U.S. Markets closed

Edited Transcript of XLNX earnings conference call or presentation 26-Jul-17 9:00pm GMT

Thomson Reuters StreetEvents

Q1 2018 Xilinx Inc Earnings Call

SAN JOSE Aug 13, 2017 (Thomson StreetEvents) -- Edited Transcript of Xilinx Inc earnings conference call or presentation Wednesday, July 26, 2017 at 9:00:00pm GMT

TEXT version of Transcript

================================================================================

Corporate Participants

================================================================================

* Lorenzo A. Flores

Xilinx, Inc. - CFO & Senior VP

* Moshe N. Gavrielov

Xilinx, Inc. - CEO, President & Director

* Rick Muscha

================================================================================

Conference Call Participants

================================================================================

* Ambrish Srivastava

BMO Capital Markets Equity Research - MD of Semiconductor Research and Senior Research Analyst

* Blayne Peter Curtis

Barclays PLC, Research Division - Director and Senior Research Analyst

* Christopher Adam Jackson Rolland

Susquehanna Financial Group, LLLP, Research Division - Senior Analyst

* Christopher Brett Danely

Citigroup Inc, Research Division - MD

* Christopher Caso

Raymond James & Associates, Inc., Research Division - Research Analyst

* Christopher James Muse

Evercore ISI, Research Division - Senior MD, Senior Equity Research Analyst and Fundamental Research Analyst

* David Michael Wong

Wells Fargo Securities, LLC, Research Division - MD & Senior Equity Technology and Services Analyst

* JiHyung Yoo

Deutsche Bank AG, Research Division - Research Associate

* John William Pitzer

Credit Suisse AG, Research Division - MD, Global Technology Strategist and Global Technology Sector Head

* Mark John Lipacis

Jefferies LLC, Research Division - Senior Equity Research Analyst

* John Vinh

KeyBanc Capital Markets Inc., Research Division - Senior Research Analyst

* Romit Jitendra Shah

Nomura Securities Co. Ltd., Research Division - Executive Director

* Ruben Roy

MKM Partners LLC, Research Division - MD

* Srinivas Reddy Pajjuri

Macquarie Research - Senior Analyst

* Tristan Gerra

Robert W. Baird & Co. Incorporated, Research Division - MD and Senior Research Analyst

* Vijay Raghavan Rakesh

Mizuho Securities USA LLC, Research Division - MD of Americas Research & Senior Semiconductor Analyst

* Vinayak Rao

Morgan Stanley, Research Division - Research Associate

* William Shalom Stein

SunTrust Robinson Humphrey, Inc., Research Division - MD

================================================================================

Presentation

--------------------------------------------------------------------------------

Operator [1]

--------------------------------------------------------------------------------

Good afternoon. My name is Sarah, and I will be your conference operator. I would like to welcome everyone to the Xilinx First Quarter Fiscal Year 2018 Earnings Release Conference Call. (Operator Instructions)

I would now like to turn the call over to Mr. Rick Muscha. Thank you, Mr. Muscha, you may begin your conference.

--------------------------------------------------------------------------------

Rick Muscha, [2]

--------------------------------------------------------------------------------

Thank you, and good afternoon. With me are Moshe Gavrielov, CEO; and Lorenzo Flores, CFO. We'll provide a financial and business review of the June quarter, and then we'll open the call for questions.

Let me remind everyone that during our conference call today, we may make projections or other forward-looking statements regarding future events or the future financial performance of the company. We wish to caution you that such statements are predictions based on information that is currently available and actual results may differ materially.

We refer you to the documents the company files with the SEC, including our 10-Ks, 10-Qs and 8-Ks. These documents contain and identify important 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 Lorenzo.

--------------------------------------------------------------------------------

Lorenzo A. Flores, Xilinx, Inc. - CFO & Senior VP [3]

--------------------------------------------------------------------------------

Thank you, Rick, and thanks, everybody, for joining us today. Sales in June -- in the June quarter increased for the seventh consecutive quarter to $615 million, up 1% sequentially and up 7% on a year-over-year basis. Growth was driven by our advanced products, which increased 8%, sequentially to a new record. Gross margin was 68.8%, in line with our guidance. Operating expense was $243 million, also as guided. We are executing to the plan we outlined at our Analyst Day in May. Operating income for the quarter increased 4% sequentially to $180 million or 29.3% of revenue.

The tax rate was 8.2% for the quarter, below our guidance of 10% to 13%, due primarily to a discrete item related to the accounting standard for the treatment of stock-based compensation. Our net income for Q1 was $167 million or $0.63 per share.

Some key points on the balance sheet and our cash flows. We ended the quarter with $3.7 billion in gross cash and $2 billion net cash after our debt. During the quarter, the company raised $750 million of fixed rate debt at a rate of 2.95% with a 7-year term. We redeemed the remaining balance of $458 million on our convertible debt on June 15.

Operating cash flow was $191 million. During the quarter, we paid $82 million in dividends, and we repurchased approximately 1 million shares for $67 million, at an average price of $65.09. We continue to execute on our share repurchase program with the intention of exhausting our authorization over the next several quarters. We currently have $615 million remaining on that authorization.

We ended the quarter with diluted shares at 266 million, which included the impact of 14 million shares from the convertible and the warrant associated with it. Due to the redemption occurring in mid-June, the full impact on share count will be in the September quarter. For a complete explanation of the impact of these instruments on share count, please refer to our convertible FAQ on our Investor Relations website. Inventory was $215 million, down $12 million from the prior quarter.

On to guidance. In the September quarter, we are expecting sales to be between $605 million and $635 million. We continue to expect growth in our advanced products. Regards to end market, we expect the communications category to be down, driven by wireless. Industrial and A&D is expected to be up. And lastly, broadcast, consumer and automotive is expected to be up as well.

Our gross margin will be approximately 69% to 71%. Consistent with the guidance we provided at our Analyst Day, we expect operating expense to increase to approximately $253 million, including $1 million of amortization. Most of this increase will be in R&D. We continue to invest in both our technology and our customer relationships as we expand our leadership in our markets. Finally, our tax rate is expected to be between 10% and 14%.

Let me now turn the call over to Moshe.

--------------------------------------------------------------------------------

Moshe N. Gavrielov, Xilinx, Inc. - CEO, President & Director [4]

--------------------------------------------------------------------------------

Thank you, Lorenzo. I am very pleased with our ongoing execution, as manifested in our June quarter financial results. We've delivered our seventh consecutive growth quarter, with sales increasing 1% sequentially to $615 million. Operating profit increased to over 29%, reflecting the strength of our business model.

We continue to benefit from our proven technology leadership over 3 consecutive nodes and its positive implication for our broad multi-market thrust. On the market side, 5 of our 8 end markets increased in the quarter. Special mention is due to our test measurements and emulation business, which delivered record revenue yet again. In addition, our broad-based channel business was a very noteworthy area of strength. We're delighted to realize the benefits from our expanded and very focused global relationship with Avnet.

As this being the recurring pattern, overall revenue growth was driven by our advanced product. This increased by 8% sequentially and 33% year-over-year. This category now represents 52% of our overall revenue, significantly up from 49% last quarter and 42% a year ago. As advanced products now comprise over half of our business, they will continue to fuel our growth going forward.

Our 28-nanometer product portfolio had another very significant industry sales record with strong sequential growth from the ISM, [TME], broadcast and consumer market. 20-nanometer sales also significantly increased in the quarter, driven by vibrant sales in wireless, primarily for pre-5G deployment. The latest surge in the June quarter sales was driven by our -- excuse me, the largest surge in the June quarter sales was driven by our 16-nanometer portfolio. We continue to receive consistent customer feedback that our exceptional execution of this node has extended our competitive lead to over 18 months. We are now shipping 22 unique products to more than 530 discrete customers, representing all end markets, a substantial increase from last quarter.

I am delighted with the consistent progress we have made in Xilinx's transformation from the premier FPGA company to one addressing a much broader market through uniquely differentiated All Programmable devices. A great manifestation of this transition is the Zynq platform, a very deliberate multi-market expansion play into the SoC space. Our (inaudible) portfolio, which was implemented in both the 28-nanometer and 16-nanometer nodes, again delivered a new sales record, while almost doubling over the past year. In that vein, we recently began customer sampling of our breakthrough 16-nanometer RFSoC family. This highly disruptive product offers an architectural revolution of 5G wireless with integrated RF-class analog technology. It provides both 50% to 75% power and dramatic footprint reduction for future 5G deployments, cable and wireless backhaul application. I'm particularly pleased with the Amazon FPGA as a service F1 deployment that enables cloud-based acceleration. The general access phase targeted for traditional hardware application developers was very successfully launched in mid-April of this year. We're currently working shoulder to shoulder with the AWS team to facilitate the second very significant market expansion phase of this program. The expectation is that within a month, our unique SDAccel technology will be made available to AWS F1 users and will enable the much larger community of software application developers to derive the benefits of FPGA acceleration.

Returning to the here and now, our September quarter guidance of $605 million to $635 million will unquestionably be driven from broad-based growth in our advanced products portfolio. We remain focused on achieving the fiscal year 2018 revenue target of approximately $2.5 billion. In addition, while we continue to invest in our leading technology and market expansion efforts, we remain committed to delivering 30% plus operating margin exiting this fiscal year.

Let me now turn the call back to the operator for questions-and-answers.

================================================================================

Questions and Answers

--------------------------------------------------------------------------------

Operator [1]

--------------------------------------------------------------------------------

(Operator Instructions) Your first question comes from the line of C.J. Muse with Evercore.

--------------------------------------------------------------------------------

Christopher James Muse, Evercore ISI, Research Division - Senior MD, Senior Equity Research Analyst and Fundamental Research Analyst [2]

--------------------------------------------------------------------------------

I guess, first question. I know it's a small part of your business but one that we're all interested in, in the data center. Can you provide an update on what you're seeing in FPGA as a service? And I guess, as part of that, how you're ramping R&D to support that and how we should think about that

going forward?

--------------------------------------------------------------------------------

Moshe N. Gavrielov, Xilinx, Inc. - CEO, President & Director [3]

--------------------------------------------------------------------------------

Okay. Well, this program, which has been ongoing for over 1.5 years with Amazon is now online. And since mid-April, I think it was April 19, it was opened up for general availability. The nature of that is that it really targets the traditional hardware FPGA designers. And that's gotten very good reception, and we worked very closely with Amazon and their design team to enable that. But the real potential for this technology, both in terms of the ubiquitous accessibility of the technology and the broadening the set of users, is to move beyond the traditional hardware designers and facilitate -- or software developers and algorithm developers to use the technology. And we have made the technology which is an abstraction layer or think of it as a wrapper, an abstraction wrapper on top of our hardware technology now available to Amazon, and that's this SDAccel. And this is sort of a journey, it's not one set of technologies. But that is, by far, the most significant portion and we expect that to be rolled out by Amazon within a month or so. So we've developed the technology, we've handed it over to them and now we're working very closely with them to make sure the customers have the best experience possible. As with every new technology, it typically has some teething challenges. And this is very advanced technology, so it undoubtedly will have that and we expect to continue to work closely with Amazon over the next 6 months so that as it broadens its usage, it becomes, a, easier to use; b, has much higher performance and gives the benefits of the parallel processing capability, which is inherent in the FPGA to the customers so they can use it in their end applications. And everyone sort of -- I know there's tremendous desire to this -- to see this happen overnight and the reality is that this probably will take 6 months to go through this phase of the software technology deployment and acceptance by the market. And at that point in time, it will clearly have the potential of going viral and expanding in a big way. With regards to all of the numbers we've given in terms of engineering investments, they already relate to this becoming the highest priority for the company. So all of the numbers that Lorenzo and the team shared back in May in New York are -- the numbers and -- it's being done by our existing engineering resource. But it is clearly an area that we are significantly increasing the investments over what it has been in the past. You shouldn't sort of assume that, that will impact the overall engineering investment. It's an issue of prioritizing the resource.

--------------------------------------------------------------------------------

Lorenzo A. Flores, Xilinx, Inc. - CFO & Senior VP [4]

--------------------------------------------------------------------------------

And I want to expand just a little bit on one aspect of what Moshe said, C.J., which is the investment we're making is primarily in SDAccel, which is intended to be broadly available technology outside of Amazon as well, right? So this is a horizontal leveraging play as well for us.

--------------------------------------------------------------------------------

Operator [5]

--------------------------------------------------------------------------------

Your next version comes from John Pitzer with Credit Suisse.

--------------------------------------------------------------------------------

John William Pitzer, Credit Suisse AG, Research Division - MD, Global Technology Strategist and Global Technology Sector Head [6]

--------------------------------------------------------------------------------

I guess...

--------------------------------------------------------------------------------

Lorenzo A. Flores, Xilinx, Inc. - CFO & Senior VP [7]

--------------------------------------------------------------------------------

John, we're having a hard time hearing you. One second...

--------------------------------------------------------------------------------

John William Pitzer, Credit Suisse AG, Research Division - MD, Global Technology Strategist and Global Technology Sector Head [8]

--------------------------------------------------------------------------------

Is this any better, guys?

--------------------------------------------------------------------------------

Lorenzo A. Flores, Xilinx, Inc. - CFO & Senior VP [9]

--------------------------------------------------------------------------------

Yes. Now it's better. Thank you.

--------------------------------------------------------------------------------

John William Pitzer, Credit Suisse AG, Research Division - MD, Global Technology Strategist and Global Technology Sector Head [10]

--------------------------------------------------------------------------------

Yes, apologies. Moshe, I won't ask about the data center opportunity, but I was curious, in your prepared comments, you started talking about some revenue, early revenue for 5G. And so I guess I'm just trying to figure out how we should be thinking about the deployment of 5G, both the trial opportunity for you, which is going to start to come, I guess, in a little bit more volume as you exit this year and more so into 2018. And then how you see that 5G opportunity developing when trials eventually move into kind of full deployment in sort of the late 2019, 2020 opportunity? How much revenue do you think you can get out of the trial phase? And then how do you think about 5G under a full deployment, that revenue opportunity?

--------------------------------------------------------------------------------

Moshe N. Gavrielov, Xilinx, Inc. - CEO, President & Director [11]

--------------------------------------------------------------------------------

Okay. Well, we agree that 5G, in terms of real deployment, it's the 2020 onward time frame. And the expectation is the needs -- the technology needs in order to support 5G, which is a massive expansion in terms of bandwidth and reduction in terms of latency, all of those -- the hardware requirements are huge. And so the expectation is that this will be a long rollout, but the overall area under the curve will be larger than it has been in previous generations of technology. And we're seeing now a massive race to use our products and in particular, the RFSoC, we believe, is a very attractive element, which should enable us to capture very significant parts of that early deployment. And then over time, what we would expect is when you move into the later phases, that will be addressed by a 7-nanometer product offering. And so we expect both of those to be addressed by our technology and, in particular, by this integration of the mix signal. When you look at the overall market, we expect wireless to sort of generally, until 2020, to move sideways. But not at the trough level; somewhere in the middle. And we're sort of seeing this. If you look at our trough and peak levels of wireless, they were -- there's about a 2:1 ratio between the 2, and it's now hovering in the middle. And in quarters like the quarter we just reported, actually, wireless went up substantially and was well above that midpoint. And in the current quarter, we can tell it's going to go down a little, just below that midpoint. And we can't sort of -- that's the nature of this business, we expect it to continue to fluctuate. But for us, we expect it to be on average at that level, which is better than most and that's due to the very strong technology position we have. And that's probably why you're seeing us do somewhat better than some of the other traditional players in this market. But to summarize, we expect it to sort of be, on average, moving sideways until 2020. And from 2020, we expect it to sort of get to the previous peak levels. And then when there will be rapid deployment, it could be above the peak level, but it won't sort of stay there the whole time. The area under the curve is very substantial, and it's just the technology requirements are much larger and the density of the MIMO and all of that is really much higher than it has been in the previous generation.

--------------------------------------------------------------------------------

Operator [12]

--------------------------------------------------------------------------------

Your next question comes from David Wong with Wells Fargo.

--------------------------------------------------------------------------------

David Michael Wong, Wells Fargo Securities, LLC, Research Division - MD & Senior Equity Technology and Services Analyst [13]

--------------------------------------------------------------------------------

Can you give us some idea of what percentage of your total revenues automotive made up? And what sort of growth you're seeing there sequentially and year-over-year?

--------------------------------------------------------------------------------

Lorenzo A. Flores, Xilinx, Inc. - CFO & Senior VP [14]

--------------------------------------------------------------------------------

Yes, so our automotive businesses is around 7% of our overall business. And it's been driven -- the strength of our automotive business has been driven primarily by the ADAS applications over the past few years as they've offset the infotainment legacy business we've had. So that will be the driver for us in the future. Let me just see if I have some sort of year-over-year. You think -- our automotive business over the past few years has grown from less than 5%, in the order of 4% of our business to where it is today. So it's been one of the fastest growers in our overall suite of end markets, David.

--------------------------------------------------------------------------------

Operator [15]

--------------------------------------------------------------------------------

And your next question comes from Joseph Moore with Morgan Stanley.

--------------------------------------------------------------------------------

Vinayak Rao, Morgan Stanley, Research Division - Research Associate [16]

--------------------------------------------------------------------------------

This is Vinayak calling in for Joe. I had a follow-up on your ADAS business. Can you just touch upon the breadth of ADAS applications where you're seeing traction as (inaudible) the diversity of the design wins you have there? And longer term, how do you see your positioning in a fully autonomous world?

--------------------------------------------------------------------------------

Moshe N. Gavrielov, Xilinx, Inc. - CEO, President & Director [17]

--------------------------------------------------------------------------------

So this is a market we have invested in for several years now, and both -- and what you're currently seeing in terms of our revenue is largely our 28-nanometer Zynq product offering. And we're seeing it -- the numbers we had shared is approaching 100 models in just over 25 manufacturers. And we're seeing that continue and we're seeing the usages of technology, it's sort of forward-looking, backward looking, inward looking, sideways looking. There's a lot of applications where it's being used. What we're seeing now is that a lot of the early designs are now transitioning to the second generation of product, which is the MPSoC, it's the 16-nanometer product offering. What we expect to happen and obviously, there's a lot of press on this, but these are still really very, very early days. And when you sort of hear about vehicles driving around, they don't really have the level 5 or anything close to that available. And the expectation is that over the next several years that, that will roll out. And we expect this to be a combination of potentially a very powerful central CPU or CPU cluster, and then a whole host of other inputs. We are targeting that using our 7-nanometer as the ultimate solution for this market, with more of a focus on the other elements as opposed to the supercomputer at the center of it all. And our expectation is that the market will evolve in a way that because things are unknown, it will take probably until 2030 before it's clear how it's evolving. And we expect to be a significant player addressing that and building on our ADAS leverage. And so if you sort of look at the ADAS and if you look at our 16-nanometer solution, we already have tremendous investment in things like security, which are essential to this, where we're way ahead of everyone else in terms of providing the best security for these solutions. And I think you're probably aware that this is one of the biggest concerns in this market going forward, because if the cars can be hijacked, that's not a particularly comforting thought, right? And so we're in the midst of it all, we're continuing to invest. The current business is driven by 28-nanometer. It's now moving. And over the next 5 years, there will be significant deployment, and we already know the model. In 16-nanometer, the MPSoC solutions, where we have a larger role. And then we're working on the 7-nanometer, which will enable us to continue to stay in the game going forward. So hopefully, that answered your question.

--------------------------------------------------------------------------------

Operator [18]

--------------------------------------------------------------------------------

Your next question comes from John Vinh with KeyBanc Capital Markets.

--------------------------------------------------------------------------------

John Vinh, KeyBanc Capital Markets Inc., Research Division - Senior Research Analyst [19]

--------------------------------------------------------------------------------

Moshe, my question for you is on competition. I was wondering if I could get your updated thoughts on how you're thinking about competition these days. Obviously, since your main competitor has been acquired, it's been no secret that they really struggled to execute. You guys clearly have dominated the last 3 nodes and even moving forward. It doesn't look like you're going to get a lot of pushback. And I'm wondering, given your success that you're getting in Zynq and embedded SoC, are you starting to see your competition becoming more of the ASICs and ASSP players? Or are you still seeing your competition as traditional FPGA? And if your competition is starting to shift a little bit, are there any sort of changes that you need to make to address that?

--------------------------------------------------------------------------------

Moshe N. Gavrielov, Xilinx, Inc. - CEO, President & Director [20]

--------------------------------------------------------------------------------

Okay. Well, that's a great question. And indeed, you're right. And actually, this is an ongoing transition which is happening. And as our devices are becoming more and more sophisticated, they're all in the systems exchanging. And whereas in the past, the programmable logic for the FPGAs used to have a relatively well-defined role, but it was more of a auxiliary capability. Now these All Programmable have very sophisticated SoCs with highly integrated programmable logic moving to the center of the customer's system. And now because we're using this technology and it's happening in conjunction with the transition to heterogeneous computing, right, which is -- people are moving beyond the traditional computing models to others where there's a need for parallelism, we have a very good match with a lot of these new applications. And in particular, if you look at machine learning and artificial intelligence, that's sort of the killer application which transcends all of these and actually can be very well addressed by our solution. As a result of that, we definitely are seeing very different competition. I would say that the market is about 60% of the -- in terms of revenue is in our more traditional historic competition. But there are -- actually, there's less players in that market, and I'm not only talking about Altera, which is now part of Intel, but you're seeing less and less competition in those markets like A&D, Broadcast and the like, they really are seeing less competition because they're very fragmented. And test and measurement is another example of those. And then about 40%, what we're seeing and this is very market specific, we're seeing different competitors. So if you look at wired, the biggest competitor would be Broadcom and in some case high-end ASICs. Now if you look at wireless, it's more ASICs than anything else. I mean, the solution for that is the high level of integration with our RFSoC. If you look at things like automotive and data center, then what you see is the big semiconductor companies there, and they are addressing it through their SoC-based solutions. And the way we're attacking it is not frontally, it's based on our strengths because we believe that the ability to use the programmable logic as a differentiator is a key element, which is very difficult to duplicate any other way. And so we're addressing those markets there. But it is a very different and competitive map than it was. It's also one which requires us to invest a lot in ease-of-use, and that's the biggest knock on our technology in the past was it would only work for hardware designers. And a lot of these companies don't even have hardware designers or have very small teams, and that's why the SDAccel and SDSoC products were so important. And those are the things that Lorenzo highlighted. The Amazon deployment of this technology not only will benefit our business with the hyperscalers but actually will enable us to compete on a broader base in all of these markets against other SoC providers.

--------------------------------------------------------------------------------

Operator [21]

--------------------------------------------------------------------------------

Your next question comes from William Stein with SunTrust.

--------------------------------------------------------------------------------

William Shalom Stein, SunTrust Robinson Humphrey, Inc., Research Division - MD [22]

--------------------------------------------------------------------------------

Moshe, congrats on the good quarter, and outlook in particular. On the quarter you just posted, the Broadcast, Consumer & Auto end market did very well. I wonder if you can maybe dig into the components there. You gave us a little bit about auto, but can you talk about the other 2 subsegment or sub-end markets there?

--------------------------------------------------------------------------------

Lorenzo A. Flores, Xilinx, Inc. - CFO & Senior VP [23]

--------------------------------------------------------------------------------

Yes. Both the other 2 elements of those end markets actually did a little bit better than expected. They're still relatively small and as everybody knows, consumer isn't a focus area of ours, but we're seeing some interesting opportunities develop in certain end market niches there that we're particularly suited for. And so we'll see how that evolves over time. The AVB strength is really, I think, a cycle, product cycle where, again, our products are particularly well suited for the evolution of that industry. And so we've seen success there. And we probably will continue our strength there. As Moshe said, we're seeing less intense competition from a traditional sense there. I will point out, auto was coming off a pretty spectacular record quarter for us last quarter. And so where it is today, we see it continuing to strengthen through the rest of the year. So again, based on the designs that we have on our 28-nanometer Zynq product line, in particular. Does that get to your question, Will? Okay.

--------------------------------------------------------------------------------

Operator [24]

--------------------------------------------------------------------------------

Your next question comes from Ross Seymore with Deutsche Bank.

--------------------------------------------------------------------------------

JiHyung Yoo, Deutsche Bank AG, Research Division - Research Associate [25]

--------------------------------------------------------------------------------

This is JiHyung for Ross. The growth expected in the Industrial, Aero & Defense segment in the September quarter was actually a little bit better than on prior expectations for a flat to slight decline. Can you discuss some of the subsegments within the IAD segment in the September quarter? And does this imply that fiscal year '18 growth can be above the prior guidance for 5% to 9% growth year-on-year?

--------------------------------------------------------------------------------

Lorenzo A. Flores, Xilinx, Inc. - CFO & Senior VP [26]

--------------------------------------------------------------------------------

So let me take the bigger question first, which is as we go through quarter-to-quarter and as each of our end markets trends through the year, there'll be pluses and minuses throughout the year. So I wouldn't read next quarter's guide as anything on the full year yet. I mean, ideally, all the positives continue, but that's just not the reality. Within the -- and each of the end markets for Q2 in the Industrial and A&D space, we actually expect each of them to be growing different degrees. So we see strength across the board.

--------------------------------------------------------------------------------

Operator [27]

--------------------------------------------------------------------------------

Okay, your next question comes from Ambrish Srivastava with BMO Capital Markets.

--------------------------------------------------------------------------------

Ambrish Srivastava, BMO Capital Markets Equity Research - MD of Semiconductor Research and Senior Research Analyst [28]

--------------------------------------------------------------------------------

Moshe, I just wanted to clarify something, and I'm not expecting your AI-related business to grow overnight, but I just wanted to revisit something you said a few quarters ago, and are we still on track for that. You had said that by -- exiting fiscal '18, it could be tens of millions of dollars. And so question, are we on track? And would you need to get additional wins beside the AWS? And also if you could please highlight the -- any details you could give us on the Baidu announcement that came out a few weeks ago.

--------------------------------------------------------------------------------

Moshe N. Gavrielov, Xilinx, Inc. - CEO, President & Director [29]

--------------------------------------------------------------------------------

So now there are 7 hyperscalers and then there's also 2 significant systems companies that are very close customers and partners. You can look at our release and you can sort of figure out the people who are in CCIX. CCIX is the protocol which addresses and is targeted at -- particularly, this market addresses this area. So as you look at those, the tens and millions of dollars do not require massive adoption by additional players. It's sort of -- it's based on what we believe and for better or for worse, this is the way it's panning out. This is based on the existing business we have with numerous players, which is going to be relatively small. The business with AWS, which I think will carry the bulk of this and then any of the other players that we're working with, all of the other 6 hyperscalers and these 2 system-level companies who are pursuing these businesses, which is to be cloud-based providers, they're likely to contribute, but not significantly this year, right? And that's -- and as they come on and as the AWS potential deployment increases significantly then that's what sort of enables us to grow beyond these numbers, but there's nothing in those tens of millions of dollars which sort of assumes huge design wins translating to revenue other than the ones that you're sort of aware of. And the Baidu one generally belongs in that other category. It's not -- we're not counting on it yet for this, for our fiscal year '18 numbers.

--------------------------------------------------------------------------------

Operator [30]

--------------------------------------------------------------------------------

Your next question comes from Blayne Curtis with Barclays.

--------------------------------------------------------------------------------

Blayne Peter Curtis, Barclays PLC, Research Division - Director and Senior Research Analyst [31]

--------------------------------------------------------------------------------

Actually, if I could just follow up on what you just said, Moshe, on the data center. One, couldn't you just level-set just how big that segment is in the June quarter? And then to your comment about the projects that you have, I'm just kind of curious if you can lineate between FPGA as a service with sort of similar offerings to Amazon versus your more direct deployments for internal use, such as what Microsoft is doing.

--------------------------------------------------------------------------------

Moshe N. Gavrielov, Xilinx, Inc. - CEO, President & Director [32]

--------------------------------------------------------------------------------

So those numbers are primarily driven by the Amazon, which is FPGA as a service. Now FPGA as a service is a lot more difficult to do than an internal deployment which has a very limited usage model. And I'm not implying that that's easy either, but that's a contained challenge. Our expectation is that the -- when the benefits of FPGA as a service, which is an incredibly aggressive move by Amazon, when those start to manifest themselves, they could, in and of themselves, lead some of these other players once they see the benefits of that to integrate All Programmable devices into their solutions. And whether they do it as a service or whether they do it for -- to enable their existing business, both would benefit from that. But you're right, in the interim, it's the -- FPGA as a service is the major driver short term. Now we are working with all of these players on numerous options. And as you have numerous times pointed out there, there's 4 categories of ways to skin the cat. And we believe that our way and the benefits of it are going to be clearer and more obvious over time. And we do expect to see the potential of that to being adopted in some of these other players for their core capability as opposed to just as a service solution. Hopefully, that answered your question, Blayne?

--------------------------------------------------------------------------------

Operator [33]

--------------------------------------------------------------------------------

Okay. We have a follow-up question from Ambrish Srivastava with BMO.

--------------------------------------------------------------------------------

Moshe N. Gavrielov, Xilinx, Inc. - CEO, President & Director [34]

--------------------------------------------------------------------------------

Ambrish Srivastava.

--------------------------------------------------------------------------------

Ambrish Srivastava, BMO Capital Markets Equity Research - MD of Semiconductor Research and Senior Research Analyst [35]

--------------------------------------------------------------------------------

I had a quick follow-up for Lorenzo. On the turns, seems like on a -- if you just look on a yearly basis, we've gone from 50 to 42. Is there a structural change in the business? We -- I'll just say "we grew up thinking" PLDs, FPGAs were very hard turns business, low visibility. Is this a structural change or reflective of the current condition we are in, i.e. visibility has extended maybe artificially because of some tightness in the supply chain?

--------------------------------------------------------------------------------

Lorenzo A. Flores, Xilinx, Inc. - CFO & Senior VP [36]

--------------------------------------------------------------------------------

So there's a couple of different things driving this. And I do think if you look back in just over the past year, you'll see a fairly big range in terms of the turns we've seen each quarter. A couple of things that we -- we are trying to get more orders in the book and make our quarters more predictable. Sometimes, it just doesn't happen. The customers don't make decisions based on our quarters. So for the previous 2 quarters, we had a relatively high book. In the 2 quarters prior to that in FY '17, we had a relatively low book. And in each of those cases, we managed to come pretty close to our guidance because we go through a couple layers of effort to understand what else is out there, what's likely to turn, where our consumers are at in their decision process. So we've seen a range of turns. I think we evaluate that internally. We are trying to get the bookings earlier in the quarter. It helps us with our predictability and it helps actually with our accounts receivable at the end of the quarter. But we can't always get what we want in working with our customers, but we do, do the rigor in understanding where the business is -- or where we think the business is going to come from at any given quarter. So I wouldn't say there's a structural change. I think we are working at, obviously, the whole optimization of our supply chain through time, but there will still be volatility.

--------------------------------------------------------------------------------

Operator [37]

--------------------------------------------------------------------------------

Next question comes from Tristan Gerra with Robert W. Baird.

--------------------------------------------------------------------------------

Tristan Gerra, Robert W. Baird & Co. Incorporated, Research Division - MD and Senior Research Analyst [38]

--------------------------------------------------------------------------------

You mentioned that you're going to be working with AWS over the next 6 months to optimize ease-of-use and performance. Is this ramp purely contingent on the customer ramp time line at this point? Or is there any update to software, anything implementation related on your side that needs to be finalized first?

--------------------------------------------------------------------------------

Moshe N. Gavrielov, Xilinx, Inc. - CEO, President & Director [39]

--------------------------------------------------------------------------------

Well, it's a joint effort. The technology we have provided that to Amazon and being packaged by Amazon in a way where have their own delivers. We continue to -- to the extent that they need more help from us, we're very motivated to help them, and it is a top priority. But I would say that in terms of anything profound from us, we've already delivered all of that. The big invention has been done, and now it's being handed over and needs to be product-ized in a way where Amazon feels comfortable with deploying it broadly. And of course, we'll provide whatever support they need, but we don't know of anything which is a fundamental problem at this point in time.

--------------------------------------------------------------------------------

Operator [40]

--------------------------------------------------------------------------------

Your next question comes from Mark Lipacis with Jefferies.

--------------------------------------------------------------------------------

Mark John Lipacis, Jefferies LLC, Research Division - Senior Equity Research Analyst [41]

--------------------------------------------------------------------------------

I have a couple of related questions, I hope you don't mind. On AWS, and the first part is a clarification, to what extent is this based -- the interface is going to be solely SDAccel versus like a Vivado interface? That's the first one. The second one is, can you talk about what you view as the competitive environment with SDAccel? Are you picking up that your competitor has anything like this? And the third one, Moshe, I believe last earnings call, you mentioned that you were not quite -- it was a bit of a surprise -- correct me if I'm wrong, but it was a bit of a surprise that you learned kind of late that AWS was working on theirs. So I was wondering if you had a sense if any of the other Super 7 guys are kind of engaged on this FPGA as a service also?

--------------------------------------------------------------------------------

Moshe N. Gavrielov, Xilinx, Inc. - CEO, President & Director [42]

--------------------------------------------------------------------------------

Okay. So 3 questions is normally 1 more than I can remember, so if I -- I might ask you to step in. But the initial deployment, they started the deployments at the beginning of the year and that was an internal deployment which was just accessible to their hardware design team. The [GA], which was on April 19, was a Vivado-oriented deployment. And that again is targeted at customers who have traditional hardware design -- designers. And the new one, which we expect to be rolled out over the next month or so, is the one which has SDAccel. So but you are right, the first one was Vivado. And now the new one will be SDAccel, targeting software developers. With regards to the 9 players who are targeting these markets, most of them are not looking at FPGA as a service, and they're actually looking more at providing acceleration capabilities as part of their cloud-based footprint but -- to enable their business as opposed to making a business of it. But having said that, there are quite a few who are looking at making a business of it, and I'm not at liberty of saying who is and who isn't. But obviously, Amazon is a tremendous homing beacon for quite a few of these. And some of them have announced capabilities, but no one is anywhere close to where Amazon is in terms of the level of deployment. And then the middle question.

--------------------------------------------------------------------------------

Lorenzo A. Flores, Xilinx, Inc. - CFO & Senior VP [43]

--------------------------------------------------------------------------------

Does the competitor have anything like SDAccel at all?

--------------------------------------------------------------------------------

Moshe N. Gavrielov, Xilinx, Inc. - CEO, President & Director [44]

--------------------------------------------------------------------------------

Does the competitor have anything like SDAccel? I'm sure they're convinced they do. But truth is that this is technology which, at the core, is some unique proprietary synthesis technology, which we believe we're the only people in the world that has that. So we believe that this sort of approach is maybe not limited to us, but we have, by far, the best, the highest likelihood of being successful there. That doesn't -- to be totally honest, that doesn't eliminate them as a competitor because they can try to address it in other ways which are maybe more limited and maybe don't give the same benefits. But they can sort of try to do that. And I'm sure that, that -- to the extent that they're interested in pursuing it, that's the most likely way that they will. In which case, we will have, likely, a very strong competitive advantage on that, too. And so that's the third of the 3.

--------------------------------------------------------------------------------

Operator [45]

--------------------------------------------------------------------------------

Your next question comes from Romit Shah with Nomura.

--------------------------------------------------------------------------------

Romit Jitendra Shah, Nomura Securities Co. Ltd., Research Division - Executive Director [46]

--------------------------------------------------------------------------------

I just wanted to make sure I understood the math to getting to 30% operating margin by the end of the fiscal year. So for the current quarter OpEx is, per guidance, $253 million, which is a little higher than what we were forecasting, but you're also forecasting gross margins to be better, too, I think, at 70%. So I guess, should we assume that you hold OpEx here at the current level, in the $250 million-plus range, and I guess on slightly higher quarterly revenue; that's how you get to 30%, or is there other dynamics?

--------------------------------------------------------------------------------

Lorenzo A. Flores, Xilinx, Inc. - CFO & Senior VP [47]

--------------------------------------------------------------------------------

So what I tried to describe at our Analyst Day, Romit, was some of the general trends over the year that we were expecting. And on OpEx, we do expect a step, as we've said, this quarter. And then a little step next quarter from tape-out expenses and then have a leveling off, a decline. And concurrent with that, as you have anticipated, a revenue growth throughout the year. That would give us the result that we've been talking about.

--------------------------------------------------------------------------------

Operator [48]

--------------------------------------------------------------------------------

Your next question comes from Vijay Rakesh with Mizuho.

--------------------------------------------------------------------------------

Vijay Raghavan Rakesh, Mizuho Securities USA LLC, Research Division - MD of Americas Research & Senior Semiconductor Analyst [49]

--------------------------------------------------------------------------------

Just taking a step back, I'm looking at the deep learning and AI on the data center side. If you look at the Comm & Data Center segment, are you still sticking with the 1% to 4% year-on-year growth, given the trends in the wireless and the data center side for fiscal '18, that is?

--------------------------------------------------------------------------------

Moshe N. Gavrielov, Xilinx, Inc. - CEO, President & Director [50]

--------------------------------------------------------------------------------

Yes, we don't see a reason to change those ranges. For all of the 3 components we provided 2 months ago, we believe that they'll end up in those ranges. Now they probably won't all end up in the middle of the range, some will be higher, some will be lower. But we don't see a reason to modify the range at this point.

--------------------------------------------------------------------------------

Operator [51]

--------------------------------------------------------------------------------

Your next question comes from Christopher Rolland with Susquehanna.

--------------------------------------------------------------------------------

Christopher Adam Jackson Rolland, Susquehanna Financial Group, LLLP, Research Division - Senior Analyst [52]

--------------------------------------------------------------------------------

For Aerospace, it seems like that's being helped by strong passenger growth year-on-year right now. Do you see kind of a reacceleration in that business? And similarly, for Defense, there are -- some believe there are kind of multiple new global defense cycles beginning. Do you agree with that? Or do you just kind of see this as steady as she goes?

--------------------------------------------------------------------------------

Moshe N. Gavrielov, Xilinx, Inc. - CEO, President & Director [53]

--------------------------------------------------------------------------------

Well, we -- I would say, we are less dependent on passenger patterns, primarily because -- well, there's a few reasons but, a, it's much broader than planes. We play in a lot of areas and even when it gets to planes, they don't immediately change the number of planes they are building. That takes years to sort of put the infrastructure in place. So what we are seeing is the latter element that you talked about, which is seeing across the board -- not only in North America but across the world -- there's a bigger investment in aerospace and defense, and it appears to be just going in that direction and the budgets are growing. And as you know, there's sort of pressure from the U.S. administration on the Europeans to pay a larger portion of the NATO bill, and one of the ways of doing that is by buying more equipment. And I think we're seeing that across the board. And it's difficult for us to predict which way it's going to go, but it doesn't appear to be going away by any stretch.

--------------------------------------------------------------------------------

Operator [54]

--------------------------------------------------------------------------------

Your next question comes from Chris Caso with Raymond James.

--------------------------------------------------------------------------------

Christopher Caso, Raymond James & Associates, Inc., Research Division - Research Analyst [55]

--------------------------------------------------------------------------------

Just a question on gross margins, which the guidance is a bit higher than what you indicated in the Analyst Day and what you've shown in the past couple of quarters. Is that mix related? If you could provide some explanation of that and what we should expect going forward. Is what you talked about at the Analyst Day still the right way to think about gross margin going forward?

--------------------------------------------------------------------------------

Lorenzo A. Flores, Xilinx, Inc. - CFO & Senior VP [56]

--------------------------------------------------------------------------------

Yes, again, I'll start with the end and then come back to this current quarter. We're still comfortable with the range that we provided for the year, which is 68% to 70%. We're not changing that as of now. For Q2, what you'll see, what we've talked about already is relative weakness in wireless and relative strength in the Industrial and A&D segment for us, which that mix shift tends to help our gross margins.

--------------------------------------------------------------------------------

Operator [57]

--------------------------------------------------------------------------------

Your next question comes from Chris Danely with Citi.

--------------------------------------------------------------------------------

Christopher Brett Danely, Citigroup Inc, Research Division - MD [58]

--------------------------------------------------------------------------------

As I'm near the end, hopefully I can trade in 1 question for 2 quick clarifications. Lorenzo, I think you said that R&D or OpEx will be trending down towards the end of the fiscal year; is that true? And should we assume then that OpEx will be down on a year-over-year basis in fiscal '19? And then are there any more discrete tax items coming for the rest of the year that you anticipate?

--------------------------------------------------------------------------------

Lorenzo A. Flores, Xilinx, Inc. - CFO & Senior VP [59]

--------------------------------------------------------------------------------

So again, I'm going to start in a different direction. We're not going to give FY '19 guidance yet. And to clarify on what we said on OpEx for the year, we're showing a step-up this current quarter, primarily driven by R&D. We'll see a -- likely see a step-up in the next 2 quarters. Tape-outs will be a big driver of that, but we're also continuing to invest in supporting the business. And then we will see -- we're likely to see a decline in the fourth quarter as, again, driven by tape-out expenses not happening, we'll get through a bulge of them by the third quarter. And like I said, we'll talk about FY '19 sometime later. With respect to discrete items, one of the drivers that I pointed out in my remarks was how we were accounting for excess stock-based compensation. And that's driven by a difference in stock price, current stock price versus historic stock price, really. So we'll see some of that going forward. We haven't sized that yet and we can't size that yet until we actually go through the quarter and see what the stock price does.

--------------------------------------------------------------------------------

Operator [60]

--------------------------------------------------------------------------------

Your next question comes from Srini Pajjuri with Macquarie Research.

--------------------------------------------------------------------------------

Srinivas Reddy Pajjuri, Macquarie Research - Senior Analyst [61]

--------------------------------------------------------------------------------

Just a question, a clarification on the wireless segment; you said it's a little bit weaker in the short term. Could you give us a bit more color by, I guess, China and India, what you're seeing out there? And then you mentioned that 5G won't happen for a couple of years. And as we bounce around these levels for the next couple of years, what's the risk that you could see some ASIC conversions in the existing 4G business, given the maturity of that business?

--------------------------------------------------------------------------------

Moshe N. Gavrielov, Xilinx, Inc. - CEO, President & Director [62]

--------------------------------------------------------------------------------

So let me start with the ASIC conversions. To the extent that they're viable, they already have been done on 4G, right, that's sort of been in production for several years. And I would say, you never take your eye off that, but I think most of the important ones would -- with the people with that capacity, probably have already happened. So I don't think there's a huge risk on 4G there. What I do think is going to happen is there's -- this past quarter, the June quarter, there was a surge, a very significant surge on the wireless business. And the current quarter, the September quarter, we expect it to go down quite a bit. The -- and we -- I expect this sort of volatility to continue over the next few years until 2020. But I don't think that the average business is likely to go away because there are ongoing deployments and there is densification and a whole host of things which we tend to see. And this more global deployment still ahead of us there. So on average, we expect it to remain this way. And yes, indeed, India, which was growing very well is taking a hiatus and China is currently not in a huge surge in terms of deployment, but we do think that some of these areas will start the pre-5G deployment and that should help fill things in, in addition to additional deployment of the older protocols.

--------------------------------------------------------------------------------

Operator [63]

--------------------------------------------------------------------------------

Your next question comes from Ruben Roy with MKM Partners.

--------------------------------------------------------------------------------

Ruben Roy, MKM Partners LLC, Research Division - MD [64]

--------------------------------------------------------------------------------

Moshe, you mentioned something when talking about AWS that was interesting to me. You were saying that you're working arm and arm or shoulder to shoulder, I guess, with the engineers there. And as you spend more time on that specific program, are you -- I'm wondering if the resources that Xilinx will need to support that type of customer and that type of program are different from traditional FPGA programs and applications. Have you gotten any sense that, that might happen? And then longer term as a service is actually being used outside of AWS by end users, does that change kind of your sales process or your resources required to support? And I think that's it.

--------------------------------------------------------------------------------

Moshe N. Gavrielov, Xilinx, Inc. - CEO, President & Director [65]

--------------------------------------------------------------------------------

Yes, so generally speaking, I 100% agree with your observation. The challenge is how to do it and how to scale it. And one of the benefits we have is a model which can scale to thousands of customers very effectively. And so what we're doing is this is being done through a combination of elements. If for every new deployment we need to add additional resource, it would not be particularly attractive. Clearly, the level of investment on the Amazon side is a huge one, but that is a singular point. It's about as complex as it can get. And that's why we're so excited because we figure out if we can make them successful, then this technology and the SDAccel and there's a whole host of other things that have been put in place, are going to be applicable to the breadth of our customers and should enable us to support the business and continue to leverage the resources we have. And so the answer is it's a combination of things, it's a transition we've been going through for several years. Amazon is the ultimate in terms of what they're trying to achieve. It's a very different support infrastructure, it's one which is built through collaborative interaction on the web, and we're figuring out how to support that. But once we have that in place, it should scale for our other customers. And I don't think -- best we can tell, it doesn't change our business model, but we need to constantly make sure that we address this to facilitate that more efficient support infrastructure.

--------------------------------------------------------------------------------

Lorenzo A. Flores, Xilinx, Inc. - CFO & Senior VP [66]

--------------------------------------------------------------------------------

Okay. Thanks for joining us today. We'll have a playback of this call beginning at 5:00 p.m. Pacific Time, 8:00 p.m. Eastern Time today. For a copy of our earnings release, please visit our Investor Relations website. Our next earnings release date for the second quarter fiscal year 2018 will be Wednesday, October 25, after the market closed. We'll be attending the following conferences this quarter: the Pacific Crest Global Technology Forum on August 7; the Jefferies Semiconductor, Hardware and Communications Infrastructure Summit in Chicago on August 29; and the Citi Global Technology Conference in New York City on September 7.

This completes our call. Thank you very much for your participation.