Rick Muscha - Xilinx, Inc. Lorenzo A. Flores - Xilinx, Inc. Moshe N. Gavrielov - Xilinx, Inc..

Ambrish Srivastava - BMO Capital Markets (United States) John William Pitzer - Credit Suisse Securities (USA) LLC Tristan Gerra - Robert W. Baird & Co., Inc. Joseph L. Moore - Morgan Stanley & Co. LLC C.J. Muse - Evercore Group LLC John N. Vinh - KeyBanc Capital Markets, Inc. William Stein - SunTrust Robinson Humphrey, Inc. David M.

Wong - Wells Fargo Securities LLC Blayne Curtis - Barclays Capital, Inc. Ross C. Seymore - Deutsche Bank Securities, Inc. Hans C. Mosesmann - Rosenblatt Securities, Inc. Christopher Brett Danely - Citigroup Global Markets, Inc..

Good afternoon. My name is Ian and I'll be your conference operator. I would like to welcome everyone to the Xilinx Fourth Quarter Fiscal Year 2017 Earnings Release Conference Call. Please limit your questions to one to ensure that management has adequate time to speak to everyone. I would now like to turn the call over to Mr. Rick Muscha. Thank you. Mr.

Muscha, you may begin your conference..

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

Let me remind everyone that during our conference call today, we may make projections or other forward-looking statements regarding future events or the future financial performance of the company.

We wish to caution you that such statements are predictions based on information that is currently available and that actual results may differ materially. We refer you to 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..

Thank you, Rick. Sales in the March quarter increased for the sixth consecutive quarter to $609 million, up 4% sequentially and up 7% on a year-over-year basis. Growth was driven by our Advanced Products, which increased 9% sequentially to a new record. Gross margin was 69.5%, at the high end of our guidance, due primarily to favorable end market mix.

Operating expense was $250 million. This was $6 million higher than guided as we accelerated some 16-nanometer tape-out expenses to extend our technology leadership and add some increased litigation expense. Operating income for the quarter increased 6% sequentially to $173 million or 28.5%.

Other income and expense was an expense of $2.2 million, better than guided due primarily to investment gains. Tax rate was 10% for the quarter due to discrete item. Our net income for Q4 was $153 million or $0.57 per share. We are pleased to have delivered on our financial plan in fiscal 2017.

We met our target of 6% revenue growth, driven by a 45% increase in Advanced Products, and operating margin was 30% for the year. This profitability led to the generation of a record $934 million in operating cash flow. Finally, our EPS was $2.32 for the year, a 13% increase over FY 2016. Now, some key points on the balance sheet and cash flows.

We ended the quarter with $3.4 billion in gross cash and $2 billion in net cash after our debt. Accounts receivables decreased by nearly $100 million, as we collected last quarter's higher than normal receivables balance. Inventory was $227 million, up $21 million from the prior quarter, with nearly all the increase coming from our Advanced Products.

Operating cash flow was $306 million for the quarter. In the quarter, we paid $82 million in dividends, and we repurchased 1.8 million shares for $108 million, an average price of $58.45. We ended the quarter with diluted shares at 267 million, which included the impact of 15 million shares from the convertible and the warrant associated with it.

For a complete explanation of the impact of these instruments on share count, please refer to our convertible FAQ on our Investor Relations website. As we have discussed, capital allocation remains a top priority for the company.

This year, we returned $855 million to shareholders through $333 million of dividends and $522 million of share repurchases. This total capital return is $90 million more than we returned to shareholders in the prior year. Our board recently authorized an increase to our dividend for the 12th consecutive year.

We continue to execute on our share repurchase program with the intention of exhausting our $1 billion authorization over the next several quarters. We currently have $680 million left on that authorization. Now to guidance. In the June quarter, we are expecting sales to be between $600 million and $630 million.

Our backlog is up heading into the quarter, and we are expecting continued growth in our Advanced Products. On end markets, we expect the communications category to be up, industrial and A&D categories is expected to be flat sequentially. And, lastly, broadcast, consumer and automotive is expected to be slightly down.

Our gross margin will be approximately 68% to 70%. We expect operating expense to decline to approximately $242 million, including $1 million of amortization. Other income will be $1 million. Finally, our tax rate is expected to be between 12% and 15%. We will provide full year FY 2018 guidance at our upcoming Analyst Meeting in New York City on May 22.

We look forward to seeing you all at that meeting. Let me now turn the call over to Moshe..

Thank you, Lorenzo. I am extremely pleased with our March quarter financial results. Sales increased 4% sequentially to $609 million, marking the sixth consecutive growth quarter. Gross margin was at the high end of our range. The combination of higher sales and higher gross margin drove our strong operating profit growth.

We continue to benefit from our diversified multimarket portfolio. Five of our eight end markets increased in the quarter. The automotive and test measurement and emulation markets drove the largest incremental sales increases, and both delivered new record revenue levels.

Overall, revenue growth continues to be driven by our Advanced Products, which increased 9% on a sequential basis and 45% overall on an annual basis. This category now comprises 49% overall of our revenue. 28-nanometer product family very significantly surpassed $200 million in the quarter, setting by far a very significant PLD industry record.

20-nanometer generated $60 million in sales, driven by a very broad base of markets. 16-nanometer sales grew significantly in the March quarter to a new record, significantly exceeding our forecast with sales from all of our end markets.

Our Zynq SoC platform, which includes both our 28-nanometer and 16-nanometer product offerings, increased nearly 20% sequentially and now represents more than 10% of our overall revenue. One of our key 16-nanometer customers is Amazon Web Services. Just last week, AWS announced the general availability of our FPGAs for cloud-based acceleration.

Our world-class silicon technology, coupled with our optimized software tools, allow AWS to offer optimizable and programmable hardware acceleration to their users.

I'm very excited by this because this has the potential to create a new disruptive business model for our technology, where the flexibility of the programmable logic is an inherent benefit over fixed-function ASICs, thereby expanding our reach into a much broader set of customers and applications.

More broadly, our investment and exceptional execution at the 16-nanometer node has extended our competitive lead to approximately 18 months. We are now shipping 14 unique products to more than 450 discrete customers. This represents a substantial increase from last quarter, where we shipped 12 unique products to 300 customers.

Additionally, we recently announced a major expansion of the 16-nanometer portfolio with our All Programmable RF SoC product family. This is a disruptive integration and architectural breakthrough to 5G wireless with RF class analog technology.

This family provides 50% to 75% power and dramatic footprint reduction, 5G, cable and wireless backhaul applications. Lorenzo mentioned we're extremely pleased that we delivered on our commitment of 6% revenue growth for fiscal year 2017. We remain committed to delivering to our longer-term operating margin target of 30% plus.

Our June quarterly guidance of $600 million to $630 million will again be driven from broad-based growth in our Advanced Products portfolio. As I mentioned before, this is already 49% of our current business, and hence helps drive our overall revenue forward.

I look forward to sharing more details on fiscal year 2018 with you at our Analyst Meeting on May 22 in New York. Let me now turn the call back to the operator for Q&A..

Our first question is from the line of Ambrish Srivastava from BMO..

Thank you very much. And, Lorenzo, good to see the accounts receivable come down. That had caused a lot of consternation last quarter. Moshe, I had a question on inferencing.

And I know that the investment case for Xilinx is not just on AI and machine learning, but there has been a lot of talk about it and you guys are showing up as well in the inferencing market. So my question is, help us understand some of the benchmarks.

And I don't expect you to get into a blog fest with the existing or potential customer, but Google put out a TPU specs against the GPU against the CPU. Where do the FPGAs stack up? Latency, throughput, anything you are willing and able to share will be very helpful. Thank you..

Okay. Well, I'll do my best. But realistically, we're going to provide a lot more information on this topic on May 22. So I encourage you to look for more detail then.

But generally speaking, first and foremost, the most valuable element we have is actually in the flexibility, and that is inherent to the programmable logic, and that is something which is very, very, very difficult to provide in an ASIC. And similarly, it's quite difficult to provide in a similar way at least using a CPU and a GPU.

And the importance of flexibility is that even though it sort of sounds like machine learning and inferencing and all of that, we are into the 70th year of hearing about it.

Reality is it's an emerging technology with very broad applicability where the benchmarks are very dynamic and the ones which are important today are going to be likely totally useless two or three years from now. And the biggest benefit we have is the fact that our devices can be programmed to match the new payloads and new algorithms, right.

Now, so that's sort of the generic thing, and I think everyone sort of misses that, and that's why I'm starting with this long semi-diatribe on this issue because that sort of is something that I think the market underestimates the inherent advantage we have there.

Now, if you sort of look at a specific set of applications, generally speaking, on things which do not require floating point, we tend to have higher performance and lower power than all other options.

Now, if you take one fixed application and you target a dedicated ASIC to it, yes, I'm sure they can do better, but then if anything changes then that ASIC all of a sudden becomes a lot less attractive on the new application, whereas the programmable logic can be modified to adapt to that.

And so generally speaking, we have much better performance than CPUs and GPUs for everything except floating point for each application and we have performance which is generally on par to a fixed but actually the biggest benefit is in the flexibility of our products and the fact they can be adapted.

With regards to specific benchmarks we can share the numbers we have with you, and we would be happy to do that. And this is one of the reasons that we're so excited about this market.

And if you sort of look at Amazon in addition to the inherent technology advantage we provide, it actually provides us with a perfect conduit to a very broad set of new customers. So both in terms of technology and access to the technology, this is a tremendous benefit for us. And we'll talk a little more about that too.

I think I've overstayed my welcome on this..

And our next question is from the line of John Pitzer from Credit Suisse..

John?.

John, your line is open..

Can you guys hear me now?.

Yes..

Yes..

Moshe?.

Yes, we can hear you..

Congratulations on the strong results. I guess, Moshe, my question is pretty simple. Up until about two or three months ago, a lot of our field work was suggesting that, while FPGAs were a great architecture for accelerators in the data center, there were still a lot of software compatibility issues.

And I guess we've been fairly impressed over the last three or four months how quickly some of those software issues, programmable issues, have been resolved.

So I guess my question to you is, relative to three or even six months ago, how is the acceleration opportunity today versus then? At last year's Analyst Day, I think you talked about $250 million of potential incremental revenue by 2020.

I wonder if you'd just give us a sense of whether or not relative to those expectations the acceleration market is actually accelerating? And I apologize for the pun..

Okay. So you're right. Last year, at our Analyst Day, we projected it to be between $200 million to 300 million in the calendar year 2020 timeframe. At the time we gave those projections, we were aware of the engagement with Amazon, but we were unaware of what they were planning on doing with it. And so for us, we have lots of customers.

And actually all of the big hyperscalers are customers, but they don't always initially share what their plans are.

What has happened over the past year, and in particular with Amazon, and that became clear only at the tail end of the calendar year, so I think it was November at the very earliest that we sort of found out what it is that they wanted to do.

We also detected that in order to make them successful, we needed to prioritize the software environment, which was something we were working on, but was probably at least a year away. We identified that that needed to be raised in terms of priority, and a lot of resource was put in place. Now, they've moved very, very quickly.

We have been delighted to support them with their very significant requirements, and at the beginning of the calendar year, they actually opened it up to limited release, and the limited release enabled FPGA experts to access the technology. So the technology became available in the cloud, but it was available to a limited set of customers.

And when they announced general availability, then that's a much broader set of customers. So the change is that Amazon has now legitimized this, is putting it in the cloud.

The software, which was clearly a handicap we had, we have invested very heavily and now the results the customers are seeing are very positive, and there's a follow-up release less than a month from now which will even broaden the target customer base even more significantly than it does.

Now, the underlying question is how much can this do for revenue, and I think it's still too early to predict that. Clearly, the potential is there for it either to be larger or for it to happen a little earlier.

It's still unlikely to impact the next 18 months in a significant way, but it is possible that by 2020, if this goes well, that the market could be larger than what we said, and this is for our technology, or it could maybe happen faster in 2019.

But what we're really trying to do is, we don't want to sort of generate any short-term frostiness because there really is not the proof in terms of deployment to justify that. There is the potential of that happening in a little further-out timeframe than what I think most people are looking at.

And so there is goodness and there is goodness in the technology. There is goodness in the opportunity, and there is definitely tremendous potential with Amazon putting this up in the cloud and making it available to everyone broadly. And they were very clear. You can look at their blog.

They sort of said this is initially being deployed in one location. If that is successful, then the implication there is that there could be significant deployment in other locations, which in of itself would be an accelerator or a potential enhancer of the business. But that first deployment needs to be successful.

And obviously both Amazon and Xilinx are working hand in hand to do the best we can to support it..

And our next question is from the line of Tristan Gerra from Robert W. Baird..

Hi. Good afternoon. Similar question in the ADAS business. How should we look at this business medium term for Xilinx? You've mentioned a high 20s market share in the past. Do you think that there is a trend where FPGAs can actually gain share against ASIC in that market? And we know the product cycles are very long.

And is programmability as critical as it is in the other market that you just discussed?.

Well, the programmability and the flexibility is valuable in this market, because this again is a market which is in flux. It is changing rapidly. What we are now deployed in is ADAS. We have a lot of design wins in advanced ADAS. Those put us in a good position to transition into the automated or autonomous driving over time.

We think that this is a process which, until it really becomes mainstream and very, very broadly deployed, it has a 10 to a 15-year horizon until that sort of happens. And at that point in time, it wouldn't surprise me if it becomes significantly commoditized.

But until it gets to that point, we do believe that the programmability is a major asset that we have. And this is a market we identified early, and we think that for the next two generations of deployments over the next five to seven years, we're in a very good position.

And our market share should grow a lot of the growth this current quarter in our automotive numbers, which hit a new record were due to a surge in ADAS.

And we'll give more data with regards to our overall practical businesses in May, but you can be sure that automotive is likely to be the highest growing or amongst the highest growing markets for our next year projections. Next question, please..

And our next question is from the line of Joseph Moore from Morgan Stanley..

Great. Thank you. I wanted to ask about the Zynq being over 10% of revenues. I was kind of excited about that. Can you just talk a little bit about what the drivers are? And you've talked about the number of design wins that you have in Zynq, the mountain of design wins, I think you've referred to it as.

Where are we in terms of monetizing that and how much do you think, how long do you think that you can grow at the rate that you have been in that Zynq product?.

Well, if you look at the Zynq products offering, it was introduced at the tail end of our 28-nanometer, so I believe the first tape-out was at the end of 2011 for Zynq.

And then silicon was probably available a year later, and we got a lot of design wins in three initial markets which it was targeted at were wireless, and it was won a lot of business in wireless. Actually it's at the core of our wireless strategy. Most of the deployment that you're seeing now in wireless at Xilinx or a lot of it is Zynq-based.

Second area, and this was a major driver for the definition of the product, was automotive, and that's a market that takes quite a bit longer, but is now hitting its stride.

And the third is industrial control, and that's where we have the proverbial mountain of design wins, which are now starting to turn to revenue, and that's just a function of how long it takes.

What we're seeing in parallel is a second generation Zynq that was targeted at the high and mid range of the market at 16-nanometer, but what we've done is we've since taken the 28 nanometer and expanded it down to lower-end devices.

So, you can tell from that that we expect it to be a big driver going forward and at 7-nanometer it's going to be part and parcel of a broad deployment in our technology, because we believe that at that point, 7-nanometer, it's going to be an integration play.

And nearly all of our customers will use the high-performance multi-CPU cores, which are part of that in their technology. So, if you sort of look at it, it's grown from 0% to 10% of our business. And my expectation is that it can grow from 10% to 20% much faster than it took to grow to 10%.

And maybe we'll try and quantify how much faster, but it's now no longer a new wonder. And just the repeat business that we can get in of itself should be a major driver in addition to the expansion that we have.

And if you go far enough into the future, it's likely that close to 100% of our business over time will be driven by the Zynq product offering. But that's probably 10 years out. That's probably not now..

And our next question is from the line of C.G. (sic) [C.J.] (27:23) Muse from Evercore..

Yeah. Good afternoon. Thank you for taking my question. I guess the question here is you guided comm and data center up sequentially. Curious if you could talk about what you've embedded in the guide for incremental growth from the data center.

And then as a quick housekeeping question, if you could offer your outlook for share count for the June quarter. Thanks so much..

Yeah. So two things. So I might frustrate you by not being very precise. But in the comms, we do show data center. We do expect data center to grow meaningfully. And again that's in the context of some of our older data center business not growing as much as we would like. And in certain cases, our design wins that we've had were slowing down.

So we do think that, in summary, that the data center growth, which is part of the overall comms growth, all the end other markets in the comms area we expect to be growing as well, but the data center growth will be driven by hyperscale.

Then with regard to share count, so I haven't been providing detailed share count guidance for the past couple of quarters for a couple reasons. One is I don't want to kind of signal exactly what we're doing on our buyback. And the second part is, the complexity of the convert and its impact on our diluted share count is hard to describe briefly.

So I would though recommend you, CJ, go to our website. We're going to update what we expect to be the impact on diluted share count from the redemption of the convert in this quarter. Like I said, it's fairly complex, but we've tried to make it understandable.

And you will see that it will have a significant impact in this current fiscal quarter, and then because of the way the accounting is done for the share count, it will actually even translate to a further reduction in the next quarter.

So I mean, I will actually, I'm happy to take a follow-up if there's something specifically you were looking for, but I'm not going to give you a precise share count guide..

No, that's very helpful. Thanks so much..

Yeah..

Thank you. Next question..

Our next question is from the line of John Vinh from Pacific Crest..

Hi. Thanks for taking my question. Question on 5G. I think you guys have talked about your 5G trial win rate at roughly 9% plus. Also if you think about 5G as versus 4G, you've got a 3x to 5x increase in radio heads, which is an area that you historically have done extremely well.

And you've also announced the RF SoC, which gives you an opportunity to expand your footprint because it integrates additional discretes. So, I was wondering if you could just put into context, how do we think about the 5G opportunity versus 4G.

Right, and conceptually it sounds like a very sizable opportunity, but I'm just wondering if you could also just talk about what are the other offsets that we should be thinking about..

Okay. So great questions, actually. There's a whole series of very detailed questions. So we will try and give you more clarity on that in May, but let me give you Reader's Digest. The broad deployment of 5G is expected to start in earnest in 2020 and we believe we're going to have a very good position there based on the design wins we have.

It is indeed quite likely that at least on the radio head side there will be a larger number than was done, you know, than were deployed in 4G. And fundamentally 5G is going to be essential in terms of enabling the infrastructure or a whole host of things which are extremely visible today, like this entire industrial IoT.

To do that well, the 5G technology is going to be a key enabler there. So we absolutely agree that the potential is larger. Our technology position is very strong.

We think that between now and 2020, there is going to be ongoing deployment for 4G, including there is going to be additional rounds even in China, which is now the tail end of I think the fifth round of LTE deployment.

And they will do a sixth round that is expected to be about two-thirds of the size of the fifth round, but that's still a very significant number. India is going strong, and there is other countries in the world where there is deployment.

So if you look at the overall wireless business, we expect it to hover around the mid-point of, for us, our high was close to $150 million a quarter. Our low was $75 million. We expect to hover around the midpoint over the next few years on average until the 5G starts in earnest.

Our technology position makes us very confident that during these next two years, all of the pre-5G deployment is likely to benefit us. And then when the real thing starts, then that's when you could see it approaching or maybe even surpassing the previous levels. But that's a few years away in terms of hitting over $150 million per quarter.

We don't see that happening any time soon. Hope that answers your question, and we'll try and give more visibility into that in May..

And our next question is from the line of William Stein from SunTrust..

Great. Thanks for taking my question.

Moshe, I'm wondering, as we head into the Analyst Day, in order to prepare for it, what should we be thinking about sort of the puts and takes that could take your top line above or below this year's roughly 6% growth?.

Well, we'll give you an annual projection. And we're delighted that we have, if you look at one year, we have 100% hit rate in terms of hitting that number. If you look back more than one year, we have a very low hit rate. And we're very proud of actually delivering the 6% on the nose and actually just exceeding that.

But we will provide you with guidance. I think the thing that gets overlooked and despite the fact that we emphasize it again and again and again is, we benefit immensely from the multimarket portion, and that sort of provides us with a lot less volatility than most other players.

And what we'll try to do is to highlight which of these markets are likely to, or have the potential of doing better, which are just likely to hit the natural growth rate and which some of them, for better or for worse are below the growth rate.

And we'll sort of highlight those, but there are sort of eight disparate markets and they all tend to have somewhat different dynamics. The ones which get the sexiest and get the most air time are data center and automotive, but we're actually doing very well on A&D. We're doing very well on test and measurement.

We've done better than most other semiconductor or all other semiconductor companies on wireless, in a time where the wireless industry has not done well. So, I think the under-appreciated part is the multimarket portion.

And if I have a request is that you'll come with some openness with regards to hearing about the other markets, even if they're less, have less curb appeal than the ones everyone is talking about..

That was a very artfully phrased question, too. Good for that..

And our next question comes from the line of David Wong from Wells Fargo..

Thanks very much.

Can you give us some idea of what you expect 16-nanometer revenues might be in the current fiscal year, fiscal 2018?.

We will give you guidelines for that in May. It's too early, but it's done great, and a lot of that is driven by the Amazon deployment, but it's not all of it. There's actually, you can just tell from the number of customers in the number of markets we're in that it's doing better.

It's actually getting deployed faster than any other technology we have seen in the past and we're also benefiting from a very favorable competitive position..

Okay, great.

And for my follow-up, can you give us any idea of whether you're seeing any pickup in defense, in sales into the defense segment?.

Well, there was a nice pickup this quarter..

It's generally been on an upward trend as design wins that we have, one in our 28-nanometer generation ramp in. As you know, in defense, it's a very, very long-tailed business. So at any point one time, we have multiple generations, multiple of our product generations being sold into the industry.

So given our competitive position and strength of design wins and our continued design wins in large opportunities like Joint Strike Fighter, we feel very good about the business..

It's already the third largest..

Largest business..

Market we're in, and it's not that far away from wired and wireless. Smaller than wired and wireless, but it's approaching those numbers..

Our next question is from the line of Blayne Curtis from Barclays..

Hey, guys. Thanks for taking my question. I just want to circle back to the AWS announcement, just kind of the concept of FPGAs as a service, or available in the cloud.

And kind of what size of consumers do you think would use this, whether you'd see any larger scale customers deploy it, or is this more for smaller? And then you can you just talk about competitive landscape there, if you expect you'd be the only offering or if you would see any from Altera as well?.

So there's seven big players, and we are talking to all seven of them. The one which has the largest existing deployment is Microsoft, and that is using the Intel slash Altera for that. We believe we have a product which is so much better than that, that obviously that is at the core of our success with all of the players.

With regards to the breadth of the usage, the beauty of this is, in particular in the Amazon case, is that this enables a broad new set of applications to evolve in a way which the traditional way of pursuing the business was just not capable of doing.

And it is inherent to the Amazon model that they don't actually know themselves yet, because they don't have the end system expertise. They're not experts in all of these areas. What they do and what they are experts at monetizing is making it easy to use and making it available and then enabling an amazing ecosystem where the players use it.

And what we are seeing and just they talked about thousands of customers applying from the very first minute, several hundreds getting access during the limited release and then being open to the potentially thousands of customers in the general availability.

Our expectation is as this unfolds, and that's why we're being cagey on providing numbers, right.

That's sort of what we're trying to – we believe this could have a viral impact and not only would it be used in a public cloud sort of applications like Amazon, but when people start seeing the benefits of that, they'll just think that even for their specific applications, they could see huge improvements.

And as this transition is happening into a world where machine learning is becoming broadly used for a whole host of applications, that these sort of open up and hence there's huge potential there. Also, that's the reason that we are loath to try to quantify it yet.

But yet, all of this change is quite a monumental one in terms of accessibility, in terms of ease of use, and the breadth of the benefits and the benefit of the programmability, because there are some misguided notions that an ASIC by definition is better than an FPGA.

This is the one set of applications which an ASIC, due to the fact that it takes years to design and it's very rigid, you could do a wonderful job designing absolutely the wrong ASIC, and then deploy it at a time where the market has changed, whereas if you use programmable logic, you can change it in situ, in live applications and it can adapt to these new applications and benchmarks, so hence the benefits.

The other benefit is that a lot of these applications, in particular on the inference side tend to have more of a integer nature to them as opposed to the traditional floating point applications.

And as a result, and it's sometimes it's a flexible integer format that they use, and there is nothing more flexible or better than programmable logic in terms of doing that. So I know I haven't quite answered your question, but I think that's the potential.

Lorenzo, also?.

Yeah. So I think an element of your question, Blayne, was maybe related to whether the applications on AWS are kind of nichy or small or whether there's some big entities using it.

And the launch partners range obviously from some smaller entities to some very large entities, and they're doing a range of applications that are fundamental to each of the companies in the core businesses. So I do think the scaling opportunity is there and very well represented..

And our next question is from the line of Ross Seymore from Deutsche Bank..

Hi, guys. Just one near-term question.

The guidance for the June quarter for industrial, aerospace and defense being flat and then down in the broadcast, consumer and auto segment, could you give a little bit more color about the moving parts in that? And then maybe a slightly longer-term aspect of it, I know you said that automotive would be a fast grower, and you'll give us more details at the Analyst Meeting in May.

But the other part of that equation, the broadcast and consumer, has been a bit of a headwind for a couple of years. So I just wondered what the trajectory of that might be over the course of the year. Thanks..

Yeah. So the industrial A&D softness is due to program-specific things in defense primarily. The rest of the business areas are relatively going to be flat relatively quarter on quarter. In the automotive, consumer, broadcast space, as we've mentioned in this call already, we've set a record in automotive.

And if you go back a quarter, we talked about kind of a low based on some inventory things. So we're going to see a little bit of dip in automotive. And by the way, though we expect a longer-term trend upward based on the strength of our ADAS design wins, we won't be surprised by some quarterly fluctuations.

But underneath that, the ADAS business is very strong. It's just kind of a quarterly correction and the rest of the end markets in that segment are flat. We will probably talk a little bit more in May about longer-term trends in audio/video broadcast, but you're right, the industry is facing some headwinds, and we are feeling those as well..

And our next question is from the line of Hans Mosesmann from Rosenblatt Securities..

Thanks. Hey, Moshe, can you give us your feel for your roadmap as you go from 2016 to 2017 and beyond in terms of process technology? And that is in the context of the update that Intel gave the Street a few weeks ago regarding their roadmap and their density advantages. Thank you..

Well, we're delighted to be with TSMC. It's one of the best decisions I've ever made. Maybe professionally, it's the best decision I've ever made. We don't regret it for one minute. They are great. You know, look at everyone who has tried someone else, including Intel. Great company, great manufacturing technology, great CPU provider.

The foundry business is a service business. It requires technology leadership. It requires support. It requires an ecosystem. TSMC is second to none, absolutely second to none. I mean, that's sort of really not something I worry about and I'm not in any way understating Intel's capabilities.

I'm just sort of saying, as a foundry, I think we're absolutely with the right foundry. What we now uniquely benefit from is, they have such a huge portion of business which comes to them from the wireless smartphones, which in a very, very short period of time enables them to get to incredible defect densities.

If you can their investment in capacity, they have for several years now invested $10 billion every year. In terms of access to technology and support technology, in terms of addressing fluctuations, in terms of requirements, they really are best-in-breed. So yes, Intel is great, has great manufacturing capacity.

TSMC is by far the best foundry in the business, and they didn't get to having $30 billion worth of foundry business without that sort of excellence. So I'm not at all concerned about access to leading-edge technology. I believe that TSMC can provide us with that and has in the past, is now.

If you remember, there was a huge excitement when Altera initially committed to Intel, and the common wisdom would be that they're two years ahead of us. I would say now they're comfortably, and this comes from customers, 18 months behind us. Right. So, we will continue to pursue TSMC and benefit from that relationship..

And our next question is from the line of Chris Danely from Citigroup..

Hey thanks, guys. Nice job of bringing the OpEx down.

Can you just talk about OpEx trends after this quarter? Do you expect it to be at the same percentage of revenues or go up or can you take it down even more?.

Chris, we'll, like all the other questions that relate to longer-term views on the business, we'll talk about that more in our Analyst Meeting..

But we are committed to, in the longer term, to get back to 30%..

30%..

And higher, right? So, it will hopefully continue moving in the right direction and we'll give more granularity to that in May..

And at this time, I'm showing no further audio questions. Presenters, I turn it back to you..

Okay. Thanks for joining us today. We'll 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 Investor Relations website. Our next earnings release date for the first quarter of fiscal year 2018 will be Wednesday, July 26, after the market close.

We'll be hosting our Analyst Meeting in New York City on May 22. We definitely look forward to seeing you there. In addition, we'll be attending the BofA Global Tech Conference on June 7 in San Francisco. This completes our call. Thank you very much for your participation..

Ladies and gentlemen, once again, we appreciate you participating in today's fourth quarter fiscal year 2017 earnings release call. You may now disconnect..