base product family that was incremental. I'd say, actually, compared to what we thought at the beginning of the quarter, about $10 million of business came in on a project that we weren't expecting. But there was actually another project that was in the neighborhood of the same size. So we continue to see improvement in our defense business, which I think maybe is a sign that the sequestration concerns and fears that may have been making the suppliers a little more reticent have loosened up a little bit. And we're forecasting defense to be up again in this September quarter as well. So we're seeing good contribution from the defense systems.

Next, we have the line of James Schneider with Goldman Sachs.

This is Gabriela Borges on behalf of Jim. I was hoping to follow up on your commentary on wireless CapEx. You mentioned your expectation for China, but I was hoping you could also provide some color on trends in North America and perhaps more broadly in other geographies as well?

Yes, from an overall LTE perspective, we've been doing really well with the whole North America rollout, as well as South Korea and Japan. And it -- everybody's kind of wanted to say, "Oh, where is the big increase coming out of North America?" We've had, in our wireless business for quite a few quarters, very strong contribution for the rollout of Verizon and LTE -- excuse me, Verizon and AT&T. And then as Sprint starts to kick in, I think that will be some incremental. So we've got a really nice North America business that has a little bit of lumpiness, and we continue to see that strength go on for the next several quarters at least. So that's very good. We've seen, of late, some strength out of the South Korea area. And as I mentioned earlier, we're yet to see anything significant out of the China deployment. A couple of quarters ago, we did ship into some trials for Europe, but we've not seen any significant signs of any build from a European perspective at this point in time.

[Operator Instructions] Our next question comes from Glen Yeung with Citigroup.

A question for you on your visibility, looking at your market share potential in the China business, given the design wins that you've seen today and the progress you've made with Kintex, what's your confidence that in this next deployment, whatever the size, whatever the timing, that your share position will be better than the last time?

The -- Glen, the question is related to China or generally speaking?

Yes. I'm sorry, the China LTE buildout.

Okay, okay. So we're very confident that it's going to be a lot better than it was in the previous node. The Kintex product is the winning product in terms of its cost and performance and power and was designed specifically to address the wireless market. It actually is also beneficial for others, but it's definitely a clear winner there. And with regards to wireless, in particular, we're expecting a very high percentage of the wireless LTE business in China to use Kintex. If you compare it to the 40-nanometer, I would say, it's several times in terms of percentage of what we got. And I think we, at most, had 20%-ish before.

It was -- in the early part of the 3G rollout in China, it was higher. And then it declined as designs moved to 40-nanometer. It started out in 65, and then they moved to 40 in China. And that's where we were disadvantaged with that product family. And the way that the LTE is going to roll out is that at the beginning, there will be some -- in certain supplier -- OEMs, there will be some 40-nanometer product, again, where we're disadvantaged. But they'll all be transitioning to 28-nanometer relatively rapidly over the initial few quarters, and we're very confident of our share position with the 28-nanometer win.

Next, we have the line of Ruben Roy with Mizuho Securities.

Jon, I just wanted to clarify one thing around the $10 million of base business that you talked about being incremental. Are you expecting any of that? Or is there any of that continuing into your fiscal Q2, that's implicit in your guidance or no?

So what I would characterize being for the aerospace and defense business, that part of that increase in base of $20 million -- the $10 million of the $20 million was due to some specific aerospace and defense projects, and then we do expect a strong up defense business. And some of them will be in base, some of them will be in mainstream products. So we do expect a strong, healthy defense business. The $10 million that I was characterizing as -- was more along the broad-based foundry EOL. We do expect that to continue into the next quarter. But again, we're forecasting base to be overall kind of flat to down-ish. So we aren't expecting any sort of big pops from this particular EOL that will go on for quite a long time.

Next, we have the line of Ian Ing with Lazard Capital Markets.

So revenue estimate's likely going up nicely this year. It seems most of the upside is 40- and 45-nanometer design wins. It seems 40 and 45 is up 20% this quarter, and it looks like, this year, you're still on track for $250 million in that 28. Is that sort of the right framework we should think of?

Your latter statement is correct. But no, the growth that we've experienced overall is very strong 28-nanometer. We happen to have a slightly stronger contribution from 40, 45. But I don't -- I would expect, over the fiscal year, revenue growth horizon that 28 becomes the much more rapid growth vector for us.

Okay, so north of $50 million this quarter and still on track for $250 million for the full year.

We were explicit that we were well north of $50 million, and we set a target of $60 million. And we've beaten all of our targets to date. Our target for the year is $250 million, and we're very confident about beating that target.

Okay, great. And then, Moshe, with the efforts at 16 and 20 nanometers at the same time, maybe you could articulate the benefits that the FinFETs give you? You said these are more for the expensive high-end parts, so is that -- is it sort of a logic shrink or lower power solutions, more integration?

Really, as you move from technology to technology, the benefits you get are typically lower power and higher performance, and achieving lower cost is more difficult and actually requires re-architecting your system and designing it from scratch to fully exploit the benefits of that. And some customers can do it, and some customers find that more difficult and it's a higher effort. If you compare 16 to 20, then 16 has the FinFET. And 16 will undoubtedly have even higher performance and lower power than 20. But 16 inherently is a more complex manufacturing process, and FinFETs are more difficult to design and require more manufacturing steps. Hence, they cost more. So you can achieve higher performance. You can achieve lower power. You can modulate between those 2. It's difficult for a given design to achieve cost reduction. So by definition, 16-nanometer addresses the high end of the product offering: communications -- Mil/Aero wired communications, Mil/Aero test and measurement, prototyping and less the other markets we address.

We have no further questions at this time. I turn the call back to the presenters.

Great. 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 second quarter of fiscal year '14 will be Wednesday, October 16, after the market close. This quarter, we will be presenting at the 2013 Citi Global Technology Conference in New York City on September 3. This concludes our call. Thank you very much for your interest and for your participation.

This concludes today's conference call. You may now disconnect. Thank you.