Advanced Micro Devices Inc (AMD) — Q2 2017 Earnings Call Transcript

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

Thank you, Rick. In the September quarter, Xilinx sales were $579 million, up 1% sequentially and up 10% on a year-over-year basis. Sales for the first six months of the fiscal year were up 7% versus the same period of the prior year. Our advanced products increased 10% sequentially and 61% year-over-year, with 28-nanometer, 20-nanometer and 16-nanometer sales reaching new records. Sales from our 16-nanometer UltraScale+ family were significant and exceeded our expectations. Our sales in the September quarter demonstrated the resilience of our multi-market model, growth in Industrial and A&D and Broadcast, Consumer & Automotive categories more than offset the expected decline in Communications. The Industrial and A&D category stood out with strength in test, measurement and emulation, as well as Aerospace & Defense. In fact, sales from test, measurement and emulation reached a record during the quarter. Broadcast, Consumer & Automotive was up slightly, better than expected due primarily to strength from AVB. Gross margin in Q2 was 69.6%, lower than guided due to the impact of a licensing agreement with Rambus. Moving forward, the impact of this agreement is expected to be immaterial to our gross margin. Operating expense was $227 million, lower than guided due to lower compensation expenses. Operating income for the quarter was $177 million or 30.5%. Other income and expense was an expense of $1 million, lower than guided due primarily to an investment distribution. The tax rate for the quarter was 7%, which included the previously disclosed benefit of $9 million or $0.03 per diluted share related to the closure of the audits of fiscal years 2012 through 2014, and a $0.02 per diluted share benefit related to other discrete tax items. Our net income for Q2 was $164 million or $0.61 per share. Operating cash flow was $184 million. Diluted shares were 270 million, including 14.8 million from the convertible, higher than guided due to the impact of our higher share price. We repurchased 1.9 million shares for $100 million and we paid $84 million in dividends. Now some key points on the balance sheet. We ended the quarter with $3.7 billion in gross cash and $2.1 billion in net cash after our debt. Inventory was $197 million, approximately flat with the prior quarter and down $16 million from the same quarter a year ago. As I turn to our guidance, I would like to refresh the key points of the annual guidance from our Analyst Day in May. At the Analyst Day, we forecasted our revenue growth to be between 4% and 8%, our gross margin to be between 68% and 70%. And our operating expense to be between $930 million and $950 million. Our performance for the first two fiscal quarters, our guidance for Q3 and our expectations for Q4 indicate that we remain centered in the range of the guidance provided at our Analyst meeting. In the December quarter, we’re expecting sales to be flat sequentially. Our backlog is slightly down heading into the quarter. We expect advanced products will continue to grow and we are expecting communications to increase with growth in both wireless and wired. Industrial and A&D is expected to decline as test and measurement comes off of two consecutive record quarters. Broadcast, Consumer & Automotive is expected to be approximately flat sequentially. Our gross margin will be approximately 69%. For operating expense, we continue to invest in accelerating our leadership position as discussed at our Analyst Day. We expect to see operating expense increase to approximately $245 million with a significant portion of the increase coming from tape-out expenses. Our operating expense outlook includes $1 million of amortization. Other income and expense will be an expense of $2 million, and will likely approximate this level again in Q4. Finally, our tax rate is expected to be 14%. Now, I’d like to say a few words on our capital allocation strategy. We are confident in our long-term growth strategy and expect to continue to generate healthy cash flows. Note our operating cash flow from year-to-date exceeds $500 million. Our capital allocation priorities remain the same: invest in our business to capitalize on our leadership position and return of capital to our shareholders with a commitment to an increasing dividend over time and through share repurchases. With respect to share repurchases, as we previously announced in May of this year, the Xilinx’s Board of Directors increased the repurchase authorization by $1 billion. Our plan is to take a more deliberate approach on share repurchases with the intention of exhausting the $1 billion authorization over the next several quarters. Let me now turn the call over to Moshe.

Thank you, Lorenzo. I’m very pleased with the financial results of the second quarter. We delivered our fourth consecutive quarter of revenue growth, resulting in operating profit in excess of 30% for the third consecutive quarter. Our overall revenue growth was, as in the previous quarter, driven by our Advanced Products. These increased again by over 60% year-over-year, delivering yet another significant quarterly record and now account for 46% of our total sales. Overall, five of our eight end-markets grew in the quarter demonstrating the robustness of our diversified multimarket portfolio. Our unparalleled execution on the 28-, 20- and 16-nanometer nodes has now delivered three consecutive generations of very significant technology leadership. Our world-class partnership with TSMC, the foundry technology and market leader, has been key to this success. In the September quarter, for the fourth consecutive quarter, 28-nanometer sales delivered a new record. 28-nanometer sales growth was enabled by our Zynq product line, which is a significant component of our market expansion efforts. Zynq product sales increased 25% sequentially, driven by automotive, wireless, industrial, and consumer applications. 20-nanometer revenue again reached a record level, significantly exceeding our $40 million target. Sales were driven by strong contributions from our test and measurement, wired, and wireless markets. We expect to ship over $50 million of 20-nanometer products in the December quarter. 60-nanometer sales grew significantly in the September quarter, similarly exceeding our forecast, while shipping to a broad base of end markets, including automotive, data center, and communications. At our Analyst Day in May, we discussed the strategic increase in our R&D investment in fiscal year 2017 to both solidify our leadership position and expand our product portfolio. We recently highlighted two significant accomplishments that demonstrate the early fruits from that investment decision. First and foremost, we reached ahead of schedule a critical production milestone with our 16-nanometer UltraScale+ portfolio, further extending our product lead to well beyond the year. We’re already shipping nine unique revenue generating products to approaching 200 active customers. Per Lorenzo’s earlier comment, we expect to deliver multiple production tape-outs in the current quarter to solidify this lead. In addition, we announced the expansion of our cost optimized Spartan, Artix, and Zynq 28-nanometer product families, targeting a wide range of applications including Embedded Vision and Industrial IoT. We recently made two announcements illustrating our significant momentum in cloud computing, another of our key emerging market expansion opportunities. Baidu has adopted Xilinx to accelerate machine learning applications in the data center and will leverage our platforms in their initiatives to develop commercially viable autonomous cars. In addition, the CCIX technology consortium, formed to bring a high-performance open acceleration framework to data centers, tripled its membership and released the specification to consortium members. We’re uniquely positioned to capitalize on our excellent execution, along with innovative and game-changing products to both accelerate share gains and drive market expansion. We remain committed to maintaining and extending our leadership while delivering on our long-term operating margin targets of 30% plus, and per Lorenzo’s earlier comments, underlining our commitment to returning cash to shareholders. Let me now turn the call back to the operator for Q&A.

Yeah. Good afternoon, guys. Thanks for letting me ask the question. Lorenzo, in your prepared comments, it sounds like you’re still on target to hit that 6% growth guidance for the fiscal year. I dislike talking about seasonality because there’s so much variance around it. But given the guide for the December quarter, that kind of implies the March quarter up about 6% versus a seasonal up 3%. I wonder if you guys can just talk about if there is anything specific you’re seeing maybe around the F35 revenue stream or any other sort of new product revenue stream that gives you confidence as you look out that far that you have that kind of above seasonal growth?

Yeah, so thanks, John. Obviously, as we close this quarter and we go into Q4, we’ll be providing you with more specifics. But I think right now, what we will tell you that underneath our view of Q4, of our fiscal Q4, is a view that communications will be firming and we will see growth in the Industrial and A&D category as well as the Consumer, Automotive, and Broadcast category; some of the reasons you’re alluding to, but there is more there than that.

Got it, and then, as my follow-up guys, Moshe, the Baidu announcement was a nice announcement around acceleration. I’m wondering if you can help us understand kind of the opportunity there. Are you sitting next to Intel processors? And then how do we think about the acceleration sort of growth developing in the near-term, because our thought process was that was perhaps going to be more a 2018, 2019 event relative to the Analyst Day you guys had earlier in the year? It seems like you’re getting some of that earlier. Do you expect to see sort of additional follow-on here or is this going to be still something you expect it to mature in the next couple of years instead of the next couple of quarters?

John, it’s the next couple of years; it’s not the next couple of quarters. We believe that this has the potential of being $200 million to $300 million by 2020. And it will have to grow to that number. It’s not going to happen overnight and it won’t happen over the next six months. It’s going to take more than six months just for these devices to go into volume production. So this will not generate the Q4 upswing that you asked about earlier.

Okay. Thanks, guys. Congratulations on the good results.

Sure, thank you.

Thanks, John.

Hi, thank you very much. I had a question on the share buyback. And I’m not very clear, to me at least, Lorenzo, what you meant by a more deliberate approach on the buyback. And I have a quick follow-up after that, please.

Sure. So I think we’ve had questions over the past few months after we announced the Board’s authorization of an additional $1 billion in May, about what our plans are. And I think in the past, we’ve characterized our approach to share repurchases as opportunistic, which is fairly ambiguous. What we’re trying to communicate is we have the intent to utilize the authorization over - and I chose the word several - next several quarters deliberately to indicate a rather more defined time-span on when we’re going to utilize the authorization.

Okay. And then a quick follow-up, Moshe, on the nodes, 16-nanometer, where is the difference in this node versus what you’ve seen in the past? And Intel just announced their product and they said that they would be shipping for revenues. So just help us understand kind of where you are in terms of being, I think, over a year ahead versus them, and then what kind of opportunities does it open for you? And if we’re just going back to the last nodes it’s clear at least to us that it opens up like a 65-35 kind of market share. So any help on the opportunity you’re seeing versus the last node will be great. Thank you.

Sure, Ambrish. So we taped it out in June of 2015. It came back in September of 2015. It was in such pristine condition that at that point in time we were able to demonstrate devices working and achieving the specs, which for silicon immediately in the first day is as good as it gets and probably better than nearly everyone else. And we sampled the parts to our customers and we have been shipping them to numerous customers in a whole host of applications. And the feedback we’re getting is between the software, the silicon and IP customers have never seen from anyone anything at this level of functionality. And as a result what we have done is we now have nine products that are shipping. And we’re moving into production and part of the bump that is happening this quarter, the current quarter we’re in, in terms of R&D and the quarter after that will be a whole slew of tape-outs, which indicate that we’re going to move into production at an early point. So in terms of sampling, we believe we are clearly a year ahead in terms of the maturity of the product and from what we’re hearing from our customers we’re much more than that. And that’s the position we’re in; it’s a very enviable position. And that’s why we are increasing our investment to exploit it. Hopefully, that answers your question.

Yeah, thank you, Moshe. Good luck.

Sure. Thank you.

Great. Thanks for taking my question. Moshe on the automotive side, can you remind us what’s driving that growth today? And then turning to this autonomous driving agreement with Baidu, is that expected to drive Xilinx units in the vehicles, so for scoring or inference? Or is it more on the deep learning side, not so much on the client side?

Okay. So our automotive business historically was more on the entertainment side. And our Zynq product offering was defined specifically to address driver assistance. And that is a huge step towards delivering autonomous vehicles. What is happening now is the part that is growing is the advanced driver assistance. And we actually shared that we expect to have 85 models in production in this year. And we remain on target for delivering those 85 models. If you noticed, I had mentioned that quite a bit of our early revenue on 16-nanometer was in automotive and that comes from the second generation of our Zynq product, which is the MPSoC. It’s a more powerful element. Now, in parallel to that as being in a pull position for this market, we are working on the next generation, which is the 7-nanometer product. And there, we expect that as the world moves from driver assistance to autonomous driving, we expect it to move in phases. There are question marks with regards to how quickly it will happen. There is a whole host of barriers regarding safety, security, and legislation that will modulate and throttle that growth. And so, we believe that a lot of the work is being done now and there has been a lot of early deployment. And what we’re seeing is that that market is expanding beyond the traditional suppliers, beyond the tier 1 providers to those suppliers, and there is a whole host of new players who are entering the market in addition and are deploying technologies which they have developed for their data center applications. And so, we expect this to continue to unfold to be honest over the next 10 years. We don’t think it’s going to all happen tomorrow. We do think there is going to be a lot of proof of concepts and very limited deployments over the next year or two. But we think that the massive volume will actually come in the 7-nanometer in the 2020 and beyond timeframe. And there will be a lot of different approaches. We believe that our technology and our solutions will be relevant to most, if not all of those targets, right. And there is quite a bit of fragmentation; there are a lot of different approaches there. But our technology fundamentally should be very viable to address all of those. And our early position with ADAS gives us good understanding and good relationships to - with the providers to address that market.

Thanks for the color.

Thanks for taking my question. Moshe, this is sort of a longer-term question. The promise of FPGAs has been that there is a very long life to current generation and legacy products. And then you lay around growth from next generation, so certainly your next generation is growing at a very phenomenal pace. But when I look at your legacy or core products they are declining at a mid-teens pace or so. Is that different from what you’ve observed in the past? I.e., that the core products are declining faster than they had in the past or am I not looking at this the right way?

No. You are looking at it right. And those numbers are difficult to argue with. I would say that first quarter effect is, are you winning or are you not winning at a certain node? And to be honest, the decline in the old product is because there are two significant generations where we did not do well, right. And if you look at the 40-nanometer, at the high-end we were late, where the product had some issues. And prior to that, there was also a generation which had its challenges. And so as a result, that is now declining at a fast rate. The good news is now we have three nodes in a row that we have had great execution and significant market share. And because it takes a long time, and you are absolutely right in that, it takes time for that to compensate for the old node. The good news is once you get to 50%, right, and we are very close to that 50%, then you have a bigger, a much stronger tailwind, and the headwind is diminishing. And so, what you should see is the percentage that has the tailwind is growing now at a faster rate and will soon reach a magnitude that it can compensate, and actually more than compensate for the products which are not growing. And so your analysis is correct. But I think you need to take that first-order impact is, did you have a good node or not have a good node. If you have three good nodes in succession and our recent nodes, then once that gets to 50% of revenue that should more than compensate for the shrinkage of the old portion.

Got it. And then just as a quick follow-up, when I look at the growth rate that Intel has given for Altera, I understand that they are not having too many details. But they are saying it’s about similar to what you’re targeting, it is about 6%, right, or so. So is it that it is too early for us to see your share gains at some of these new nodes, because if you have been gaining so much share at 28 and then 16, and at 20-nanometers, why isn’t Xilinx growing faster than the growth rate that Intel has mentioned around Altera? Thank you.

Yeah. It’s very difficult for me to answer that question. And it’s even more difficult for me to answer it since they’re part of a very large company with a lot of things going on. And so, what I would say is we believe in the 6%; we need to deliver the 6%; if they deliver 6% or higher, good for them; we need to deliver 6%. All of the published numbers with regards to new nodes show that Xilinx share at 28 is well over 60%; at 20 is significantly higher than that, and obviously at 16, 14, right. We are in terms of shipments for revenue we have all of the shipments for revenue at this point and we have over a year lead. So I think we’re in a good position and that position has strengthened over the past three generations. And we now believe that that is at the root of enabling us to deliver 6% growth and the comparisons are going to be more difficult in the future.

Hey guys, thanks for taking my question. You mentioned 25% sequential growth in Zynq and you mentioned autos first. Just curious if you could give any primers as to how big auto is getting within that bucket or maybe what that business is growing for the fiscal year?

I’m sorry, Blayne. When you say what that business are you referring to auto or was it…?

You said - autos were the first one you mentioned so I was curious if you could give any color as to how big auto is getting within that segment and what that may grow for the fiscal year.

I think last time I got tangled up in the Zynq and the auto. So auto, again the dynamic in auto is we are seeing the decline in the older line business, which is infotainment, and the growth in the ADAS part of the market, which is driven by Zynq. In aggregate, the automotive business for us is around 7% or 8% of our overall business. We think that will grow over time as the ADAS growth rate outpaces the decline in the infotainment part of the business.

Got you. And I was just curious, your perspective on the wireless market, a decline in September, but then slight growth. Just curious just - what you think the end-market is doing and whether you’re benefiting from any share gains?

Well, what we see in terms of the end-markets are the customers we have, have been doing relatively well with; I don’t think I would call it strength. I would call it stabilization; in the China market and growth in India, both contributing to our year-over-year growth. Now, we had - we’ve said this before, we had a relatively poor year last year in wireless. Again, aggregate though this year we’re expecting the overall segment to grow a little less than 10%. But what we’re seeing right now through the year is not unexpected for us, which is it will be up and down throughout the year a few million dollars a quarter. But again getting to a place where we show year-over-year growth. So it’s going to be a few million dollars of cycling in and out each quarter. We were down last quarter from Q1. We expect to be up in the third quarter from the second quarter. So that’s not unexpected from what we had looked at for the year in wireless.

Hey, thanks guys. Just first a clarification on the share count. So you said, it’s going to be a little more deliberate. I think your shares are basically the same they were five quarters ago. So is the goal to slowly take the share count down or keep it flat over time?

The share count should go down. I think, if you’re - and Chris, you need to always keep into account when you’re looking at our diluted shares, the impact of the convertible that we have, which will obviously increase the diluted shares if our share price goes up. So overall though, we’re tending to bring the share count down.

Yeah. Okay, great. And then, just another clarification on the data center opportunity, Moshe, you had a question on Baidu earlier, and you said $200 million to $300 million by 2020, is that just Baidu alone or is that total data center?

No, that’s total data center.

Okay, great.

And we already have - we have two in production. We expect to have five in production in a year or two, five of the major seven in production.

And still on track for the tens of millions of revenue next year?

Yes.

Thanks very much. Can you give us an idea of what your total Zynq revenues in dollars are at the moment?

We’re looking it up.

Yeah. I just want to make sure. In aggregate, Zynq will be between 6% and 8% of our overall revenue.

Okay, great. And what functions are the Xilinx chips typically used for in ADAS? And what’s your dollar potential per car?

Okay. So the way to look at it is we just talked about the 8%, right, which is automotive, over 50% of that 8% is in ADAS. And those - right, so let use 5%, right. So we are looking for the year, if you look at our target it’s about $120 million for the year, give or take, which will come from ADAS. So a lot more than that if you look at all automotive. And this is spread over 85 models. The devices typically are used at the high-end and mid-range, but now are spreading down. And they are in the tens of dollars per unit.

Yes. Thank you. Just a question on the gross margin guidance, obviously going to the midpoint of the full year expectation, 69%, it is down 60 basis points sequentially at the midpoint. Mix is slightly less favorable. And the Rambus licensing headwind is going away. So I’m just trying to understand the trend there to get to that guidance.

Are you talking specifically about Q3, Ian?

Your Q3, yes.

Yeah, so I think, if you look at my end-market discussion, where we talked about communications strengthening, we have a wireless element to that. And the test and measurement part of the business coming off of record-highs. You can rationalize the mix down on gross margin.

Great.

There are other factors obviously, but if you - that’s the most kind of straightforward way to look at it in terms of mix.

Okay, great. And just a question here, obviously, it’s a period of elevated 16-nanometer investments, a lot of new parts in each of the families. I’m just trying to understand, I mean, how confident that you are that each part that is being defined can generate some pretty good ROI? Is each of these parts sort of addressing a specific application in a top customer and do you see almost like tailored parts now in terms of resources?

Yes, the parts are - if you look at the entire family, there are lots of family members. And what we try to do is to make sure that the mix of resources addresses the different needs of each of the market. If we had perfect planning and vision of what the customers really would need as opposed to what they tell us they need, then I would tell us we are doing great. Based on our experience you look back and then you find out that some have overachieved and some have underachieved. But we do try to target the different requirements. And sometimes we get it more right and wrong. But it’s not a perfect science and it’s not necessarily because our ability to predict is impaired. The customers’ needs tend to change and what we find is that they end up requiring different mix and higher performance, etc. And so, that’s the beauty of our business and that we have this whole range of products. So if we miss it on one, we can typically get it on the larger one, albeit sometimes it comes at a margin cost to address the specific market requirements.

Okay. Thank you.

Sure. Thank you.

This is Vinay calling in for Joe. I wanted to follow up on the Industrials and A&D segment. And can you just talk about now what kind of visibility do you have in that segment? And more importantly, how should we think about the long-term growth trajectory for the different buckets in that segment?

Okay. So this is a really mixed category for us in light of the elements which you’re thereafter. Then in certain cases, with very large customers, and large and more public deals like the Joint Strike Fighter in A&D, we have fairly good visibility up until the actual letting of the purchase orders on when you can see the business. But in general, we can map those over time. Test and measurement, to a large degree, with some customer concentration, we know we’re doing extremely well there based on our leadership products; it’s just unmatched. There is no competition for the capability we bring there. And so we have relatively good visibility to the program health. What we actually can be positively surprised by in that area is the success of our customers, as their platforms win in the market. Industrial is a lot of customers for us. And so, what we are typically doing is looking at a few specific opportunities with customers, and then looking at an aggregation on the broader customer base and seeing what we behave there. So each of these actually has relatively strong growth prospects, kind of defined by the markets they serve. Industrial, as we said before, driven by the Industrial Internet of Things and platform adoption will be broad-based and grow, but at the nature of that business is more slowly over time. A&D has actually been a relatively bright spot for us. We continue to win broadly and designed outside of the Joint Strike Fighter and the Joint Strike Fighter follow-on technology insertions, and so we’re seeing a lot of broader program strength there. In test, measurement and emulation, what you will see is that our leading products would gain share and grow relatively rapidly. And then will be filled in by other elements of that category like semiconductor test business as that broadens our footprint. So again it’s a very complex category for us to provide one answer to. Hopefully, I gave you some color.

Got it. That’s really helpful. From a follow-up I want to talk about 5G; you talked about 5G build starting in 2018 timeframe. But can you provide us an update there in terms of what are you hearing from carriers in terms of roadmap and more importantly, what does that mean for you in terms of content opportunity?

So 5G, the market has changed quite a bit and up to a couple of years ago. It looked like it was going to get pushed out. And definitely the standard is not quite stable and isn’t complete yet. So there is quite a bit of work that still needs to be done to come up with the standard. But what we have seen over the past two years is that all of the major players have shifted to a desire to be first and have significant market share. And we’re seeing them all race. And we believe that our 16-nanometer portfolio and our 20-nanometer portfolio is being used in close to 100% of the early prototypes, which are being used to demonstrate the technology. What we do expect is that the massive deployment will start in 2020 onward. And that the initial deployments, between now and 2020 are happening and they will - there are various flavors, 4.5, 4.9; there is a whole host of numbers which are being thrown around; pre-5G, 5G prototypes. But we are in a good position for all of those, but we can tell just based on technology position, but we do believe that the massive deployment will be 2020 and onward with the combination of programmable devices. And then for some of the companies they will definitely over time design ASICs to address that market. As it stabilizes and as the standards become clearer and the market requirements become clearer. Hopefully, that helps.

Got it. Thank you.

Hi, guys. I have one question and one housekeeping item as a follow-up. And the main question, pulling the wireless incomes back to a little bit of a near-term timeframe, generally, what gives you confidence both the wireless and the wireline business will grow both in the December and March quarters? They’ve been notoriously lumpy over the last couple of years. And so I’m a little surprised that you’re confident in sequential growth in both of those quarters. So any color you could give there would be helpful.

Yeah, I mean, the straightforward answer is on wired we actually have seen the design-wins, that I think we’ve alluded to several times in the past, actually beginning to ramp. Obviously, there is some aggregate macro CapEx uncertainty that could impact that. But underneath what’s going on in the industry that the products we’re in, OTN and GPON, we tend to see those ramp with infrastructure build-out internationally and domestically is happening. But you’re right, that can be lumpy. And wireless, again, I’ll go back to the answer I said earlier. We have actually visibility into our customers’ demand on us and their view on what they need to build out in particularly in China and in India on 4G and then the pre-5G business that Moshe referenced just a second ago internationally, North America included, that’s what we see.

Great. And then I guess just a one housekeeping one back to the share repurchase plan that you talked about. Just any color you could give, Lorenzo, on the percentage of the $3.5 billion in gross cash that you have that’s held onshore and what percentage of ongoing free cash flow is generated onshore?

So the way to answer that is of the $3.7 billion of gross cash, about two-thirds of it is categorized as permanently reinvested offshore. But about a third of the total has been deemed tax covered, I guess if you will, for the United States, so approximately two-thirds is available in the U.S. for use without book tax implications. It will be some cash tax implications on about a third of the total. Okay, and the mix of it as we generate it overtime, it’s a little dynamic.

Did you say two-thirds of it was offshore or one third was offshore?

Well, it’s - so there are two-thirds of it offshore, but two-thirds of the total are - so half of that two-thirds, so one-third of the total. We have basically provided tax provision for domestic purposes. So we could utilize that without having a U.S. tax book consequence.

Got it. Thank you.

Yes. Thank you. Just a question on the gross margin guidance, obviously going to the midpoint of the full year expectation, 69%, it is down 60 basis points sequentially at the midpoint. Mix is slightly less favorable. And the Rambus licensing headwind is going away. So I’m just trying to understand the trend there to get to that guidance.

Are you talking specifically about Q3, Ian?

Your Q3, yes.

Yeah, so I think, if you look at my end-market discussion, where we talked about communications strengthening, we have a wireless element to that. And the test and measurement part of the business coming off of record-highs. You can rationalize the mix down on gross margin.

Great.

There are other factors obviously, but if you - that’s the most kind of straightforward way to look at it in terms of mix.

Okay, great. And just a question here, obviously, it’s a period of elevated 16-nanometer investments, a lot of new parts in each of the families. I’m just trying to understand, I mean, how confident that you are that each part that is being defined can generate some pretty good ROI? Is each of these parts sort of addressing a specific application in a top customer and do you see almost like tailored parts now in terms of resources?

Yes, the parts are - if you look at the entire family, there are lots of family members. And what we try to do is to make sure that the mix of resources addresses the different needs of each of the market. If we had perfect planning and vision of what the customers really would need as opposed to what they tell us they need, then I would tell us we are doing great. Based on our experience you look back and then you find out that some have overachieved and some have underachieved. But we do try to target the different requirements. And sometimes we get it more right and wrong. But it’s not a perfect science and it’s not necessarily because our ability to predict is impaired. The customers’ needs tend to change and what we find is that they end up requiring different mix and higher performance, etc. And so, that’s the beauty of our business and that we have this whole range of products. So if we miss it on one, we can typically get it on the larger one, albeit sometimes it comes at a margin cost to address the specific market requirements.

Okay. Thank you.

Sure. Thank you.

This is Vinay calling in for Joe. I wanted to follow up on the Industrials and A&D segment. And can you just talk about now what kind of visibility do you have in that segment? And more importantly, how should we think about the long-term growth trajectory for the different buckets in that segment?

Okay. So this is a really mixed category for us in light of the elements which you’re thereafter. Then in certain cases, with very large customers, and large and more public deals like the Joint Strike Fighter in A&D, we have fairly good visibility up until the actual letting of the purchase orders on when you can see the business. But in general, we can map those over time. Test and measurement, to a large degree, with some customer concentration, we know we’re doing extremely well there based on our leadership products; it’s just unmatched. There is no competition for the capability we bring there. And so we have relatively good visibility to the program health. What we actually can be positively surprised by in that area is the success of our customers, as their platforms win in the market. Industrial is a lot of customers for us. And so, what we are typically doing is looking at a few specific opportunities with customers, and then looking at an aggregation on the broader customer base and seeing what we behave there. So each of these actually has relatively strong growth prospects, kind of defined by the markets they serve. Industrial, as we said before, driven by the Industrial Internet of Things and platform adoption will be broad-based and grow, but at the nature of that business is more slowly over time. A&D has actually been a relatively bright spot for us. We continue to win broadly and designed outside of the Joint Strike Fighter and the Joint Strike Fighter follow-on technology insertions, and so we’re seeing a lot of broader program strength there. In test, measurement and emulation, what you will see is that our leading products would gain share and grow relatively rapidly. And then will be filled in by other elements of that category like semiconductor test business as that broadens our footprint. So again it’s a very complex category for us to provide one answer to. Hopefully, I gave you some color.

Got it. That’s really helpful. From a follow-up I want to talk about 5G; you talked about 5G build starting in 2018 timeframe. But can you provide us an update there in terms of what are you hearing from carriers in terms of roadmap and more importantly, what does that mean for you in terms of content opportunity?

So 5G, the market has changed quite a bit and up to a couple of years ago. It looked like it was going to get pushed out. And definitely the standard is not quite stable and isn’t complete yet. So there is quite a bit of work that still needs to be done to come up with the standard. But what we have seen over the past two years is that all of the major players have shifted to a desire to be first and have significant market share. And we’re seeing them all race. And we believe that our 16-nanometer portfolio and our 20-nanometer portfolio is being used in close to 100% of the early prototypes, which are being used to demonstrate the technology. What we do expect is that the massive deployment will start in 2020 onward. And that the initial deployments, between now and 2020 are happening and they will - there are various flavors, 4.5, 4.9; there is a whole host of numbers which are being thrown around; pre-5G, 5G prototypes. But we are in a good position for all of those, but we can tell just based on technology position, but we do believe that the massive deployment will be 2020 and onward with the combination of programmable devices. And then for some of the companies they will definitely over time design ASICs to address that market. As it stabilizes and as the standards become clearer and the market requirements become clearer. Hopefully, that helps.

Got it. Thank you.

Hi, guys. I have one question and one housekeeping item as a follow-up. And the main question, pulling the wireless incomes back to a little bit of a near-term timeframe, generally, what gives you confidence both the wireless and the wireline business will grow both in the December and March quarters? They’ve been notoriously lumpy over the last couple of years. And so I’m a little surprised that you’re confident in sequential growth in both of those quarters. So any color you could give there would be helpful.

Yeah, I mean, the straightforward answer is on wired we actually have seen the design-wins, that I think we’ve alluded to several times in the past, actually beginning to ramp. Obviously, there is some aggregate macro CapEx uncertainty that could impact that. But underneath what’s going on in the industry that the products we’re in, OTN and GPON, we tend to see those ramp with infrastructure build-out internationally and domestically is happening. But you’re right, that can be lumpy. And wireless, again, I’ll go back to the answer I said earlier. We have actually visibility into our customers’ demand on us and their view on what they need to build out in particularly in China and in India on 4G and then the pre-5G business that Moshe referenced just a second ago internationally, North America included, that’s what we see.

Great. And then I guess just a one housekeeping one back to the share repurchase plan that you talked about. Just any color you could give, Lorenzo, on the percentage of the $3.5 billion in gross cash that you have that’s held onshore and what percentage of ongoing free cash flow is generated onshore?

So the way to answer that is of the $3.7 billion of gross cash, about two-thirds of it is categorized as permanently reinvested offshore. But about a third of the total has been deemed tax covered, I guess if you will, for the United States, so approximately two-thirds is available in the U.S. for use without book tax implications. It will be some cash tax implications on about a third of the total. Okay, and the mix of it as we generate it overtime, it’s a little dynamic.

Did you say two-thirds of it was offshore or one third was offshore?

Well, it’s - so there are two-thirds of it offshore, but two-thirds of the total are - so half of that two-thirds, so one-third of the total. We have basically provided tax provision for domestic purposes. So we could utilize that without having a U.S. tax book consequence.

Got it. Thank you.

Hi, thanks for taking my question. As we look at 4G, it appears that we’re in the kind of the latter stages of the rollout here in China. I think most of the operators in China forecasting 4G CapEx down over the next several years. And obviously in mature markets, it’s probably rolled out. So as we look forward to 5G, which likely is going to roll out in a few years in 2018, do you think you’ll be able to continue to grow your comps revenues over the next few years or are there other offsets to that?

Well, 4G indeed has peaked and is no longer growing. But there are still deployments that are happening on 4G. Even in China, there are deployments. There are waves of 4G deployments and they are still going through those. And whenever there are waves, you see some growth. In India now, we are seeing - it’s a smaller market, but we’re seeing - we’ve been fortunate to get the right design wins. And those are being deployed in major way in India. Our expectation is that there is a level of densification, which is happening even in 4G, part of which we’re benefitting from. And then, we are already benefitting from the 5G early deployment. But that tends to be not huge volumes, but nonetheless it’s nice revenue for us. So we expect - the takeaway on wireless is we expect it to continue to be choppy. What we are seeing is on wired communications that now is growing at the nice rate and that’s where our technology lead - and that tends to be a less volatile market anyway and that’s where the technology leadership, in particular, initially in 20 and now even in 16 is helping grow our business. So we do expect communications overall to grow. I don’t expect it to grow as a percentage of our business, generally speaking over the next few years. And it’s been fluctuating between 40% to 45%. And I don’t expect it to grow beyond the 45%. Neither do I expect it to shrink below the 40%. And that’s sort of the general elements. And from time-to-time, when there will be massive deployment cycles on wireless, it could sort of push it up, but on average, we don’t expect that to change.

Got it, thank you. And my follow-up question is on kind of the Joint Strike Fighter. It seems to be more of a measured ramp this time around versus Phase 1, which was a little bit more lumpy. What is giving you better visibility this time around? And can you talk about kind of the magnitude of the second tranche? And how far through the ramp of Phase 2 are we at this point?

So we have a little bit better visibility. I think we said before, but it bears repeating that for all of the elements of the second cycle or the cycle we are on now, from the technology insertions. We’ve won all the designs available to us. So that’s how we get this visibility. We do expect it, and actually I’m at a loss right now for the aggregate number, but we can get back to you, but this is going to be - our view is it’s going to be spread over, let’s call it, 10 years in terms of the buying patterns and the build out. And, again, we’ll see actually business from the new technologies. And we’ll also see some of the business from the older technologies in the form of spares, replacements and things like that. So our visibility, John, is based on the design wins we have and the discussions we’ve had with the government primes because they need to tell us when they are going to want them and expect to provide business to us. But it is not concentrated, right. That is the definite difference from what we saw in the last phase.

Great. Thank you.

Thanks for joining us today. We will have a playback of this call beginning at 5:00 PM Pacific Time, 8:00 PM Eastern Time today. For a copy of our earnings release, please visit our Investor Relations website. Our next earnings release date for the Third Quarter Fiscal Year 2017 will be Wednesday, January 25 after the market close. We will be attending the following conferences this quarter; the NASDAQ conference in London on November 29; the Credit Suisse Technology Conference in Scottsdale on November 30; and the Barclays Global Technology Conference in San Francisco on December 7. This completes our call. Thank you very much for your participation.