Inphi Corp
Q3 2016 Earnings Call Transcript

Published: 2 Nov 2016

[Operator:]

Good afternoon and welcome to the Inphi Third Quarter 2016 Earnings Results 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 conference is being recorded. I would like to introduce your host for today's conference, John Edmunds, Chief Financial Officer. You may begin.

[John S. Edmunds:]

Thanks, Ciara. Good afternoon, everyone. Thank you for joining us today to discuss our financial results for the third quarter of 2016. I'm John Edmunds, Chief Financial Officer. And with me today is our CEO, Ford Tamer. I'll begin with the Safe Harbor. Then Ford will give you an overview of the business. And after that, we'll provide financial summary for Q3 and the outlook for the fourth quarter of 2016. Then we'll be happy to take your questions. Now, for the Safe Harbor, please note that during the course of this conference call, we may make projections or other forward-looking statements, both about Inphi and/or the recently announced proposed acquisition of ClariPhy. 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 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. For this year in particular, you will notice we are reporting based on continuing operations and that historical numbers in the press release have been adjusted to reflect our recent sale of our Memory business to Rambus. This is being done in accordance with GAAP rules for reporting discontinued operations. However, for convenience, we have included a reconciliation to the historical reports for both the GAAP and non-GAAP reporting on our website for the seven quarters ended September 30, 2016. Now, to begin our review of the quarter, let me turn our call over to the CEO, Ford Tamer. Ford?

[Ford G. Tamer:]

Thanks, John, and good afternoon. Thank you for joining us for our third quarter 2016 earnings call. Today, I'll summarize our financial results and guidance, discuss our acquisition of ClariPhy Communications, and highlight key announcements in the quarter. Then John will cover the details of our Q3 financial results and Q4 guidance. This was another terrific quarter for us and one in which we reaped the rewards of our investments while we continued to plant seeds for future returns. Once again, we delivered a very solid quarter, beating consensus projections on both the top and bottom lines. In the quarter just ended, we generated revenue of $70.7 million, non-GAAP operating income of $20.9 million, and EPS of $0.46. These represent sequential increases of 17% on revenue, 42% on non-GAAP operating income, and 44% on non-GAAP EPS compared to the results we reported in Q2 of 2016. It also represents year-on-year increases of 49% in revenue, over 100% in non-GAAP operating income, and over 119% in non-GAAP EPS over the results we reported one year ago in Q3 2015. All in all, it was a very strong quarter for Inphi. This strong performance will also have a positive impact on our annual guidance for 2016. In Q4 2016, we now expect revenue in the range $74.7 million to $76.7 million, or up 7% sequentially at the midpoint. Before I discuss the product strengths that contributed to these strong results, let me tell you about our acquisition of ClariPhy Communications. For those of you unfamiliar with the company, ClariPhy is a leading provider of coherent DSP, targeting service provider solutions and the long haul and metro markets. The ClariPhy solution fits squarely into the Inphi product matrix alongside our optical TIAs and drivers. This acquisition will enable us to offer complete platform solutions to our long haul and metro customers and help us leverage our key enabling technologies across multiple market segments. As you know, we've always refined our product portfolio, cutting those products that don't fit and adding new ones to complement our already strong solutions. The ClariPhy acquisition exactly fits that pattern. At first, it was our customers who encouraged us to add a coherent DSP solution to Inphi's existing optical TIA and driver offerings to enable a multi-vendor coherent DSP ecosystem. As we studied the possible options, we determined that ClariPhy could strengthen our commitment to the long haul and metro market segments. It would also complement both our direct detect and PAM solutions between and inside data centers. We envision ClariPhy as a very strategic addition, both for 2017 and more significantly for 2018 and beyond. Moving to the specifics of the transaction, we have agreed to pay $275 million in cash, and we will absorb certain ClariPhy liabilities. We expect the transaction to close at or near year-end. We do not expect this acquisition to have much impact to our Q4 results, but expect it to add approximately $45 million to $50 million in revenue in 2017. As we ramp into 2018, we expect new product introductions to significantly add to our top line. While we see a slight dilutive impact to earnings in the first quarter of 2017, we do expect this acquisition would be an accretive transaction starting in the back half of 2017. We also look forward to welcoming the talented ClariPhy employees to the Inphi team. In summary, we're confident that our organic investments and the ClariPhy acquisition will create a compelling portfolio for our customers. This will meet a greater percentage of our customers' needs for interconnect and long haul and metro, between data centers and inside data centers. Now, let me turn back to our existing portfolio and our strong organic results in the just-completed quarter. Inphi continues to enable our customers to be successful and grow their market share. In Q3, we aggressively moved to extend our market leads on several fronts. Beginning with long haul and metro, we made several important announcements in September at ECOC 2016, the European Conference on Optical Communication. In Dusseldorf, we showcased our quad-linear driver in a surface mount technology or SMT package. This newest member of our 32-Gigabaud coherent product family extends our portfolio of cost-effective 100 to 200-gigabit per second products. We're scheduled to ship a significant number of these low-cost, high-performance drivers by year-end. Additionally, at the ECOC conference, we announced sampling of the world's first 64-Gigabaud dual-channel linear TIA. This advanced component uses coherent technology to support data rates of 400 to 600-gigabit per second on a single wavelength. Together with a companion driver, Inphi now offers a very compelling 64-Gigabaud solution for our Tier 1 customers who are moving data across hundreds or thousands of kilometers. This flexible solution works with multiple modulation formats. As a result, it is capable of dynamically changing transmission capacity and reach, depending on network traffic and capacity needs. This supports our customers' march to ever more efficient, faster networks. Moving on to networking interconnect inside data centers, we raised the bar by announcing the world's first 400-gigabit PAM4 solution for next-generation CFP8 modules. The solution includes our PAM 4 DSP as well as the corresponding linear TIA and driver. With this announcement, we can now offer our router customers an even more efficient low-power, end-to-end platform to move data an at accelerated rate, both between and within data centers. Finally, in our DCI Edge segment between data centers, I'm very pleased to report that our ColorZ DWDM platform is on schedule. We are in full field trials at multiple customers and have received production orders that support our current revenue forecasts through the first half of 2017. We continue to see tremendous interest from additional customers, interest that we expect will be visible in our financial statements as we continue to ramp in 2017. As we look ahead, we firmly believe Inphi's future is very bright across many customers, market segments, products and geographies. We are pleased with our positioning to support our customers' success in the marketplace. We are, as we've always maintained, in the right place with the right products, with the right team, at the right time. Now, let me turn the call over to John to go over the details of the recently ended quarter and to offer guidance for the final quarter of the fiscal year. John?

[John S. Edmunds:]

Thanks, Ford. Now, let me recap the financial results. In the third quarter of 2016, Inphi reported revenues of $70.7 million which, as Ford told you, was up sequentially in Q2 by 17% and up year-over-year by approximately 49%. Core communications revenue for the nine months year-to-date has grown by 61% and for the year is now expected to have year-over-year growth in the range of 60% to 65%. We define our core communications as amplifiers, drivers as well as 10, 40, and 100-gig physical interface products, all of which sell into the service provider and data center interconnect markets. The growth is coming from all three areas, with drivers growing a bit faster this year from a smaller base. The core communications products continue to represent approximately 80% of communications year-to-date and while the transport and legacy communication products represent the remaining 20% year-to-date. As expected, transport and legacy in Q3 was down a bit from Q2 and will continue to decline over time. For this year, we expect transport and legacy to be just below 20% of overall revenues. Gross margins on a non-GAAP basis in Q3 came in at 73%, which was down about 60 basis points from the same quarter in Q2, or about half of the decline we had previously forecasted. In Q4, we expect the gross margins to decline 70 to 170 basis points, again based on the relative forecasted mix of products. All in, we estimate non-GAAP gross margins for Q4 to be in the range of 71.3% to 72.3%. We continue to be confident that gross margins can remain above 70% for all of 2017. On a GAAP basis, we had GAAP positive operating income in Q3 of $10.2 million, up from Q2 of $4 million and compared to a $1.7 million operating loss in Q3. One year ago, the delta between GAAP and non-GAAP operating income is approximately $10.7 million, which is essentially the same as Q2 and Q1 and is made up of several basic recurring adjustments. You'll recall in previous quarters we discussed the GAAP measure includes stock compensation expense of $7.2 million, purchase accounting related adjustments totaling $3.2 million and divestiture related differences in non-GAAP reporting of discontinued ops of $300,000. Finally, we had $3.2 million of noncash debt cost amortization recorded in other income associated with the convertible bond issues. There was also a gain on the sale of a private company investment, yielding a $1.1 million profit and a $1.2 million overall associated tax benefit of all of the GAAP to non-GAAP adjustments. The Q3 GAAP net income resulted from continuing operations of $6.6 million or $0.15 per share. This comparison net income from continuing ops in Q2 of $900,000 or $0.02 per share. It also compares to $1.9 million or a loss of $0.05 per share in Q3 of 2015. The Q3 2016 GAAP net income was more positive than Q3 2015 mainly from better operating income. Now, for further clarification and analysis of the operating results, let's turn to some additional non-GAAP measures and comparisons. On a non-GAAP basis then, net income for the third quarter of 2016 was $20.6 million or $0.46 per diluted share, which is more than double compared to non-GAAP net income of $8.7 million or $0.21 per share in Q3 2015. Non-GAAP operating expenses for the third quarter totaled $30.7 million. This was up $0.9 million from the $29.8 million in the June quarter, due primarily to the higher personnel expenses. Continuing operating expense also benefited from this quarter from about $1 million in favorable revisions to estimates related to discontinued operations reporting. In Q4, we will not get the same benefit. We will also be taping out our first 16 FinFET part for development which will cost about $1 million of additional expense in Q4. Along with additional hiring, in total, we expect operating expense to go up in a range of $1.6 million to $2.6 million in the fourth quarter, or be in a range of $32.3 million to $33.3 million. Overall in Q3, we were able to deliver non-GAAP operating margin of 29.6% which compares to 24.4% in Q2. The improvement is primarily due to higher volume of revenue in Q3. In Q4, due to additional operating spending and slightly lower gross margins, we expect operating margin to be between 28% and 29%, GAAP other expense of $2.1 million, when reduced by the noncash debt cost to amortization of $3.23 million, which provides GAAP net other income of $1.13 million. We had profit on the sale of a cost-based investment at $1.14 million. So net other income of non-GAAP was about $10 million for the quarter. Included in this number is cash interest expense on the convertible debt of approximately $0.76 million, which is offset by interest income of $0.52 million. This resulted in net cash-based carry costs associated with the convertible debt of about $250,000 for the quarter. With regard to Q3, non-GAAP tax provision due to a higher mix of international business, we have been able to project an effective rate for 2016 of 8%. We are also enjoying a discrete tax return to provision correction benefit of $1.1 million. This has resulted in non-GAAP taxes being considerably reduced in the quarter to $300,000, or 1.4% of pre-tax. In Q4, we currently expect a non-GAAP effective rate of 8%. The change in rate would produce about a $0.03 benefit to non-GAAP earnings per share in Q3. Cash income taxes paid for the nine months year-to-date was $226,000, about 20% of what we had paid through the same period last year. Now turning to the balance sheet, overall cash of $693.5 million was up from Q2's balance of $327.8 million. We continue to have strong cash flow from operations in Q3 of $25.6 million compared to Q2's $11.6 million. About $5 million of the improvement was from a higher base of income of $5 million, and about $14 million from better working capital management. The $25.6 million plus $4.8 million in employee option and ESPP proceeds were then offset by $4.5 million in capital expenditures and $2.4 million in stock withholding on RSUs, resulting in about $23.5 million in net cash increase. This then combined with $78.7 million coming in from the sale of the memory business and $6.3 million from the sale of a cost-based investment. In addition, we had $258 million in proceeds which we raised from the second convertible bond issue. That accounted for the $366 million increase in cash and investments from $328 million to $694 million. We continue to expect approximately $6.5 million of capital expenditures in Q4 of 2016. Accounts receivable increased by $2.4 million and DSO increased from 50 days sales outstanding in June to 51 days outstanding at the end of September. Inventory increased by $3 million during the quarter. However, inventory days and turns were down to 69 days and 5.2 turns at the end of September compared to 66 days or about 5.4 turns at the end of June. Payables increased $5 million in the quarter or 50 days, which is up from the 41 days outstanding at the end of June. 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. In addition, this guidance does not include the impact of the acquisition of ClariPhy which is likely to close at the end of the quarter. ClariPhy is currently shipping $8 million to $10 million in revenues with product gross margins that are higher than Inphi's. With synergies, operating income will be approximately plus or minus 5% to breakeven and combination is expected to be mildly dilutive to the first half 2017 and accretive in the second half of 2017 and accretive overall for the year 2017. In the context of Inphi's overall results, we are forecasting the dilution to be so mild that I doubt investors will see or recognize the dilution in the results. Due to new product introductions coming in the middle of 2017, significant growth and accretion are expected from the ClariPhy products in 2018. Again, for Inphi standalone, our forecast of Q4 revenues from continuing operations is up sequentially in a range of approximately 5.5% to 8.5% or $74.7 million to $76.7 million. We expect non-GAAP gross margins to be in the range of 71.3% to 72.3%. We expect non-GAAP operating expense to be in a range of $32.3 million $33.3 million. We are currently estimating non-GAAP effective tax rate to be 8% for Q4 2016. We're also confident these components should align, resulting in non-GAAP operating margin in the range of 28.1% to 28.9%. This should also lead to non-GAAP net income of between $19.4 million and $20.5 million, which on approximately 44.6 million estimated diluted shares would result in estimated non-GAAP earnings per share of between $0.44 and $0.46 per share. We also estimate stock-based compensation expense to be between $7.1 million and $7.3 million. In addition, we expect about $3.2 million from purchase accounting adjustments in Q4. We also expect $5.9 million in noncash amortization of debt expense. Finally, these increases should also generate a related tax offset of approximately $950,000. This would imply a GAAP result on net income in the range of $4 million to $5.1 million. GAAP earnings per share would then be in a range of $0.09 to $0.11 per share. A more complete reconciliation of the Q4 GAAP net income is forecast and 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 have today during the general Q&A period. And now we'd be happy to take your questions. Ciara?

[Operator:]

Thank you. And our first question comes from the line of Quinn Bolton from Needham & Co. Your line is open.

[Quinn Bolton:]

Hey, guys. Very nice job on the results and the guide in the core business. I wanted to ask first on the ClariPhy acquisition, obviously, one of three merchant suppliers of coherent DSPs. Can you give us your view on what the TAM for the merchant DSP market might – my sense was there was only about $120 million. So I'd love to get your thoughts if that's a good number, what that TAM might be growing at? And then a related question, obviously, one of your big customers, Acacia, is a supplier of merchant DSPs to some of its customers. And it sounds like you may start to compete a little bit with Acacia now in the coherent DSP market. So how do you handle that relationship going forward?

[Ford G. Tamer:]

Thanks, Quinn. The TAM is estimated to grow from about $200 million or so today to $500 million, and the growth is going to be driven by a move to lower geometries, so today the DSPs that are being fielded are 16-nanometer, that's going to move to 7-nanometer and that transition should help ASIC type of solution, a commercial ASIC type of solution. And then there is also a move to lower power DCO, a metro type application, again, there probably would be more and more willingness on some of the customers that we have to potentially partner for those offerings. Finally, I think there is a demand from module partners for merchant DSP solutions that are standalone that could go help them build either DCO or daughter card to provide their customers. We believe those three trends bode well for the increase in the commercial TAM for this product. Regarding Acacia and NEL, they're both very strong partners, and we hope to continue those partnerships. They're obviously areas where we compete, but we do believe that a rising tide will help all boats and we believe it's a market that's going to be expanding significantly for all of us, and so we believe we'll be able to continue to hopefully have some partnership as well as, in some cases, we may have to compete. I think we're being pulled into this by our tier one OEM customer as well as by our modular customers.

[Quinn Bolton:]

Got it, understood. And then just on the core business, obviously very strong results there, Ford. Obviously, some nice RFPs it sounds like coming out for China looking into 2017. Can you just give us a sense what your visibility looks like based on some of those new RFPs and the metro build in North America?

[Ford G. Tamer:]

A very good question, Quinn. We are actually positive on short term and long term. We do believe 2016, in China, was driven by the buildout of the national backbone, and as we go into 2017, that will be moving to buildout of the provincial backbone for tier two and tier three cities and some new RFPs for the phase 12 of that buildout. Then moving 2018, we do see 5G becoming a much bigger tracks there. In North America, obviously AT&T and Verizon have been rolling out 40 gig and 100 gig, and we recently AT&T announced this intention to build a 400-gigabit network, and that's very good news for us because we believe we've got the right products for 400 gigabit both ourselves and with our partners. Again, longer term, the North American market will benefit from 5Gs. We do believe we're at the beginning of a multi-year optical super cycle as you call it.

[Quinn Bolton:]

Great. Thank you. I'll cede the floor.

[Operator:]

Thank you. And our next question comes from Joe Moore from Morgan Stanley. Your line is open.

[Joseph Moore:]

Yes. Thank you. I wonder if you could give us a little bit more color on the competitive vision of the ClariPhy products and what's the differentiation in coherent DSPs and maybe touch specifically on power consumption which I understand is a key driver for that space?

[Ford G. Tamer:]

Yes. Thanks, Joe. The ClariPhy team, from its heritage, has been a high performance team so they've always been very good at getting every half DB and being able to extend the reach for submarine application, very Long Haul application. This is really where they really shine. They were the first ones to come up with what's called a CL200 product which was first of its kind when they first introduced it. After that, Acacia came in with the lower-power product and they launched that piece on the metro but now they have a pretty strong pipeline of products coming up for both metro and long haul they were very positive about. So that ClariPhy team really comes from a, as I said, high performance component product, they're adding low-power to the road map and we believe they're going to be very competitive. There's a couple things to keep in mind as we look at this. ClariPhy designs its own chips, so that includes analogs, DSP core, physical design. They also do ASIC for other customers. So it opens up a couple of potential opportunities for us. Number one, from a competitive platform cost positioning, it provides a very good, effective platform costs because they do their own ASIC. They're not depending on other people to do their ASICs for them. Number two, some of our customers are actually asking for either IP or for us to do custom ASIC for them because these ASICs are becoming very expensive and now we have the scale to potentially get into this ASIC business for our Tier 1 customers. And so it opens up a couple of very interesting positions for us that are different than some of the other players in this market because ClariPhy owns all of their own tools, their own backend, their own analog, their own layout, et cetera so that's a big advantage.

[Joseph Moore:]

Okay. Great. Thank you for that. And then a separate question. You had mentioned field trials at multiple customers for ColorZ between data centers. When might you see multiple customers contributing to the revenue in that space?

[Ford G. Tamer:]

So we're still very much on track as I said in my prepared remark. Our order book already reflects demand that's higher than what's in our forecast for the first half of the year. We expect to be able to announce a second customer in the first half of the year. We're on track to increase yield and improve capacity, or improve yield and increase capacity. We're very pleased with the progress we're making on ColorZ. On track.

[Joseph Moore:]

Great. Thank you very much and congratulations on the deal.

[Ford G. Tamer:]

Thanks, Joe.

[Operator:]

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

[Ross C. Seymore:]

Hi, guys. Just wanted to follow up on the ClariPhy side of things, and no pun intended, clarify the revenue growth that you said. You said $8 million to $10 million in revs, but for the full year it was kind of $45 million to $50 million. Is that kind of the growth rate that that asset and the revenue side is growing year over year? And any kind of color you can give on the year over year growth rate of that business will be helpful.

[Ford G. Tamer:]

Sure. I think what John was talking about, the $8 million to $10 million is the current run rate. We expect next year to be $45 million to $50 million in 2017, and expect 2018 to be upward of $75 million, call it $75 million to $85 million range, but upward of $75 million.

[Ross C. Seymore:]

Great. And then for your core business, could you just talk a little bit more about actually the ColorZ side of things? You talked about the new customers coming in last quarter, you talked a little bit about getting the capacity set before the revenue growth rate would run. Is there anything that changes the size of the business that you expect to come from Microsoft? And is the aggregate size for 2017 as a whole rising versus what you had originally expected on, say, the last quarterly conference call?

[Ford G. Tamer:]

The answer is yes. The aggregate demand that we're seeing from the market is significantly higher than we had expected on the last conference call so we are much more positive and confident in the demand in the market for those set of products. And we're really going to be limited next year by capacity and our ability to ramp that capacity.

[Ross C. Seymore:]

I guess one final clarification for John on the accretion/dilution side of things. When you talked about the dilution being de minimis and hard for us to even notice, was that in response to just the first half of the year so the magnitude of the dilution early in the year is going to be de minimis? Or was that kind of the full year?

[John S. Edmunds:]

No, that was with respect to the first half of the year. We expect the transaction to be accretive in the second half, particularly the fourth quarter, and we'll be measurably accretive in 2018.

[Ross C. Seymore:]

Great. Thanks, and congrats on the deal.

[John S. Edmunds:]

Thank you.

[Operator:]

Thank you. Our next question comes from the line of Tore Svanberg from Stifel. Your line is open.

[Tore Svanberg:]

Yes, thank you and congratulations on the strong results. First question is on ClariPhy and this becoming a bigger merchant market. Could you just elaborate on a little bit of that, Ford? You said that your customers are kind of pulling you in, so I'm just trying to understand the dynamics there, if this is sort of across the board or is it a few customers? And what is it about the move to lower geometries that perhaps really opens up the merchant market more?

[Ford G. Tamer:]

Sure. So I think there's probably five market segments, Tore, where we could maybe categorize this in. So the first one is the sort of heritage of ClariPhy, this long haul and submarine where every 0.5 dB matters and these are – I won't say the name of the customer, but these are all the well-recognized name in long haul and submarine. The second market is really China where we have strong both OEM customers as well as module customers who want to do module for this Chinese OEM. The third one is the non-China metro, so this represents both North America as well as Europe. We have some strong OEMs in both North America and Europe pulling us into this. The fourth segment is the 1RU cloud boxes, so they have both ASICs as well as standard offering that go into 1RU boxes that are serving the data center interconnect, both the edge where we actually have ColorZ as well as the longer-reach metro and long haul market. And finally, the number five segment is the customers that are building their own modules for DCO-type application. They have a new product coming up which is a low-power product that is getting tremendous traction with module customers wanting a merchant solution for the next-generation DCO modules. So these are the five set of customers, Tore, that are pulling us into this.

[Tore Svanberg:]

That's really helpful. And a question on the Q4 outlook for continuous growth. Can you maybe identify some of the key drivers there? I assume you get the continuous penetration from ColorZ but any comment on the driver business or the TIAs? Just trying to understand the actual growth in Q4.

[Ford G. Tamer:]

Yes. So if you look at Q4, let me divide it up into the key segments. So maybe first if I started with the TIA and driver, we're seeing a few things happening there. Number one, ACO continues to ramp from a small base. We have the DCO and China continues to ramp as well in a very nice way. There's obviously one major customer who obviously is the leading in that segment, but we expect several new customers to come in with DCO offerings. We have what has been an interesting trend which we now better understand, the emergence of what we call daughter cards in the China market. So in the China market, some of the customers, to be able to go to market faster are deploying daughter cards that are sort of, instead of ACO, they're putting daughter cards in some of these systems. So these are sort of the three types of form factor, the ACO, DCO and daughter card in this optical market. Regarding the speed, 100 and 200 gig is still the workhorse of what's being deployed, so our 32-Gigabaud driver in this SMT format I just talked about is really going to be one of the fastest growing drivers in the Q4 and into 2017. We're starting to see 400 gigs of flex coherent starting to ramp and we expect it to grow nicely into 2017 and then 2018. Finally, the latest and greatest 64-Gigabaud is starting to be in field trial and we expect that to be, again, starting to contribute in 2017 and going to 2018. On the networking side, what's probably not has been as well understood is everybody's focused on PAM, but really where the action is today is really this NRZ 100-gig. So we're seeing a tremendous amount of growth in both our CFP and CFP2 which is more of the traditional router market for both the router product of CFP and CFP2. And then we're seeing the QSFP grow with our CDR product so that's starting to grow. And finally, we're starting to see PAM being deployed on 40 gigabit toward the end of this year and next year into 200 and 400 gigabit. That's sort of what's happening in Q4. Obviously, we also have a 10 and 40-gig client that goes into server platform that is starting to come up, and finally ColorZ. So quite a few growth vectors into Q4, Tore.

[Tore Svanberg:]

Very good. Just one last question if I could. You mentioned CDR ramping and QSFP. Will see any CDR contribution from CFP2 at all from Inphi next year? Or should I still be thinking primarily QSFP?

[Ford G. Tamer:]

No, you're 100% right. I'm sorry to have missed it. You are right. The CDR actually, the biggest market right now is that CFP2. And so that CFP2 is growing very nicely for our CDR product and specifically was one large customer, one system OEM who does their own modules there, that's been a very nice growth area for us.

[Tore Svanberg:]

Great. Congratulations again, guys. Very well done. Thank you.

[Ford G. Tamer:]

Thank you, Tore.

[Operator:]

Thank you. The next question comes from the line of Wayne Loeb from Cowen. Your line is open.

[Wayne Loeb:]

Hello, thank you for taking my question. Can you talk about your plans and timing to expand your position inside of data centers beyond the existing revenue streams? How significant is the contribution from 100-gig gear boxes CDR now? And recently, the company talked about 10% of core communication revenue was coming from within data centers. Where do you think this can go in a year, two year horizon?

[Ford G. Tamer:]

Thank you, Wayne. I'm going to take the benefit maybe of a new corporate pitch that has been uploaded to our website, so if you go to the About Inphi page and you can pull it up from there. It's called Corporate Update, or it's also under Investor section. So it's in a couple of places. And if you put that corporate overview, and if I take you to two slides that are critical here, slide 34 and 35, we've been getting this question a lot, Wayne, about what's inside data center. And that's a good summary to help walk people through this. So if I go to slide number 34, you can see that we have categorized the markets into enterprise, service provider, and data center. And so first, if we go into enterprise, what we ship in enterprise today is both 10-gigabit and 40-gigabit PHY that sit on the motherboard next to the server processor. And that's a nice inside data center type of revenue that we have today. The second one is the service provider, and there we're seeing the gear box product, as well as the CDR product. And those go into CFP as well as CFP2 that go into router that are the edge of the data center. So again, those are products that have been growing nicely on more of the NRZ or 01 type of modulation format. We're now starting to see a wide adoption of a PAM into quite a few router market with what's called a CFP8. And that's really for 400-gigabit. And that's going to be going and starting to ramp early next year, with again, quite a few customers at the edge of the data center. Then if we go into data center, per se, on just a more traditional what people think of data center, the first two are one on server, one on the router, so they're both in data center, but this last category is maybe more what people think of data center. There we're in both QSFP and SFP today, longer term committed to support OSFP and QSFPDD, which are the 200-gig and 400-gig format. So on the QSFP and SFP, this is ramping today with the CDR platform and is going to move through the PAM platform. What we're excited about is IEEE just adopted the PAM for 50-gigabit per second format. And that's really going to help standardize PAM and provide a standard for 200-gigabit inside the data center. And 400 gigabit could be either way. It could be either 4 by 100 or 2 by 200. So I think we'll support both. So this added standardization of 50-gig and 200-gig is really going to help this nx50-gig PAM4 increase momentum and increase industry adoption. The final thing I'd say is the next generation is only going to be PAM. So this will, again, help PAM become the standard in the data center. So that's a brief overview of the different segments we play in, Wayne.

[Wayne Loeb:]

Okay. My follow-up question, so China has been a real driver for core communications growth and you talked about how this year was the national system being built, the national infrastructure, next is going to be city and metro. But overall, can you give me an opinion on what inning you think we are in the high-speed China buildout?

[Ford G. Tamer:]

We're one year into it, so your job is to tell me how many innings, how many years you think we have. My view is will be a three to four-year type of game, so we are in the first year. So we should have another couple, two to three years to go.

[John S. Edmunds:]

I think Ford also mentioned on the call that the national buildout had largely driven in 2016 and that they're now progressing onto the regional buildout which is more far-reaching and will take longer, I think, for them to build out. So there is, we think, a fair amount of upside yet to go in China.

[Wayne Loeb:]

How do you see the scale of the China metro opportunity compared to the long haul?

[John S. Edmunds:]

Oh, much larger.

[Wayne Loeb:]

Thank you very much.

[Operator:]

Thank you. And our next question comes from the line of Vivek Arya from Bank of America. Your line is open.

[Vivek Arya:]

Thanks for taking my question and congratulations on the good results and the ClariPhy announcement. For my first question, I was wondering if you could give us your target operating model gross margins, operating margins, once you include ClariPhy. So if you have some target model we should be thinking about as you exit 2017 or the back half of 2017, I think that will be very helpful.

[John S. Edmunds:]

Good questions, Vivek. We haven't actually sat down to forecast exactly where we think they'll come out. We do think that they'll improve, probably more dramatically in the fourth quarter 2017 and into 2018. As I mentioned in past calls, we're continuing to invest for growth, and we would expect our operating margins to stay in kind of a high-20s range as we make those investments and hopefully drive more revenue growth as we move forward.

[Vivek Arya:]

Got it. On organic growth, so this year, of course, a very strong 60% plus – I'm talking about the core communications. Ford, do you think as we look into next year, this is still a, I don't know, 45% or 50% kind of growth opportunity including – and I think the expectation is sort of mid-20s million from the ColorZ product. So overall growth opportunity and where in that, how big ColorZ can be from the visibility you have today?

[Ford G. Tamer:]

Vivek, thanks for the question. As you know, we've always like to take a step at a time. So we've delivered some pretty strong growth this year. We're not forecasting any growth of any similar growth next year in the current consensus number or our current forecast. So we'd like to take it a step at a time and get more closer to 2017 and get more evidence that this is really going to be the same strengths in 2017 before we start predicting that. As you heard my answer to prior comment, we do believe there's quite a few significant growth vector that could make this repeatable. But it's too early for us to pound on table and say this is what we're going to do because there will be numbers that are very different. The numbers that we're all forecasting. So we'd like to keep to our more conservative view right now and then go a step at a time. And then maybe get more bullish as we enter 2017 and progress.

[Vivek Arya:]

I understand. That's fair. And just the last one. With optics and the history is that of boom and bust cycles, what do you think makes this cycle different? And is your portfolio adequately diversified now from a product or a customer and a geography perspective? Because there's just been a lot of volatility among your stock and several of your peers because of that concern around the diversity of whether it's customers or end market and so forth. What makes you think the optical cycle could be different this time?

[Ford G. Tamer:]

Yeah, I think there's two factors I think that make it different. The first one in the cloud is these are very large companies with Google, Amazon, Microsoft, Facebook and the BAT, Baidu, Alibaba, Tencent coming online in China; they are spending real CapEx. The latest number I have on this CapEx this past quarter is that Amazon, Microsoft, Google combined grew significantly in their CapEx. And so this is not any more – this is not companies that don't exist that are not spending real money. These guys have some real profits and after some really large cloud computing market segments. So we do believe that's one large difference. The second one on the telecom side is we believe the metro and long haul will be followed by very significant opportunity in 5G wireless. So we believe the 5G wireless opportunity would be significantly higher than metro and long haul. And so we're focused on that opportunity as well. So on both the telecom side as well as the data center side, we believe that there's some differences this time. The third one I'd mention is, on the last time around, the optical buildup happened and the enterprise went and deployed copper. Here if you look at the cloud data center spending and deploying this current generation, and they're not going to stop at 100, they're going to go to 200 and 400, at that time, the enterprises are going to start adopting 25-gig servers. Today, they're mostly on 10-gig servers and from there they're going to have to go to optical as well. So the enterprise, if they're going to move to 25-gig server, is going to have to deploy this same optical type of interconnect that the cloud guys are deploying. The copper is not going to cut it anymore. Hence we believe the enterprise following the cloud is yet one more reason why this is a more prolonged cycle than prior cycle. I have here the numbers for the CapEx spend for Amazon, Google, and Microsoft and their CapEx is over 33% year-on-year. So that's significant increases year-on-year on CapEx spend for this last quarter.

[Vivek Arya:]

Okay. Thank you.

[Operator:]

Thank you. And our next question comes from the line of Richard Shannon from Craig-Hallum. Your line is open.

[Richard Cutts Shannon:]

Well, hi, Ford and John. Congratulations on great results and thanks for taking my questions. A question on ClariPhy

[Ford G. Tamer:]

Yes. If you look at ClariPhy today, they have about 20% share on this $200 million and we're being pulled into this by Tier 1 customers. So the names that are pulling us into this are Tier 1 customers absolutely, on both the OEM side as well as the module side. So we've got a set of customers on system OEM and module pulling us into this opportunity on the merchant silicon now. In addition there is a significant opportunity for us to do ASICs for the same set of customers, so they may not necessarily want to have the same solution as everybody else. So this gives us two opportunities. Number one opportunity is for us to do an ASIC for them versus purely their own ASIC. The second opportunity is because ClariPhy owns its own back end, ClariPhy has been and is able to include some proprietary IP for this particular customer that's not available to anybody else. The good news going 16-nanometer and going to 7-nanometer, that chip looks like the size of multiple football fields. So you've got plenty of space to include customer specific IP into these chips that allow that customer to differentiate compared to everybody else. And so actually, that's one of the main reasons ClariPhy is winning some of these future generation is because these customers can then put specific IP into different shapes. I mean, it's not just a vanilla DSP. It's their own IP that's embodied into this DSP.

[Richard Cutts Shannon:]

Okay. Great. That's very helpful, Ford. Thanks for that. Second question for me is on your linear drivers and amplifiers. Inphi's had a historically very strong share in the amplifier side. And I think John's comments talked about relatively higher growth on the driver side. I assume mostly from the linear versions of those. Can you talk about where your share is, where it is now, where it could go with these linear products? It sounds like you're having some great success. Any way you could characterize or quantify, that would be great, please.

[Ford G. Tamer:]

Sure, Richard. Thank you. So our linear driver grew about 100% year-on-year, so very fast. So, obviously, we're growing much faster than the market. And that seems to indicate we're taking share. And this is in a couple of places we're taking share in the transition of the industry from limiting to linear. We're also taking share because in addition to our gold box driver, we've now introduced the surface-mount technology, or SMT driver, that are now fully qualified and ramping to mass production and, again, taking share. So I think a few factors that are allowing us to take share here.

[Richard Cutts Shannon:]

Okay. That's helpful. And last quick question for me, guys. Module formats, DCO versus ACO, which one do you think will be a bigger driver overall for your business going forward for the next few quarters?

[Ford G. Tamer:]

Here is the good news. We want to be totally neutral. We're a component vendor. We want to support our customers. And so we've got customers doing ACO. We've got customers doing DCO. We've got customers doing daughter cards. And we've got customers wanting to buy just discrete parts to do whatever they'd like to do. And so we support all of the above. And we're not going to take sides. I mean, our job is to provide components. And we keep telling our teams and ourselves that we're only successful if our customers are successful. So we're a component vendor. The only way we make money is if our customers ship their module or ship their systems. So we support all of the above. And again, same strategy on the components, we want to support all the components. So we want to have TIA. We want to have driver. We want to have DSP. We want to have optical PHY. We want to have silicon photonics. We want to have all the components needed by our customers to build these modules or line card or daughter cards. So that's our strategy.

[Richard Cutts Shannon:]

Okay. Very helpful. Keep up the good work, guys. Thanks a lot.

[Operator:]

Thank you. And I'm showing no further questions in the queue. I would like to turn the call back over to John Edmunds for closing remarks.

[John S. Edmunds:]

Thank you, Ciara. Ford and I would like to thank everyone for joining us today. We plan on attending the Needham Next-Gen Storage and Networking Conference in New York tomorrow on November 2 at the Millennium Broadway Hotel. We'll be at the Stifel conference in Chicago next week on November 10. And we'll be at the Credit Suisse conference in Arizona on November 30 through December 1. We'll be at the Raymond James conference in New York on December 6, the JPMorgan Access Day in San Francisco on December 8, and the Bank of America conference in Boston on December 15. Again, we'd like to thank you for joining us today. And we look forward to speaking with you in the future.

[Operator:]

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