Inphi Corp
Q3 2015 Earnings Call Transcript

Published: 28 Oct 2015

[Operator:]

Good day, ladies and gentlemen, and welcome to the Inphi Corporation Third Quarter 2015 Earnings Conference Call. At this time, all participants are in a listen-only mode. Later, we will conduct a question-and-answer session and instructions will be given at that time. As a reminder, this call is being recorded. I would now like to introduce your host for today's conference, Deborah Stapleton. Please go ahead ma'am.

[Deborah Stapleton:]

Thank you and good afternoon, everyone. We appreciate you joining us today to discuss the financial results for the third quarter of 2015. With me today are John Edmunds, Chief Financial Officer; and Ford Tamer, our Chief Executive Officer. John will begin with the Safe Harbor, then Ford will give you an overview of our business. After that, John will provide a financial summary of Q2 and the outlook for the fourth quarter of 2015. Then of course we'll be happy to take your questions. I'll now turn it over to John.

[John S. Edmunds:]

Thanks, Deborah. Please note that during the course of this conference call, we may make projections or other forward-looking statements. These forward-looking statements and all other statements made on this call, which are not historical facts, are subject to a number of risks and uncertainties that may cause actual results to differ materially. These forward-looking statements speak only as of today's call. We do not undertake any obligation to provide updates after this conference call. For further information regarding risk factors for our business, please refer to our Registration Statements as well as our most recent annual and quarterly reports on Forms 10-K and 10-Q, all filed with the Securities and Exchange Commission, accessible at www.sec.gov. Please refer, in particular, to the sections entitled Risk Factors. We encourage you to read these documents. Also during the course of this conference call, we may make reference to non-GAAP financial information. A reconciliation of this information is included in the press release and on our website, which is available at www.inphi.com. This information is not a substitute for GAAP and should only be used to evaluate the company's results in conjunction with corresponding GAAP measures. Now, to begin our review of the quarter, let me turn the call over to our CEO, Ford Tamer. Ford?

[Ford G. Tamer:]

Thanks, John. Good afternoon, everyone, and thank you for joining us for our Q3 2015 earnings call. Q3 was another very solid quarter for Inphi, and we are pleased to deliver these results to our shareholders. We are announcing record non-GAAP revenue of $62.4 million for the quarter, which represents 72% year-over-year and 3% sequential increases respectively. Our non-GAAP gross margin remained strong at 68.4%. We are also announcing record non-GAAP operating margin of $12.7 million or 20.3% and non-GAAP EPS of $0.25, which is a 108% increase over our $0.12 EPS one year ago. Our free cash flow this quarter was a record $14.2 million, bringing our overall cash and short-term investment balance to $101.7 million. On the product and customer fronts, this was a busy quarter for Inphi. To walk you through the events of the quarter, let me return to our familiar FedEx, planes, trains, and trucks analogy to describe Inphi's data movement interconnect progress. Starting with planes, our optical interconnects for the long-haul and metro market. In late September, we attended the annual European Conference on Optical Communications or ECOC show. There, we announced the sampling of a new surface mount linear driver. This is an adaptation of our very successful quad linear driver in a more cost effective packaged, designed for higher volume, 100 Gigabit and 200 Gigabit coherent applications. We developed this surface mount package to meet our long-haul and metro customers' significant existing need for higher density and lower cost solutions. We are also seeing increased customer traction with our new 45 Gigabaud optical interconnect product family. We are confident that both the surface mount and the 45 Gigabaud Amplifier and Driver product will contribute to our revenue growth in 2016. Moving onto trains. Our networking interconnects within the data center. Also at ECOC, we showcased the world's first, 40-Gigabit, 50-Gigabit, 100-Gigabit, and 400-Gigabit Ethernet PAM4 Solutions for cloud interconnect. By doubling the bits transmitted at the same baud rate, PAM4 is now recognized as the preferred modulation method for more efficient, optical, and copper interconnects, both within the data center and between data centers. Inphi's PAM4 100-Gigabit dual lambda technology was demonstrated at ECOC and is ready and available for designers to build line cards and modules today for next-generation data center platforms of tomorrow. We remain confident that our networking interconnect solutions will continue to ramp in 2016, driven by our 10-Gigabit and 40-Gigabit physical layer devices, our clock/data recovery or CDR and the many PAM4 designs currently underway for 40-Gigabit, 50-Gigabit, 100-Gigabit, and 400-Gigabit. And finally, the portion of our business that correlates to trucks is our memory interconnects inside server and systems. There, we remain ready and well position to regain share. Specifically, this quarter, we announced sampling of the second-generation of DDR4 registers and buffers, which consume less power, while allowing DRAM to run up to 3200 mega-transfers per second with increased capacity. Because of this increased performance and reliability, we anticipate strong market acceptance, continued market share gains, and revenue ramp in the first half of 2016 and continuing through mid-2017. While announcements are key indicators of Inphi's progress, what really matters is how our customers react. Our module, system OEM, and data center customers expressed robust interest and demand in our new optical, networking and memory interconnect products. With that, you can see that in Q3, our planes, trains, and trucks continue to roll along on track, delivering data faster and further. As we progress through the fourth quarter and look ahead, I am pleased with our product positioning in the various markets. We remain confident in our 2016 outlook, driven by a resumption of the business in China and our multiple new product introductions. I offer a sincere thank you to the strong team here at Inphi that is working 24x7 to make all this possible. Now let me turn the call over to John for a discussion of the financial results. And I look forward to addressing any questions you might have at the end of this call. John?

[John S. Edmunds:]

Thanks, Ford. Now, let me recap the financial results for Q3. As Ford mentioned, in the third quarter of 2015, Inphi reported revenues of $62.4 million, which was up year-over-year by approximately 72%. The nine months year-to-date revenue was $182.6 million on a non-GAAP basis, which also represented 80% year-over-year growth. While this growth was due in part to the October 2014 acquisition of Cortina, even when we add the Cortina 10 and 40 Gig SerDes business for the first nine months of the year into the base, the core communications business continued to show approximately 47% organic year-over-year growth, and we expect that growth to continue through Q4 of this year. We define our core communications as Amplifiers and Drivers as well as 10 Gig, 40 Gig, and 100 Gig physical interface products. Our communications business continued to represent roughly 80% of our total revenues through Q3. The core communications products continue to represent 65% to 70% of communications, while the legacy communications products represented the remaining 30% to 35%. As was the case in Q2, in Q3, we continue to see some strength in transport and legacy revenues relative to expectations. As expected in Q3, sales of product into high-speed memory applications were up sequentially as sales of DDR4 products grew on top of sales of DDR3, which remained relatively strong. We expect sales of our DDR4 products to keep growing into Q4. This should bring us to revenues for the year in high-speed memory in the range of $55 million to $58 million. Gross margins on a non-GAAP basis came in at 68.4%, slightly ahead of where we expected, and including approximately $300,000 of IP licensing revenue in the quarter. Based on the current forecasted mix for Q4 and an ongoing stream of IT licensing revenue, we expect the gross margins in Q4 to be in the range of 68.5% to 68.7%. On a GAAP basis, we were closer to breakeven in GAAP operating income in Q3, now at a $200,000 loss, as compared to Q2 when we had a $2 million operating loss. The delta between GAAP and non-GAAP is approximately $11.5 million, made up of five basic adjustments. As discussed previously, the GAAP loss includes stock compensation expense of $7.2 million, purchase accounting related adjustments totaling $3.8 million, as well as some acquisition transition expenses, which were short-term retention bonuses of $1.8 million that were part of the purchase agreement. Finally, we also had related tax effects to these additional expenses along with valuation allowance adjustments and other tax charges totaling a $1.3 million tax benefit, all netting to $11.5 million of additional expense in the GAAP books. The Q3 GAAP-based result, a loss of $1.1 million or $0.03 per share, compares to a net loss of $6.9 million or $0.22 per share in Q3 of 2014. The Q3 2014 GAAP net loss was negative more as a result of inter-period tax allocation between the quarters throughout 2014. A more telling comparison was the GAAP pre-tax loss of a year ago of $1.8 million which now compares to Q3 of 2015 GAAP pre-tax loss of just $130,000. The more current Q3 2015 number also included $5.6 million of largely purchase accounting adjustments related to the acquisition of Cortina. In fact, this is what we would see on a non-GAAP basis, about a $6.5 million improvement in net income from the third quarter of 2014 and $3.9 million through the third quarter of 2015 at $10.4 million. Now for further clarification and analysis of the operating results, let's turn to some additional non-GAAP measures and comparisons. As I just said, on a non-GAAP basis, the net income for the third quarter of 2015 was $10.4 million or $0.25 per diluted share, which was up slightly and consistent with the non-GAAP net income of $9.9 million or $0.24 per share in Q2 of 2015. Non-GAAP operating expense for the quarter totaled $30 million. This was up $0.5 million from the previous quarter, as expected. Again, in Q3 and Q4, we have reverted to a higher level of customer co-funding for R&D, reflected as an offset to R&D expense. This is serving to offset a higher program spend on certain projects, which, on a net basis, still drove expenses higher by approximately $400,000 in Q3 and will drive incremental spending up higher in Q4 by an estimated $800,000. As in 2015, customer co-funding of R&D will continue into 2016, but it's expected to be somewhat less and will again vary by quarter. Overall, in Q3, we were able to deliver an operating income margin of 20.3%. In Q4, with slightly higher gross margins and slightly higher operating spending, we believe we can deliver operating margin in the range of 20.4% to 21.2%. With regard to the non-GAAP tax provision in Q3 consistent with Q2, we are currently expecting approximately 18%, which does not include the benefit of a restoration of the federal R&D tax credit for 2015. In fiscal 2014, when the R&D credit was restored in Q4, our non-GAAP effective tax rate for the year dropped to approximately 13.2%. We are uncertain whether the credit will be restored again in the fourth quarter of this year. To affect fourth quarter reporting then would depend on the bill passing through the U.S. Congress before the end of the calendar year. Now turning to the balance sheet. Overall, cash was up $14.3 million in the quarter from $87.4 million to $101.7 million. This was due to strong cash flow from operations in Q3 of $19.8 million, up from a comparable cash flow from operations in Q2 of $13.5 million. The improvement is largely due to favorable changes in the working capital, which generated about $4.5 million. The company also had capital expenditures of $5.6 million, which was up as expected from the $3.2 million in Q2. This also allowed us to generate $14.2 million of free cash flow, which is up and consistent with the $10.3 million in free cash flow we generated in Q2. We continue to expect a similar level of cash expenditures in Q4 of 2015. Accounts receivable came in at $33.2 million or 48 days sales outstanding as compared to $35.5 million or 53 days sales outstanding at the end of June. Inventory on the balance sheet at the end of September was $20.8 million, which included $0.7 million of acquisition-related step-up in inventory value as compared to $24.4 million at the end of June, which included $1.1 million of step-up in inventory value. Excluding the relative step-up in inventories, inventory days was down to 92 days or four turns at the end of September compared to 111 days or about 3.3 turns at the end of June. The inventory balance is still a bit higher than normal due to a variety of reasons, which we've covered in the past. We continue to work on bringing the inventory balance down over time. Payables decreased to $5.4 million or 24 days in September from $10 million or 48 days outstanding at the end of June, due largely to the timing of inventory receipts in the quarter. Now, let me recap the business outlook for Q4. I remind everyone, again, that the following statements are based on current expectations as of today and include forward-looking statements. Actual results may differ materially. We do not plan to update nor do we take on any obligation to update this outlook in the future. Our forecast of Q4 revenues is in a range of up approximately 1.5% to up 4.7% or $63.3 million to $65.3 million. We expect non-GAAP gross margins to be in a range of 68.5% to 68.7%. We expect non-GAAP operating expense to be in a range of $30.4 million to $31.2 million. We are currently estimating the non-GAAP effective tax rate to be 18% for the year, again, this does not reflect or include a potential reduction if Congress renews the federal R&D credit during the fourth quarter. We are confident these components should then align, resulting in non-GAAP operating margin between 20.4% and 21.2%. This should then lead to non-GAAP net income of between $10.6 million and $11.4 million, which, on approximately 42.4 million estimated diluted shares, would result in estimated non-GAAP earnings of between $0.25 and $0.27 per diluted share. We also estimate stock-based compensation expense to be between $7.6 million and $7.8 million. In addition, we expect about $3.2 million from purchase accounting-based adjustments. In Q4, we also have another quarter of flow-through of stepped-up inventory value estimated at about $600,000 and we expect about $500,000 in acquisition transition-based expenses. This would imply a GAAP net loss in the range of $500,000 to $1.3 million. GAAP loss per share would then be a loss in the range of $0.01 to $0.03 per share. A more complete reconciliation of the Q4 GAAP net income forecast is attached to the last page of the press release. We will not update this outlook during the quarter until the time of the next quarterly earnings release, unless Inphi publishes a notice stating otherwise, so please ask any questions you may have today during the general Q&A period. And now, we'd be happy to take your questions.

[Operator:]

Our first question comes from the line of Quinn Bolton with Needham and Company. Your line is now open.

[Quinn Bolton:]

Hey, guys. Congratulations on the nice results and outlook. Ford, lots of excitement, it sounds like, around the PAM4 opportunity, all the way from 40 Gig all the way up to 400 Gig. Now that you've got another quarter under your belt, you're recently back from the ECOC show, can you give us a sense what the PAM ramp timing looks like as you come through 2016 or 40 Gig and 50 Gig single-laser opportunities still most likely the first to ramp, or do you see opportunities for 100 Gig modules getting pulled in? And then I've got a couple of follow-ups.

[Ford G. Tamer:]

Thanks, Quinn. So, you are right. 40 Gig and 50 Gig will be first to ramp followed by 100 Gig, and that's going to be in the second half of 2016, and we expect the 400 Gigabit to ramp in the first half of 2017. Having said this, I do want to put this in context. We look at overall 100 Gigabit ramp for our business. We could discuss the long haul and metro at some point. In the data center, we have our original Inphi 100 Gigabit organic businesses, which is the CDR and the Gearbox, and that's still growing strong and expected to keep growing in 2016. We also have a new low-power CDR that is being rolled out into four lambda CWDM application and we expect that to ramp, along with the Silicon Photonic 100 Gigabit module partner for enterprise in a big way in 2016, as well as both DML and EML-based 100 Gigabit modules for cloud data centers in the second half of 2016. So, 100 Gigabit story includes many driver, one of which is PAM and we're still very confident on PAM and had some good qualification progress in field trial in the third quarter.

[Quinn Bolton:]

Great. And then just a sort of follow-up on that commentary. If I look at the Laser and Driver business to date, I think it's mostly been driven by the long haul or metro coherent applications. And if I'm correct, you have fairly small revenue today from Drivers and Amplifiers in the Datacom – 100G Datacom modules, is that correct? And if that is correct, as you look into 2016, how big could that Datacom module be for the Laser Drivers, and CDRs and Amplifiers?

[Ford G. Tamer:]

Quinn, that is correct. So, our business today on the Amplifier and Driver is up over 95% long-haul, metro. We do expect a platform solution bundled with our PAM offering to come to market in the second half 2016, and that's where we are excited about, and that could be a significant opportunity for us. We have not yet quantified it, but we are confident that this will allow us to reach our revenue goals for next year.

[Quinn Bolton:]

Great. And then just switching over to the memory interface market. One of your newer competitors, that market recently announced that they had some technical issues and, therefore, we're going to miss the Intel call cycle. Can you give us any updates that you've seen on the competitive front as we prepare for the next platform ramp – server platform ramp from Intel next year?

[Ford G. Tamer:]

Yes. We continue to gain share. So, we have said on the prior call, we will double our revenue in the second half of 2015 and we are doing what we said we would do, so our revenue is on track to double from the first half of 2015 to the second half of 2015. And we're also on track to continue to gain share and double again in 2016 and continue to gain share through the mid-2017, so we're still confident in our – the statements that we made in the past.

[Quinn Bolton:]

Okay, great. Thank you.

[Operator:]

Our next question comes from the line of Joe Moore with Morgan Stanley. Your line is now open. Please go ahead.

[Joe L. Moore:]

Great. Thank you. I wonder if you can just talk a little about the service provider environment. You alluded to the slowness in China that you saw in Q2 and that that will get better next year. Can you – what do you see there in Q3?

[Ford G. Tamer:]

Yes. Thank you, Joe. We have seen a resumption of the China demand. The big pause there was the Phase 11 of the China mobile deployment, which impacted us and some other vendor. And what we've seen is now increased customer demand, especially for our 100 Gigabit coherent amplifiers. And we believe this will continue through Q4 and into next year, Joe. So, we think the delay was a temporary situation and we're excited about the growth that this could mean for us moving forward, especially combined with some of the latest Infonetics report that are indicating a 70% growth in coherent ports year-on-year from 2015 to 20216.

[Joe L. Moore:]

Great. Thank you. And then separately, can you talk a little bit about 25 Gig, 100 Gig adoption in the data center and how sensitive you are to the timing around that adoption over the course of the next year?

[Ford G. Tamer:]

So, if you look at the latest corporate presentation that we've uploaded to our website, we are now forecasting a small delay in 2016 ramp. So, we've taken the number down from 800K – 100 Gigabit port in 2016 down to 600K ports, and that's based on a Crehan Research report that we can forward to you. But the good news is then the same Crehan Research report is more bullish on that adoption in 2017 and 2018, bringing the numbers up actually in 2017 and 2018. Most of our revenue for the new low-power CDR and the new PAM offering and the accompanying amplifier and driver were assumed to be second half 2016 for us anyhow. So, the delay, we believe, is about a quarter, and for us, it's not as big of an impact. It's a small impact to the second half, but it could actually play in our favor because it reduces the time from introduction to the time our products are available by about a quarter, which could play to our advantage.

[Joe L. Moore:]

Great. Thank you very much.

[Operator:]

Our next question comes from the line of Tore Svanberg with Stifel. Your line is now open.

[Unknown Speaker:]

Yeah, hi, nice results. This is Eric (0

[John S. Edmunds:]

So, Eric (0

[Unknown Speaker:]

Great. So, obviously, then you're getting good feedback from your Gen 2 part. On the PAM4, besides your largest customer who was kind of in the midst of – or largest competitors in the midst of being kind of acquired, are you seeing anybody else on the competitive front?

[Ford G. Tamer:]

Yes, actually the largest competitors are the ones not making noise. They're actually a range of smaller companies, mostly private, that are actually either partners or competitors. They're actually mostly partners. So we've got a range of private companies that we are partnering with to take this to market, and we are – we've got one major competitor that's also a private company that you probably know who I'm referring to.

[Unknown Speaker:]

Got it, great. Thank you very much.

[Operator:]

Thank you. Our next question comes from the line of Vivek Arya from Bank of America Merrill Lynch. Your line is now open.

[Vivek Arya:]

Thanks for taking my question. Maybe, Ford, to start with. I think you mentioned that you were seeing some improved demand from the China business. I know it's probably a little early. But how sustainable do you think this recovery is? Because we have seen in prior cycles where there is a deployment over a couple of quarters and then there is absorption of all that capacity. So just conceptually, how is your visibility as you look out into the first half of next year?

[Ford G. Tamer:]

Vivek, if I could predict a macro, I probably would not be sitting on this call. But I think from where we sit today, we feel confident that Phase 11 of the China mobile infrastructure has resumed. So, based on the data we have today, we're seeing obviously orders flowing back. And the good news, the inventory levels were fairly low, so we're going to see a direct impact into Q4 and Q1. We – at this point, we're calling this going back to normal, so we're not – we believe that this Phase 11 has been approved and moving forward. But that's the data we have today. I mean, it'd be hard for us to say that it may not stop again in six months. I mean we don't have that visibility unfortunately. As of today, it's moving forward.

[Vivek Arya:]

Understand. And maybe, Ford, as a follow-up, so next year, there are expectations that the number of base station deployments may go down. Could you help us understand – I assume that you're probably a little bit more tied to say backhaul and the core of the network and then the metro network rather than base stations and radios per se, is that a fair assumption? So, can you still benefit even if the number of base stations goes down, but just because there are more subscribers coming on, there's a need for greater capacity? So there are parts of the network where they will still need to spend.

[Ford G. Tamer:]

Vivek, you're right on. We are more tied to the number of subscriber as we are to the number of base station. We really are predominantly a backhaul type of story. So, the predominant revenue we get from the infrastructure business is really our driver and TIA business, and that is tied to really long haul and metro type offerings, which are directly correlated to number of subscriber as opposed to number of base station.

[Vivek Arya:]

Got it. One last one as a very quick housekeeping. If, John, you could give us some color around the relative sequential growth for your comms, memory and IP business in Q4, and if you have some color for us on how IP could trend in 2016, it would be very helpful? Thank you.

[John S. Edmunds:]

Yeah, Vivek, I think in Q3, we had about $300,000 of IP licensing revenue. We expect a touch more than that in the fourth quarter this year and then we'd expect kind of a modest level again next year, similar to the amount of IP revenue that we'll have this year. There are opportunities there and we'll kind of continue to pursue them as they come along. So, and then, otherwise, I think we talked already about our high-speed memory business growing sequentially in the quarter and that communications grew per se, although not as – not quite as strong in relative terms as the high-speed memory did. So, we did have a good quarter overall. And when you look at the year-over-year growth, the core communications is still sustaining this 47% year-over-year growth capability.

[Vivek Arya:]

Okay. Thank you.

[Operator:]

Our next question comes from the line of Ross Seymore with Deutsche Bank. Your line is now open.

[Ross C. Seymore:]

Hi, guys. Thanks for letting me ask the question. Congrats on solid execution in a difficult time. I guess, Ford, you've gone through a ton of different design wins, drivers, new products, et cetera. If you were to put those all together and just think from a dollar perspective as opposed to a percent perspective, how would you rank-order the biggest drivers of growth as we go into 2016?

[Ford G. Tamer:]

So, our – the biggest driver of growth as opposed to biggest revenue period, right, I mean you're looking for growth. Remember, some of these are going from small base. Obviously, the new product as a percent growth would probably be the largest growing ones, just because we're starting from not much, Ross. So, I could take you through, if you'd like, the different businesses and we could talk about magnitude and percent, Ross, if that makes sense, right.

[Ross C. Seymore:]

Sure. Yeah, that'd be great. Thank you.

[Ford G. Tamer:]

So, if we start on the memory side, as we said, the DDR4 is going to be high growth. It would double again from 2015 to 2016, so that'd be a nice growth factor for us. But if you keep in mind that it would be offset by a declining DDR3, we then overall expect that to dampen that growth on the memory side, right, between the DDR4 growing in a very nice way and DDR3 declining. On the long haul and metro markets, that's the largest business we have today, and it's been growing at the 50% corporate number for the overall communication business. And the number of ports next year is expected to grow tremendously by 70% year-on-year, and we also have some new offerings taking us into metro with the surface mount driver. Now there, what happens is we are under pressure from a price decrease. So, the number of ports would be offset by the price decrease and would be helped by the continued transition to linear and the continued entrance into new markets such as Metro, and CFP Digital, and CFP2 analog. So, we've got, in that market, great expectation for the CFP digital as well as some new design we got on CFP2 analog. And then coming to networking, the core business, which was the 100 Gigabit networking, as well as the 10 Gigabit and 40 Gigabit physical layer devices we acquired from Cortina are still expecting to get some nice growth driver as the 100 Gigabit continues to ramp and we get some more designs with the 10 and 40 Cortina into servers. The PAM and the new low-power CDR are going from a very low base, almost no revenue in 2015 to speak of, so that growth should be very high just because you're starting from a very small number. And finally, transport is doing better than expected. So transport is actually holding up better than expected and we do expect some incremental potential revenue. So, you put all this together and we are confident in getting to the consensus number that is currently out there. But we are – we want to be cautious on setting expectations ahead of ourselves, and would like to keep executing and delivering a step at a time.

[Ross C. Seymore:]

That's great to hear. There are so many different drivers. I guess to summarize all of that from a revenue perspective, if you put it together, it sounds like the mix between towards your memory and your communications in general will probably skew a bit towards the comp side further in 2016, is that fair?

[Ford G. Tamer:]

Yeah. I think fair, Ross. But both businesses are expected to grow. And so, you know, I would expect to see something in the 20% to 25% range for the server side of the business over time and something in the 75% to 80% range for communications business. That mix will generally continue.

[Ross C. Seymore:]

And then I guess the follow-up to that, John, you guys have done a great job on the gross margin this year. Given the new products, some of the ASP transitions that, Ford, you talked about, how should we conceptually think about the puts and takes to gross margin as we go into next year, again, with the IP revenues assumed to be relatively similar year-over-year?

[John S. Edmunds:]

In general, I think we can maintain the gross margin levels. I would always assume they're going to go down 100 basis points a quarter and hope that the reality proves us wrong and our job is to bring in new technology, which we're doing. And I think that's going to allow us to maintain a fairly stable gross margin picture over the course of 2016.

[Ross C. Seymore:]

Great. And then my last question will be on the OpEx side of things. You talked about comfort with the consensus and wanting to keeping that under control for next year. But just again, conceptually, when you look at the new product launches you have and the areas of investments that you look forward to for next year, how should we think about the general OpEx trend that you guys are planning to run for 2016?

[John S. Edmunds:]

Well, I think, again, we'll see OpEx rise in 2016 and some of that will just be a function of not having quite as much customer co-funding as we had in 2015 here. But in general, we're, I think, comfortable with Street estimates right now for 2016 OpEx, which I think are trending around $135 million or so, and that's, obviously, up from this year, but again it's just a function of the customer co-funding per se.

[Ross C. Seymore:]

Perfect. Congrats. Thanks again.

[John S. Edmunds:]

Thank you.

[Operator:]

Our next question comes from the line of Srini Sundararajan with Summit Research. Your line is now open.

[Srini Sundararajan:]

Hi. Thanks for taking the question. Congrats on your good performance. My question is, have any of the mergers strengthened any of your competitors?

[John S. Edmunds:]

Yes, of course. We had a 800-pound gorilla just became 1600-pound, Srini.

[Srini Sundararajan:]

Right. But with respect to your business, what has been the effect of the mergers – with respect to your business, meaning even though it became 1600-pound, but that's relative to size, but...?

[John S. Edmunds:]

At this point, the business continues to grow. We remain very focused on growing the business, achieving the best possible shareholder return and supporting our customer success. We understand the benefits of scale. In fact, for us, Cortina was a very successful acquisition for both Inphi and Cortina employees, customers, and shareholders. So we remain open to possibility of scaling on our side.

[Srini Sundararajan:]

Okay, great. One more question. What is the reason for your high days of inventory, and then what actions might you be taking to reduce it?

[John S. Edmunds:]

Srini, this is John. We disclosed a couple quarters ago that one of the things that came along with the Cortina acquisition was a supply of a 10 Gigabit framer, a very successful part that Cortina's shipped hundreds of millions of dollars over time, but they had a last time buy notice from their foundry before the acquisition. And we received the last time receipt of inventory in the March timeframe, so we received a fairly solid chunk of inventory that we're shipping overtime here. We don't really have any concerns about being able to move that inventory or ship it, but it is, you know, putting a chunk of inventory on the balance sheet. I'm not too worried of 92 days. I'd rather have it come down to around 80 days over time here, but we're making progress with it, so I'm not too uncomfortable at this stage.

[Srini Sundararajan:]

Okay, great. Thank you very much.

[Ford G. Tamer:]

Thank you.

[Srini Sundararajan:]

All of my other questions have been asked and answered. Thank you.

[Ford G. Tamer:]

Thank you, Srini.

[Operator:]

Our next question comes from the line of Brian Alger with ROTH Capital Partners. Your line is now open.

[Brian Alger:]

Thank you. Good afternoon, guys, and I'll echo the congratulations. Wondering if we could look into the 2016 opportunity. Obviously, sounds like long haul and metro continued to be the right main drivers from a revenue standpoint. But as we look at the 100 Gig opportunity inside the data center, can you maybe walk us through what changes are occurring with the largest data center providers in terms of moving towards Inphi and why we think whether it's the second half of 2016 or 2017, what specifically is going on within those data centers that from an architecture standpoint is coming our way, so to speak?

[Ford G. Tamer:]

The biggest change, Brian, is being driven by the introduction of 25 Gigabit switchers and NIC cards. So, these new controllers and new switches are driving the uplink to 100 Gigabit. And because of the new switchers having the need for very dense interconnect, the latest introduced switchers can support up to 32 QSFP, 100 Gigabit module. The need for power-effective and cost-effective interconnect is skyrocketing. And if you look at the different presentation by various data center participants, everyone is pointing to the big bottleneck being the 300-meter to 2-kilometer 100 Gigabit interconnect. Inphi is supplying all the guts that go into a QSFP 28 module, and we do believe that we've got all the interns (0

[Brian Alger:]

Okay, great. And just as a follow-up, John, you mentioned that the core communications was up about 47% year-on-year, splitting hairs between where we're targeted. How should we think about core communications in 2016 taking everything into account between the obvious large port growth in the metro and the emergence of Datacom and 100 Gig? How should we think about that core communications growth in the 2016 timeframe relative to 2015?

[John S. Edmunds:]

It should be a fairly similar growth pattern, Brian. I think we've talked, and I think you'll see it on our latest presentation charts on the website, but you'll see the growth in 100 Gig ports expected for 2016, so I think around 63%. So – obviously, some ASP erosion, but the 50% growth rate is not unreasonable relative to that port count growth.

[Brian Alger:]

Great. That's all I had, guys. Thanks a lot.

[Operator:]

Our next question comes from the line of Ruben Roy with Piper Jaffray. Your line is now open.

[Ruben Roy:]

Thank you. Ford, I just had a follow-up and clarification on some of the 100 Gig commentary that you're making as it relates to PAM4. Is there still an opportunity that somewhat materially, would you say, around two lambda 50 Gig, or is this PAM4 opportunity really skewed towards 100 Gig and something that we should expect starting in second half of next year? Last quarter, you had talked about some initial shipments spring of 2016. I'm just wondering if the (0

[Ford G. Tamer:]

So, Ruben, maybe the way to think about our PAM4 is, first, it's going to be 40 Gigabit, okay. So, what the data centers are doing is adopting our solution first for 40 Gigabit because right now, there is a solution out there that is exclusive to one system OEM. And so, we've got many other OEMs adopting our PAM4 solution as it's a cost-effective and power-effective 40 Gigabit solution, and so that you'd probably see that deployed first. Second, you'll see the 100 Gigabit, so 100 Gigabit would be on the heels of that 40 Gigabit. We expect the 40 Gigabit to probably ship in around mid 2016, then we expect the 100 Gigabit PAM4 solution to ship in Q3. Prior to that though, on 100 Gigabit, sort of when I talk about QSFP and the whole gut, the CDR solution could ship slightly before the PAM solution or around the same time in Q3 as well. So, we start with 40 Gig, then we go to 100 Gig, with both CWDM and PAM and then we expect the 50 Gig to be probably end of 2016, early 2017. And then we expect the 400 Gigabit to be the first half of 2017. So, that's sort of the succession of the different speeds. Hopefully, that clears it up.

[Ruben Roy:]

That's very helpful. Thank you, Ford. I guess, just a quick follow-up. You mentioned showing a solution that works with the Silicon Photonics partner. Silicon Photonics has come up a little more recently in some of the discussions around tradeshows and some of the companies that are participating. Would you say that that that's a technology that is starting to look like it's going to enter commercialization, I don't know, over the next 12 months to 18 months, or is it still something that's kind of out there?

[Ford G. Tamer:]

Actually, it will enter commercialization in a big way next year. So in two areas, we – you probably know there's a strong vendor in the long haul, metro type of space who is coming to market with a digital CFP solution, and that's a strong partner of ours. There's also, in the data center space an LR4 type of solution. There is a strong system OEM, which has an internal silicon photonic solution and that's a strong partner of ours. So we – in both the long haul, metro as well as data center markets, we do expect Silicon Photonics to happen in a big way in 2016. And we're, at this point, getting good visibility into improved ramp into 2016 for both segments. Now, there are other partners that would be coming to market later in the year that have not disclosed and were not ready to disclose. But we do believe Silicon Photonics starts in 2016 and grows strongly from there on.

[Ruben Roy:]

Thanks Ford.

[Operator:]

Our next question comes from the line of Gus Richard with Northland Capital. Your line is now open.

[Gus Richard:]

Yes, thanks for taking my question. Based on the guidance you gave for your memory business, is it reasonable to assume that the DDR3 revenue is going to be about $45 million sort of midpoint, am I getting that correct?

[John S. Edmunds:]

Yes, that's about right. It would remain in the $43 million-$45 million range, Gus.

[Gus Richard:]

Okay. And then what's your expectation for that market for next year?

[John S. Edmunds:]

Well, DDR3 will continue to decline and DDR4 will grow. And right now, we'd expect that they at least offset each other. We think DDR4 will grow a little bit more, then DDR3 will decline. So we should have a little bit of growth overall or we may have more growth than that. We just want to wait and see how the design wins unfold and what volumes actually ship.

[Gus Richard:]

Got it. And then just to be explicit, on the OTN business. Was that down sequentially? Can you talk a little bit about what you're seeing in that market?

[John S. Edmunds:]

We've just seen some relative strength there, Gus. And what I would say is that there were design wins where people have equipment that's continuing to ship, and there may be other newer solutions or platforms that just haven't come out online as fast as people thought originally, and so people continue to ship with the Cortina framing and mapping solutions, and we're happy to oblige as long as they're continuing to ship in volume.

[Gus Richard:]

Okay. So that business is effectively running flat with last year, or is there just a slight decline? Any color again would be helpful.

[John S. Edmunds:]

Again, what we said is that the declines have been less than what we had expected.

[Gus Richard:]

Okay. All right. Thank you very much. Appreciate you taking the questions.

[John S. Edmunds:]

Thanks, Gus. Yeah, bye-bye.

[Operator:]

Thank you. And our last question comes from the line of Richard Shannon with Craig-Hallum. Your line is now open. Please go ahead.

[Richard Cutts Shannon:]

Hi, guys. Thank you for taking my questions. I guess, just a couple from me. I was reading through my notes from earlier in the call about – a question about share in DDR4. John, I think it was you who was responding to the question. I think you mentioned kind of 25% number for next year and then maybe 40% number in 2017. I wrote in my notes that you've mentioned just registers. Did you mean to be that both for registers and buffers, or is it specific to one or the other?

[Ford G. Tamer:]

Yeah, I meant for both. In reality, we're shipping registers today and through this year and then we'll begin – we think this would buffer sometime next year, so that will enter into the mix and will contribute toward that share build.

[Richard Cutts Shannon:]

Okay, perfect. That's what I thought. Second question, regarding the new drivers. Ford, you'd mentioned some pricing pressure in the driver space and ASP ratchet down. I guess just a couple things. When and how is that happening? Is that based on competitive pressures, and are you also seeing the amplifier side or is that kind of resisting the pricing pressures?

[Ford G. Tamer:]

Richard, this is a regular ASP decline that we've been experiencing for the past few years. So, it's not anything unusual. It's just a regular course of running that business.

[Richard Cutts Shannon:]

Okay. Fair enough. I guess just last question for me. Regarding the customer co-funding, it sounds like you're expecting some funding to go through at least part of next year. When do you expect to see the product result of this work? Is this something that will happen before the end of next year, is it further out than that?

[John S. Edmunds:]

No, we expected to see the fruits of the latest customer co-funding shipping in the back half of next year, Richard. So the projects tend to run 18 months to 24 months depending on what we're doing or what we're trying to put together for somebody, and then sometimes there's a follow-on project. But we're already shipping product today and starting to ship in volume for things where we've got co-funding a year and a half to two years ago. So it's kind of an ongoing...

[Richard Cutts Shannon:]

Okay. Can you say whether the programs are ongoing whether in the communications or memory space or both?

[John S. Edmunds:]

They generally have been in the communication space, but they'll be in both.

[Richard Cutts Shannon:]

Okay. Great. That's all the questions for me guys. Thank you very much.

[John S. Edmunds:]

Thank you.

[Operator:]

Thank you. I'm showing no further questions. I would like to turn call back to John Edmunds for any closing remarks.

[John S. Edmunds:]

Okay. I would like to thank you guys for joining us today, and we want to let you know we're planning to attend the Stifel One-On-One Conference in Chicago on Thursday, November 12, the Raymond James Conference in New York on December 8, and the Wedbush Conference in Los Angeles on December 9. Again, we would like to thank you for joining us, we look forward to speaking to you again in the future.

[Operator:]

Ladies and gentlemen, thank you for participating in today's conference. This does conclude the program, and you may all disconnect. Everyone have a great day.