Good day, and welcome to the Altera Corporation Third Quarter 2014 Earnings Conference Call. Today's conference is being recorded. And at this time, I would like to turn the conference over to Scott Wylie. Please go ahead, sir.

Good afternoon. Thank you for joining this conference call, which will be available for replay telephonically and on Altera's website shortly after we conclude this afternoon. To listen to the webcast replay, please visit Altera's Investor Relations web page, where you will find complete instructions. The telephone replay will be available at (719) 457-0820, and use code 258712. During today's prepared remarks, we'll be making some forward-looking statements. In addition, management may make additional forward-looking statements in response to questions. In light of the Private Securities Litigation Reform Act, I would like to remind you that these statements must be considered in conjunction with the cautionary warnings that appear in our SEC filings. Investors are cautioned that all forward-looking statements in this call involve risks and uncertainty and that future events may differ from the statements made. For additional information, please refer to the company's Securities and Exchange Commission filings, which are posted on our website or available from the company without charge. With me today are John Daane, our CEO; and Ron Pasek, Chief Financial Officer. After Ron and John's initial remarks, we will take your questions. Prior to the Q&A session, the operator will be giving instructions on how you can access the conference call with your questions. I would now like to turn the call over to Ron. Ronald J. Pasek: Thank you, Scott. Sequential Q3 revenue growth at 2% was at the high end of our guidance range. Q3 revenue represents the sixth consecutive quarter of sequential revenue growth and is a 12% increase from Q3 prior year. 28-nanometer revenue approached $100 million, a 30% sequential growth rate, which indicates Q3 was another quarter of 28-nanometer share gains. Our 28-nanometer share for the quarter was well over 40%, which is higher than a 38% overall share against our direct competitor. New products grew 9% sequentially, 45% year-over-year and were 56% of total revenue. Mainstream products were down only slightly, and mature products down 10%. Lead times improved for nearly all nodes, with most of our products getting close to normal lead times. As a result, we saw customers trim their orders, which resulted in a book-to-bill less than 1. OpEx came in at the lower end of the expected range, which yielded an EPS number higher than consensus. Share repurchases remained substantial, as we spent $144 million to repurchase 4.2 million shares. We remain opportunistic with our quarterly repurchase activity, quite sensitive to share price. Again, we are on track to repurchase nearly $700 million this year, which, combined with the dividend, will far exceed our goal of returning 60% to 80% of cash flow from operating activities to shareholders. In Q4, we see revenue declining 2% to 6%, mainly due to softness in telecom and wireless and to a lesser extent, other. The softness in telecom and wireless is mainly telecom. Wireless is flat sequentially, which is stronger than we originally anticipated. As a result, gross margin for Q4 should be relatively flat to Q3, with vertical mix and intra-customer mix and wireless adding a bit of pressure to gross margin. 28-nanometer revenue will continue to ramp in Q4. Given the fourth quarter OpEx guidance for R&D of $111 million to $113 million, and SG&A of $78 million to $80 million, our OpEx for the year is as we expected, albeit with $2 million more in R&D and $2 million less in SG&A. As is our practice, we will, on December 10, issue a press release followed by a brief conference call to walk you through the elements of 2015 guidance. Now let me turn the call over to John. John P. Daane: Thank you, Ron. Our third quarter growth was driven by higher-than-expected revenue in the computer, consumer, networking and telecom markets. Wireless had a moderate decline in the quarter on lower TD-LTE China shipments. For the fourth quarter, we expect wireless to be flat sequentially as FTD-LTE shipments in China and 3G and 4G shipments in other geographies make up for the continued decline in China TD-LTE shipments. This is a positive development since 3 months ago, we expected wireless to decline. Overall, the forecasted Q4 decrease in revenue is due to the sequential declines in telecom, networking and some of our other category businesses. Computer should, again, grow as we continue to lead in the server acceleration market. In Q3, we released our MAX 10, very low end FPGA family and the customer acceptance and product ramp have been exceptional. We have now shipped products to several hundred customers. The Arria 10 midrange family continues its momentum with a record opportunity pipe. Our Stratix 10 product with the combination of Intel's 14-nanometer Tri-Gate process technology and our new HyperFlex architecture, provides high-end customers with unparalleled FPGA performance levels and to date, also has a record opportunity pipe. Our focus on improving customer engineering productivity continues to pay dividends, as we offer the only FPGA-based OpenCL compiler, and our Quartus II software platform has the industry's fastest compile times and performance. In summary, we are pleased with our likely 2014 outgrowth of the semiconductor industry and PLD industries and our corresponding market share gain. Now let me turn the call back to Scott.

We would now like to take questions. Please limit your questions to one at a time, so that we give as many callers as possible the opportunity to ask questions during the call. Operator, would you please provide instructions and poll for questions?

[Operator Instructions] At this time we will go to Romit Shah with Nomura Securities. Romit J. Shah: I wanted to dig into gross margin. Ron, I noticed that gross margins have -- since Q1 of 2012 have only been up twice, and it was a little bit light in Q3 and you're guiding it down a little bit in Q4. You mentioned mix. But just seeing this longer-term trend it's hard not to believe that there is an issue with either pricing or some other structural issue that's weighing on gross margin. Could you comment on that? Ronald J. Pasek: Yes. Romit, we typically explain gross margin on a quarterly basis and it's generally almost always just mix related. That's all you hear us talk about. And that's, in fact, how we started the year. We had a much more robust year in telecom and wireless in particular. So nothing's change in that respect. We'll give you guidance for next year on the call we do on December 10. Romit J. Shah: But seeing that your telecom business was down in the third quarter and down again in the fourth quarter, why would -- and that's lower-than-average gross margin, why would your overall profitability still be coming down? Ronald J. Pasek: So remember, Romit. Telecom wireless is 2 pieces wire and wireline -- wireless and wireline. Wireline for us is actually very healthy gross margin that's actually down in Q4. That's what's putting pressure on it. From the prior guidance we gave for Q4, quite a bit has changed. We guided originally wireless down in Q4, now it's flat. That's the biggest variance. We also have some intercustomer mix I talked about. And then in Q4 in particular, you have wireline weaker as I said, and industrial is a tad weaker as well. So... John P. Daane: I think if you go back to Q3 also, what you'd see is consumer was strong as we call that out, as a strong vertical and consumer like wireless tends to have lower gross margins from a vertical perspective. Romit J. Shah: Okay. If I could just one last question, John. You guys filed an 8-K, I think it was last week for a change of control provisions and the changing control provisions goes through a lot of the rank and file and relative to your comps, the provisions just seem to be a lot more generous. And I guess my interpretation was that, perhaps the company is for sale, but I wanted to, I guess, hear from you, how we should think about that filing? John P. Daane: Romit, so I'll answer from my perspective. So about -- I would say, 15 to 18 months ago, the compensation committee of the Board of Directors in combination with their independent compensation consultant, they do an annual survey of other like companies, particularly, those that we have in our case. There's about 22 companies that we say are like companies to Altera. And they found that a vast majority of those companies had both change in controls and severance agreements for their management. And they decided that they were going to go forth and put those into place. When they asked me, if whether we should do that I said, "I don't really care, I'm indifferent," if you were to ask me, "I'd say, don't bother because we really need them." But they decided they wanted to do it. Hasn't been our highest priority item. So it's taken us well over a year to do it. I think if you actually look at the agreements that we put in place, both on changing control and on severance, which are for the executive staff, which were me and my staff only. So I think there's probably 9 of us in total that have that. They actually are undermarket. So it's not a very aggressive either on the severance side or on the changing control side. So again, not anything that was a priority for us, not anything that the management here drove or -- and certainly, nothing in particular from an action going on in the industry that had us go do this. I hope that answers the question.

Our next question comes from Ambrish Srivastava with BMO.

Just had a question on the telecom and the wireline side. What's the reason behind wireline continuing to be weak? Is it also inventory correction or seems more like demand driven? John P. Daane: Yes. So if you look at our business, our -- both networking, which is in a separate category and telecom were up for us in the third calendar quarter, and we're forecasting both of those to be down in this quarter. It's interesting and that is in contrast to our competition and sort of seeing the opposite, which is they were down last quarter and they're expecting it to go up this quarter. And I think, therefore, if you look at the 2 of us it's probably more just program timing from particular customers, inventory from particular customers, some competitiveness perhaps of particular customers. But I don't think between the signals that I'm seeing from them and what we're seeing that there's any particular long-term trend here, per se.

Okay. And then my follow-up on the 14-nanometer. Is there any change to the schedule? I know couple of quarters ago, you had pushed it out by a quarter. Where are we on the tapeout schedule? John P. Daane: Yes, the tapeout schedule our goal remains to tapeout Q1. And we expect to have silicon still in the summer and we've ultimately achieved a number of very significant and major milestones towards this. The -- we're excited about this product. Its the combination of Intel's 14-nanometer Tri-Gate technology and a new FPGA architecture, HyperFlex, which really puts Stratix 10 into a performance category all alone in the high end. We really are literally, achieving 2x performance on a large number of designs we've engaged with, a very large number of opportunities. And the reason that we're so excited about this is again, if you go back and look at the FPGA industry, 50% of the revenue of the FPGA industry is in the high end. So we're going to be in a density class performance class, all to our own. And we think that will allow us to capture a vast majority of the high-end market as we roll that product out next year.

We'll move to Doug Freedman with RBC Capital Markets.

When we look at the revenue line, there's definitely quite a bit of movement in the different segments. But your performance isn't that far off of what I would call normal seasonal. Can you maybe talk about what you're seeing? And how we should think about next year, possibly shaping up from a seasonality standpoint, given that you do have some major programs that, I believe, are moving around? John P. Daane: I would say, in general, that it's very difficult. I guess we've had this question for many years and we've always said seasonal, you can get averages if you look at quarters, but it's -- there's such a heavy variation around that number that it's really something that cannot be used as a guidance or a guidepost when you look at the following year. We've seen carriers accelerate CapEx spending or decelerate CapEx spending. Obviously, communications industry being 50% of the PLD industry revenue, that can have an effect, GDP changes can have an effect. So there's many changes that can go on. Overall, I think what we've seen is some pretty healthy growth in our wireline this year. Now we're seeing some very heavy growth in computer, which we would expect to continue. Wireless has actually been a very pleasant surprise in the back half of the year because we did expect that the trail off. How that plays into next year is hard to predict at this point. So we're just going to stick with our guidance of one quarter at a time.

Thanks for that color. If I could for my follow-up sort of dig into a little bit of what some of the other questions have alluded to. If we look at your business model, it really has gone through quite a bit of cycle change, where you see your gross margins and operating margins really move on a pretty wide range, especially if you look over the last 7 to 10 years. Right now, it seems to be running at what I would call the lower end of your normal operating range. Are there anything outside of what you've already pointed to in terms of mix that you think you can do to drive your operating margins higher? And then the second part of that would be, just what is your thoughts on how you want to manage your balance sheet going forward? Ronald J. Pasek: Yes. So Doug, as I'd articulated last year at our investor meeting. You're going to see us continue to grow the topline, continue to stabilize margin and grow our OpEx for the next several years at a much slower rate than the top line, as we've done this year. So you have seen our op income this year if we, of course, hit the midpoint of Q4 guidance. It will increase from prior year. Our intent is to keep doing that. So with respect to the balance sheet, as you've seen, we've been doing a large repurchase this year. We're doing -- we've done 5 years of consecutive increase to the dividend. So we're definitely using the balance sheet to beef up the return of cash to shareholders.

This time we'll move to Vivek Arya with Bank of America Merrill Lynch.

John, I know it's probably a little hard to say but do you think that the PLD demand in China wireless could be higher, lower or about the same next year? John P. Daane: It's difficult to predict at this point, exactly where it's going to be. I think, generally, when you look at wireless deployments, what you see around the world is most of the carriers have a fairly gradual deployment of their systems. And so their CapEx spend remains fairly stable. China Mobile, from what we've seen over the last 5 years, tends to be far more choppy. They do deployments, they do them in phases. And they may do 6 or 9 months of very heavy deployment and then takes 6 or 9 months off. And so as such, they can cause the number to, both from a revenue perspective as well as a shipment perspective, to go up and down a little bit. What we've seen this year is China Mobile do a lot of spend, it's ramping down right now. They are expected to ramp up another phase next year. I'm not exactly sure when that's going to happen. What we are seeing though, is where China Mobile is really ramping down right now and has -- as it did last quarter, we're seeing a lot of other things fill in for that gap. So we're seeing FTD shipments to China Telecom, in particular, pickup, and then we're also seeing shipments to other countries both 3G and 4G North America, as an example, India has picked up, Europe has picked up, and that's also filling in for the natural decline that you would see from China Mobile. How this plays out each quarter again, because China Mobile can be choppy, it's a little bit unknown yet. But overall, if you were to take a step back and say, "Where are we in the overall deployment of the LTE?" It's still very, very early innings. We've seen some deployment in China and the U.S., and Japan and South Korea, it's just starting in Europe. Again, India just turned on for 3G, 4G. We're seeing some other developing countries like Brazil, start to spend on their 3G network. So I think you could see, in general, the spending do well for many years. How it behaves on a quarterly basis, though, is going to be a question mark, simply because of China Mobile.

Got it. And in the first half, there was a concern that perhaps there was an overshipment of FPGAs relative to other basestation component. Is that still the case? Or do you think that there are no other component shortages? And you're supplying to demand and there is no over or under supply of FPGAs for LTE-based stations? John P. Daane: So in general, what has gated the overall deployment has been power amplifiers. I think, that's very well written and very well known. If they remain, as far as we can see, on allocation today, that will probably gate your top-end shipments. Because generally, customers will want to get together. All of the components so that they can rapidly assemble and deploy systems and they're not holding inventory. We have not seen the decline in wireless FPGAs spend that either of our competitors saw last quarter and this quarter. So I can't really tell you why they saw what they saw, and I don't really have an explanation for it. From our perspective, the revenue is really gated by power amplifiers and it has continued, albeit it has changed every quarter. It's continued to be fairly, fairly healthy.

Got it. And then one last one if I may, according to Xilinx, and I'm sort of reading from their transcript here. At the 20-nanometer node, independent customer feedback indicates that they have a 1-year lead over the competition in terms of silicon maturity. I assume you do not share that view, and if you could just give a sense of how you see the 20-nanometer share playing out. John P. Daane: Yes, I think on 20-nanometer, we're doing extremely well with our product line. We put out our software about 4 to 6 months earlier. We have a number of customer engagements, as far as we can tell we're shipping more broadly to customers at this point. So we think we're going to do very well in the 20-nanometer node versus our competition. Now remember, we are not doing a high-end 20-nanometer family as they are, so we're doing a midrange family. So long term, could they have more 20-nanometer market share? Possible because again, they've got more designs that they'll have out. What we're doing is we're really saying high end at 20-nanometer using a planar transistor does not make a lot of sense. The power consumption will be far too high. We're skipping that and we're going to 14-nanometer. So we'll be a generation ahead of our competition next year in terms of process technology and both on -- when you look at the features that you derive from that, in terms of performance, power and density, it makes a lot more sense to be doing what we're doing, particularly when you look at the return for the R&D investment. Again, skipping, doing a high-end 20 and moving directly to 14 allows us to ultimately reduce some of the 20 spending, move to 14, be a generation ahead and get a lot more bang for the buck in terms of our overall R&D performance. So we think we're in very good stead midrange to midrange in 20-nanometer. And again, if you take a step back, this really makes up for what was a less competitive midrange 28-nanometer family.

We'll go to Blayne Curtis with Barclays.

Just wanted to follow back up on the gross margin question. If you look at September, I guess, gross margin came in a little lighter, wireless was down. And then I guess, just longer term here, you're below what you've laid out for more long-term target. So I would assume if this China spend does happen, the mix should shift back to wireless, so that actually should be a headwind. So just kind of your thoughts on where you are today and if you still can get back to those levels. Ronald J. Pasek: Yes. I think, what -- if you look at the growth rates of some of our higher-margin verticals, they will continue to grow the next several years that will naturally have mix in our favor. So that part of why we're going to stick to the 67 to 70 range. I don't view that as unachievable at all. John P. Daane: I think you have to remember also that like wireless, consumer is a lower-margin component of our business. And consumer was up very strongly in the third calendar quarter and that business continues fairly, strongly into this quarter as well. So there are vertical mixes, as Ron has pointed out, it can even be within verticals. Wireless is an example, basestations tend to have some higher gross margin than radios just because of the volume differences. Complexity differences of the products and therefore, it's a fairly complex equation in order to compute.

And same question. Just -- there's a lot of moving parts in the telecom wireless segment into December. I'm just curious, just more broadly in your industrial auto businesses, just have you -- what have you seen, I mean, in the terms of the environment? There's been a lot of mix debate as to what's going on, just kind of curious to your views as your broader business. So you're guiding it down, I'm assuming that's seasonal but are you seeing anything better or worse? John P. Daane: So in general, what we would say at a high level is our visibility is lower going forward than it was before, and that really is because as our lead times have come back to normal levels, customers do not have to place backlog as long as they were before. And therefore, we don't have visibility out as long as we did a quarter ago. Generally, I would say, we don't see any major weakness in any particular market, vertical or end market per se. There are puts and takes of particular customers. There are some changes in geographical things. But I don't know that I'd stand here and say, for instance, that it's really a bad environment. I wouldn't stand here and say, it's a huge growth environment. I think it's been a fairly steady environment all year, how that behaves next year we'll have to see.

We'll take a question from Joe Moore with Morgan Stanley.

You had alluded to the strength in computer as being a function of, at least, partly server acceleration. Can you help us now that that's kind of getting going. Can you talk about -- help us size that opportunity and is the fact that Intel has kind of endorsed that been a driving factor for adoption there? John P. Daane: I think the -- I would guess, and this is my guess and underline the word guess. Intel's announcement is really more of a -- an understanding of the benefit, which probably was fed back by the customer base. So I would say to a large degree, my view is the customer base is driving this. The customer base realizes the advantage of using FPGAs for acceleration within a lot of data center applications from a power performance perspective. We've done announcements with Baidu and with Microsoft, and have many other companies working with us. We invested in OpenCL quite early, and that is necessary component to have server customers utilize FPGAs. And so we've been really, I think, the lead company in this particular space. If you just look at what other components have been used for acceleration to date, it's predominantly been GPUs. We think there's a couple of hundred million dollars worth of that sort of business that can be replaced. The reasons that we are a better component than GPUs is our power consumption is about an order of magnitude lower than a GPU for the same performance levels of acceleration. And the benefit there obviously, for a customer is their #1 spend in the data center is on power, not the equipment itself. So we think overall, it's at least starting a couple of hundred million. But as we see the acceptance of this grow and new applications where they're skipping directly to an FPGA as their companionship for acceleration that it should grow well beyond that. We've estimated this could be long term $1 billion opportunity.

Okay. And then separately, the fact that you're getting your lead times down, you said it had some bookings impact. Have you seen any revenue impact from that? Have you seen -- I mean, obviously, there's correction in this quarter. But is that related to the lead times in our -- is there any kind of cancellation activity of the longer-dated backlog? Ronald J. Pasek: It hasn't affected revenue at all. No, and the cancellations and reschedules are pretty normal. So what -- to John's earlier point, we've talked about this before. Typically, when your lead times start to extend, the behavior by customers is to place orders out beyond the current quarter. And so likewise, when your lead times come in, they stop placing orders out as far. So you do lose some visibility. But then they have much better performance from a shipment standpoint.

This time we'll move to question Tristan Gerra from Baird.

Just looking at your December quarter guidance, given that China Mobile raised their target by 200,000 units, if we assume $300 average content, that will translate into about 6% of your September quarter revenue, in order to expect wireless would now track flat versus a prior expectation, and say decline for [indiscernible] that really make up for a 6% difference in your December quarter revenue outlook and if not, what will be the offset? Do you expect not to ship everything by end of this year? Or is there a weakness in other end markets that will represent a significant offset to that? John P. Daane: I would say, Tristan, the peak of the shipments into TD for China, which is really China Mobile, peaked probably in Q2 and has come down somewhat. Overall, shipments in general are gated by power amplifiers supply, which has been on allocation all year. I think to understand that equation, which is probably important is there are twice as many power amplifiers used in TD than FTD. And since there's more TD as a percentage this year than probably there will be next year, there isn't necessarily an incentive for a lot of the power amplifier companies to build out a lot of capacity, since they know that simply the shift in mix may mean that they have more than enough next year. But right now, they remain something that gates the overall shipments and what's happening in the industry. How long does that last is something that I cannot answer. Because we've been unable to get a clear picture from our major customers as to when they expect the allocation to end. So right now, we have seen TD pull down. We've seen FTD for China pull up, and we've seen, again, shipments in the other geographies for 3G and 4G pick up as well.

Okay. And a quick follow-up. On the last earnings call, [indiscernible] talked about high silicon planning on using their ARM architecture, at the 16-nanometer FINFET node by 2016. Does that materially change FPGA positioning into wireless infrastructure, given high silicon [indiscernible] going to basestation. John P. Daane: I'm sorry, Tristan, we lost some of that. But let me probably answer, I think, what you may have been implying. We do see in our industry, ASIC replacements that happen from time to time. As the cost to implement an ASIC continues to go up with every new node, the number of ASICs implemented continues to decrease. And so 15 years ago, we might have seen hundreds of ASIC conversions a year, we see very few, if any, now that happened generally. And it really will only be in very high-volume markets. Because again, you have to look at the economics of spinning your own chip and putting all the engineering time into that plus the software mask and wafer cost versus just using a PLD. And so what's happened is if conversions happen and they do happen from time-to-time to all of us in the industry, generally, we absorb those in the year because where we may have a few conversions that happen, we literally will still have hundreds of people who used to use an ASIC pick up and start using PLDs for applications. And so we have far more business that flows to us than escapes. And again, I use this sort of the guidepost that when I started here 14 years ago, the conversion point was roughly about 10,000 units a year. Where it was more economical to move to an ASIC than a PLD. Now we see that in the low millions of units per year. And so generally, when you look at the industry there are very few applications that have the volume to justify an ASIC conversion in the businesses that we're in. Because again, we're not in handsets, we're not in PCs, we're not in wearables. And those types of applications are the few in the industry left. So if I were to mention Huawei specifically, I don't see anything out in front of us that would say, that there's an ASIC conversion that I know that's going on, that would hit us going into next year.

At this time will take a question from Christopher Rolland with FBR Capital Markets.

Just first a clarification. You mentioned FD-LTE, it sounded like there was an uptick there. Was there any FD in China at all? And then Mike, I guess, my real question, so a little bit more detail I need to touch on it, on the networking computing storage side. So nice quarter there, but perhaps a little bit more on some of the drivers. Is it just the hyper scale guys? And why the downtick in 3Q? Is it just off of strength in 4Q? Sorry, 4Q is it off of strength in 3Q? John P. Daane: Okay. Let me see if I can get all that in. So on networking, strong quarter last quarter is coming down this quarter. I think that's just more of timing and customers than it is anything else. For computer, computer was up last quarter. We expect it to be up this quarter. I think the primary driver within computer right now is in the data center side. If I were to look last year, one of the other drivers was on storage for flash, for storage arrays. I would say, this year it's far more onto the computer side itself. And then I'm sorry, I didn't get the other question.

There was -- sorry about all those. There was FD in China. John P. Daane: Yes. So if I look at FD in China, absolutely, we've seen FTD pick up in China, particularly for China Telecom. A little bit for China Unicom as well. And that combined with FD shipments in other geographies -- and 3G shipments in other geographies has made up for the decline, or I would say, temporary decline. Because China Mobile will go to through another phase in the future. But certainly, decline this quarter in China Mobile TD shipments.

Great. Thank you for the color and detail there. One last question, just on OpenCL, I think we're on to about 2.0 late last year. Do you think that that's finally kicking in here, is that part of what we're seeing there? And how robust is that? Do you think that if it's a little bit more robust, we could have -- that could spur an option and inflection further? John P. Daane: Well, I think certainly, without it, you haven't solved the programming solution. Our customers will implement a design in FPGA using RTO, register transfer language, open -- it's used for ASICs. It's used for ASSPs. It's used for FPGAs. Only hardware engineers understand how to do that and there aren't as many hardware engineers as there are software engineers. Our customer, the customer base in computer, the people that own data centers are not going to hire a lot of hardware engineers to program the FPGA. They want to do it in C, like they do the CPU. And so having an industry standard language, it's got to be a parallel processing language because you're essentially using the FPGA or in the past, the GPUs accelerator off of the CPU. So it's another series of course you can think of that are inside of the FPGA that are doing acceleration, so it's parallel processing. OpenCL is the only industry standard language. That's why we did the compiler. And that's giving us a distinct advantage in the market. I think it is being broadly adopted right now and is a necessary requirement for the customer base in the data center area long term for this particular business. So I do think adoption is doing very well in that space and think it will continue. And it's not just by us, it's also by all of the CPU vendors and GPU vendors, right. So you're going to see that Nvidia and AMD, as an example and GPUs have OpenCL compilers, as well as companies like IBM and Intel. So it's a fairly broad technology at this point.

We'll take the next question from Srini Pajjuri with CLSA Securities.

I guess, Ron, a couple -- last couple of quarters, you've been saying that your lead times have extended, and you actually give us a guidance for wireless even for the out quarter. Can you talk about where your lead times are and do you have visibility beyond Q4 as it stand? Ronald J. Pasek: Yes. So our lead times have come down quite a bit and I actually said this wold probably happen through the year we would catch up and we have. Most of them have actually improved from Q3 -- in Q3. And in fact, most are close to normal. So as a result, what you don't get is that elongated pipeline with visibility to the following quarter at all. So no, we've really lost that visibility as a result of lead times getting closer to normal.

Okay. And then John, you mentioned FDD is starting to happen. Can you maybe quantify the opportunity there relative to TD? How big do you think FD is going to be in terms of maybe basestations or even if there's a difference in terms of your own content, if there's difference in FPGA content between TD and FD? John P. Daane: So in general, I wouldn't say that there's necessarily a significant difference in content TD versus FTD. Remember, FTD is shipped in the U.S. as the predominant standard. It's probably you'd view the predominant standard in Japan. Different countries, different carriers will do different things. But there really are 2 systems that are being deployed for 4G. One is TD and the other is FD. And again, carriers choose and they ramp up and ramp down at different points in time. But overall, as we mentioned earlier, we do think we're in a very early innings overall, of the overall LTE total shipments to the world.

We'll take a question from John Pitzer with Crédit Suisse.

This is Ryan Carver in for John. I guess its my turn with the dead horse. So if I think about the FTD coming on sooner than expected in boosting the wireless business from down to flat, if I sort of think of China Mobile, I think, your competitors talked about a potential reacceleration of the first half of next year. So if I think sort of FTD is happening potentially sooner than expected and then a reacceleration of China Mobile in the first of next year. What's to prevent or I guess, what are your thoughts about the second half of next year and the potential for all 3 of those carriers decelerating their deployment plans at the same time versus maybe prior expectations for FDD to perhaps carry a bit more of the weight in the back half of the year? John P. Daane: Well, it's difficult to predict exactly how the rollout is going to happen worldwide. I mean, the only thing you can learn from history is that China Mobile tends to be more choppy. If you go back to the TD-SCDMA deployments that happened over a series of years, they would do those in phases. Again, buy for 6 months, 9 months and then stop deploying for a period of time to digest what they've put in place. Mostly other carriers seem to be far more steady in terms of what they deploy. And so based on that, it's a little to hard to predict in the future exactly how deployments are going to go. And obviously it's been somewhat gated by the overall supply of power amplifiers. We haven't seen the downdraft in revenue and wireless that our competition has. So when they're talking about a coming back up, that maybe a little bit of a different phenomena from us. Because we haven't seen our business in wireline -- wireless, excuse me, declined significantly over the year. So I can't really add any color to what happened there or what's causing it, or why their business would change going out into the future quarters. And so, yes, I do think, in general, in FTD, obviously in China FTD is early and will continue to grow as both China unicom, and telecom, and particularly telecom try to catch up with China Mobile's deployment on LTE. And then we've also seen FTD pick up in North America again, in terms of shipments and starting in Europe. So I think we're very early. Again, I can't predict how it's going to play out on a quarterly basis, but we do feel that wireless is a good business for many years.

Got it. And then as my a follow-up question, you guys have commented about OpEx slowing over -- going out in time. Obviously, you're bifurcating the nodes in terms of high-end going to '14, and the mid- and low-end staying, it sort of the legacy nodes. What are the potential impacts of doing that, especially at the mid- to low-range as the ASIC and maybe specifically, the ASSP products catch up from a process node perspective and are able to offer sort of an as good or better performance option on like-for-like nodes. And, I guess, maybe specific to 20-nanometer, incumbency is important to think about the business are there design windows that have -- that you kind of know of that made going to 20 going right to 14 more advantageous from a design win perspective? I guess, what gives you confidence that when you bring 14-nanometer to market, you'll be hitting design windows that will allow you to take advantage of that kind of market advantage? John P. Daane: So there was a GSA, which is the Global Semiconductor Alliance panel that happened a few weeks ago. And we're one of the major ASSP companies, their CEO was on the panel and he said that they're not going to aggressively move to new process nodes. They're staying at 40. They're not moving to 28. They're not moving to 20. And they're not moving to 14. And the reason is they can't get the cost reduction moving forward in new nodes. So to a degree, what's happened is because harder to achieve the silicon cost reduction for their small designs, in particular cases, where they have analog components. Analog does not shrink generation to generation. They're finding that the cost reduction isn't there, and the investments keep getting higher and higher. So ultimately, we're seeing fewer designs done at newer nodes. It make sense for us to move forward to newer process nodes because we can increase the density and decrease our cost basis still, which then allows us to replace more of this business in ASICs and ASSPs, where people simply can afford to move forward in new processes. And therefore, their cost basis is not going down and it opens up an even larger opportunity for us. So I don't -- I think the dynamics that we've talked about for the last 10 to 15 years in terms of fewer ASICs, fewer ASSPs anybody in the infrastructure side that doesn't have the volume really needing the use of programmable, those are absolutely intact. And I would say, there's even an acceleration of that. I mean, even volume customers in the last couple of months we've had a number of very high-volume customers coming into us, looking to move from an ASIC to an FPGA, simply because they can't afford to implement ASICs any longer. So I'm very confident that we're in a very good space and happy to be here rather than trying to figure out what I do because the technology is working against me. In this case, Moore's Law works for us.

Got it. Real quick if I can. 20-nanometer revenue in the quarter? John P. Daane: 20-nanometer, it's -- what we'll do is when we get to about $1 million in a quarter of shipments, we call it out and then we give you percentages going forward. It's been something, I think, we've done for 5, 6, 10 years, somewhere in there. And we haven't reach $1 million yet. So as soon as we do, we'll let you know.

Sorry, I mean 28? John P. Daane: Oh, 28 is... Ronald J. Pasek: 28 was approaching $100 million. It was about $94 million for the quarter.

And we'll move to William Stein with SunTrust.

This is Joe Meares filling in for Will. I know it's kind of a small portion of your -- of the company revenues. But I'm wondering if you could just tell us how automotive dividend in the industrial, military, and auto end market in Q3, and how you're expecting that to do within the Q4 flat guidance for that end market? John P. Daane: There's nothing really major that happened with that market from an industrial or automotive or military in Q3. It was all kind of moving well, moving sideways sort of numbers, nothing specific to point out.

All right. And then any puts or takes there for autos going into Q4? Or kind of the same ... John P. Daane: Nothing that's going to be meaningful. One way or the other when you look at our numbers. I think, again, what we try to do is call out, including some of the segments that are in the other business, which we haven't typically talked about. Whenever there's a major mover in any quarter, we'll call those out. The rest are slightly up, slightly down, flat. But nothing of significance.

Next question will come from Parker Paulin with Wells Fargo Securities.

Hoping that you might be talk a little bit more about how your relationship with Intel has been developing, and if you've seen any changes in that over the last quarter or so? John P. Daane: I think our relationship with Intel has been very good. All of our suppliers have been excellent in terms of information flow, strength of the technology, relationships. We are a different business, obviously, than some others. And we do what we call semicustom designs, some smallcustoms, some standard sale. And so we need a lot more information in order to do our full custom work than your average ASIC or ASSP company. And so what we found is cooperation and information flow has been excellent. So far so good and I think, if you look at what they've announced in terms of their schedule, they're going to be in very heavy volume production well before we prototype our first unit. So this will be a case of using a very mature technology from a timeline perspective. And so everything is going good.

And we'll move to Gust Richard with Northland. Auguste P. Richard: John, back in the day when you went from 2G to 3G, your content and basestations doubled. I was wondering if you can give a little bit of color as you go from 3G to 4G, let's call it, in backhaul basestation and radio head, how do you see that -- your dollar content shifting as a percentage, if you will? John P. Daane: We haven't really looked at this in a long period of time. So I don't know if I can comment on specific numbers. In terms of architectures, I would say there was a transition in wireless a couple of years ago, where we saw in radios, some people that used predominantly an FPGA-based system moved to an ASIC plus an FPGA, which has tended to be what you see as a general architecture around the industry. So that really hasn't changed, I would say in the last 3 years. And on the basestation side, I would say the architecture in terms of the types of components that are in the systems, generally haven't changed 3G to 4G over the last several years. What has changed is with the complexities of going from 3G to 4G generally, the content that they need, in other words, the size of the FPGAs has gone up. And generally, they will always be at the leading edge for the performance in power. As to what that's done with the dollar content, we really haven't gone back to refresh that, so I don't know that I could give you an accurate number. But again, I would say with every generation, there's nothing that's really changed in architecture the last couple of years. And our -- because of the complexity, I would say, they're using every generation larger FPGAs from us. Both in the radios and in the basestations.

And we'll take a question from Alex Gauna with JMP Securities.

I know you commented on things being generally flat in industrial aerospace and defense. I'm curious your key rival in FPGA saw a much stronger surge in that category over the last quarter, in a very fundamental differences in your positioning that would explain that? And I'm also wondering if you saw any ripples from orders in consequence of the Russian sanctions that are going in place and does that create any [indiscernible] for you in some of the out quarters? John P. Daane: So to answer that, I guess let me start with the first one. If you kind of go back in time, remember that Altera in the '90s exited the military business. And so what we both categorize as the older products, mature products are -- Xilinx is going to have a much, much larger share in military, for that matter Lattice and Microsemi would than us. Simply, because we weren't a player within the defense military side of the business for those older or mature products. We really had no play at that time. So I'd say generally Xilinx is going to have, I would estimate of the PLD industry a vast majority of the revenue that's there. The other thing is we're not doing the obsolescence program on technology, which tends to accelerate revenues. So that may be why you see 2 differences between us and them, in terms of that segment. There are probably, as you pointed out, and we talked about there are 3 differences between us. We had networking and telecom up in Q3. Theirs was down. Theirs is up this quarter, ours is down this quarter. That's probably just timing. We haven't seen the downdraft in wireless. And then for us, we just don't enjoy the older military business and we certainly from an obsolescence perspective, aren't doing anything that pulls in revenue there. So those would probably be my 3 points that I could make it. And again, there is well, you look at these companies and you assume they'd be very similar in terms of, you both sell PLDs and or all of these sell PLDs and all of you have -- you are in all of these accounts. There's a large difference and it really is almost impossible for me to guess exactly what's happening at any particular timeline with any of our competitors. Russia, I'm sorry. So, yes, we did see some pull in for Russia. This was because in industrial and this is because as the EU changed its law in terms of allowing companies to ship, really the way it did that is it took away companies bulk licenses to ship equipment in and you need to get a company-by-company license. And so you saw some acceleration in industrial for shipments into Russia. But I don't know that I call that out as tremendously meaningful to what's going on in Q3, Q4 from our overall business.

Okay. And also I wanted to ask TSM now has recently announced to pull in for their 10-nanometer FINFET. I'm wondering what that might mean for you and what you think about what appeared to be accelerating timelines across the board for FINFET resources that you can take advantage of? John P. Daane: Yes, I think you're seeing certainly, TSMC, Samsung and Intel all be very aggressive with their roadmaps and their investments, recognizing that there is a class of customers such as ourselves, processors, GPUs, baseband chipsets, they are very interested in advanced technology nodes. And it does allow a lot of opportunity. Far too early for us to discuss what our plans are and what we're doing right now for 10-nanometer.

Next question will be from Harlan Sur with JPMorgan.

I apologize if I missed this, but in your storage subsegment, how did that trend for you in the September quarter and maybe your expectations kind of just directionally for the December quarter, and maybe just some discussion about kind of major drivers up or down within that subsegment? John P. Daane: So the computer and storage area was up in Q3. Sequentially, it's going to be up again in Q4. Major driver is on the server acceleration piece.

Great. And then obviously, there's been a lot of questions around wireless in China. If I look at your non-wireless business in China, more of the broader base, kind of industrial base markets, how did this segment trend this September quarter, and maybe your expectations here in the December quarter? John P. Daane: Yes. China for us and industrial has been extremely strong. It's up very significantly in the September quarter year-over-year. And so that business has been doing very, very well for us.

We'll take a question from John Pitzer with Crédit Suisse.

Just a quick follow-up on your comments about the complexities in the movement from 3G to 4G. I think that Xilinx is on record are saying that their major traction with 28-nanometer at the basestations is really more at the Kintex, at the midrange of the product stack. So maybe can you just refresh us the target areas for where you guys are going on the TD and FD-LTE and where that content growth you guys maybe seeing versus Xilinx, which is shipping more midrange than the high-end stuff? John P. Daane: So we're shipping more high-end than low-end in our mix. We weren't as successful in the midrange. We do have some midrange business, but not as much as our competition would have. So if you look at the basestation, generally basestations use high-end. Radios will use midrange or low-end. So if you look at us, I would say more of our wireless business is high-end, low-end, some midrange as a mix. I don't know if that...

At this time, we have no further questions in the queue.

Great. Thank you, operator. As to -- as we're wrapping up this afternoon, a few final items. As to conferences this quarter, we will attend the Wells Fargo GMP Conference in New York on November 12. And in December on the 3rd, we'll present at the Credit Suisse Technology Conference in Phoenix. Lastly, a reminder as Ron mentioned, we will issue a 2015 guidance press release on December 10 with a conference call to follow. This concludes Altera's Earnings Conference Call. Thanks for your interest and participation.

Once again, this does conclude today's conference call. Thank you for your participation.