Inphi Corp
Q2 2017 Earnings Call Transcript

Published: 8 Aug 2017

[Operator:]

Good day, ladies and gentlemen, and welcome to Inphi's Second Quarter 2017 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 follow at that time. As a reminder, this conference is being recorded. I would like to introduce your host for today's conference, Ms. Deborah Stapleton, Investor Relations. Ma'am, please go ahead.

[Deborah A. Stapleton:]

Good afternoon, everyone, and thank you for joining us today to discuss Inphi's financial results for the second quarter of 2017. I'm Deb Stapleton, and 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 third quarter of 2017. Then, we'll be happy to take your questions. John?

[John S. Edmunds:]

Thanks, Deb. Please note that during the course of this conference call, we may make projections or other forward-looking statements about Inphi, including references to our prospects and expectations for 2017 and beyond, the projected growth and size of our markets and our customers and market share, new products and design wins, as well as the integration 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 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 Form 10-K and Form 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 company website 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. You may notice that we are reporting, based on continuing operations in 2016, to reflect our Q3 2016 sale of our memory business. This is being done in accordance with GAAP rules for reporting discontinued operations. In general, the numbers we refer to in the conference call will be for continuing operations. 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, and thank you for joining us today. Let me start with a few highlights from Q2. Our Q2 revenue was slightly higher than our forecast, with revenue of $84.4 million, representing 39% year-over-year growth. Despite headwinds in China metro and long haul, we still delivered strong year-over-year organic revenue growth of 21% in Q2, excluding the revenue from the ClariPhy acquisition. Demonstrating the resiliency of our financial model in a quarter where market demand was challenged, our non-GAAP operating income was 20.8% in Q2. After taxes, we achieved $0.35 in non-GAAP earnings per share ahead of consensus, which represents 12% growth over our earnings per share in the same quarter one year ago. Our inter-data center interconnect solution ColorZ continued to ramp, crossing 10,000 cumulative unit shipments by end of June with excellent quality in live field operation. The majority of these shipments was to our lead cloud customer. However, we also shipped a significant number of units to other customers and believe to be on track to have a second customer start production by the end of this year. We continued to increase our design win momentum with customers worldwide across our 50 Gig PAM, 100 Gig PAM, our new metro coherent DSP and our 64 Gigabaud 16 QAM and 64 QAM capable TiA and driver product lines. That said, as you heard from us last quarter and from others over the past few weeks, we remain patient regarding a full demand resumption in China metro and long haul markets. This is causing us to guide to a flat Q3 revenue at the midpoint and remain cautious about the second half of the year. However, because of our strong product roadmap, accelerating design wins and ramping orders for our inter-data center interconnect product, we remain confident about our differentiated product positioning and the path ahead. Now, let me take this opportunity to summarize what we view as Inphi's strong position in the three markets we serve; metro and long haul, between data centers and inside data centers. I will start with the metro and long haul market segment, which has received the most attention lately due to the pause in the China infrastructure rollout. This segment includes most of our TiA and driver product lines and our newly acquired coherent DSP from ClariPhy. The TiA and driver business represented over 60% of the company's revenue at its peak in Q4 2016 and Q1 2017. They now represent under 40% of Q2 revenue and Q3 guidance. This decrease is due to an over-inventory situation in China, which has been at the forefront of discussions with investors for the past few months. Inphi and the rest of the industry have been unable to properly forecast this inventory situation and the visibility remains poor. Having said that, as the inventory situation improves, we're seeing positive trends emerge. On the driver side, the inventory for our Surface Mount or SMT drivers has mostly been cleared. We have seen a resumption of orders for SMT drivers from the largest OEM in China for its CFP-DCO margin. On the Glovebox driver, we expect a healthier demand in Q4. But until we see a resumption of demand to normal order levels, we'll remain cautious. On the other hand, inventory remains high on our TiA products that go into coherent receivers. That situation may need additional time to clear itself up. Our TiA and driver product categories currently generate about the same amount of aggregate revenue. So this gives you a sense about the timing of the recovery. Furthermore, the general consensus is that we will have to wait until after the China Congress meeting in late September, early October to understand the timing of new large tenders for the next phase of the China infrastructure build-out. Together, these trends explain our guidance for Q3 revenue to be flat at the midpoint and why we're remaining cautious on the second half of the year revenue. However, there's also a good news. Our design win momentum is accelerating in this metro and long-haul market, while Inphi continues to gain share and grow our business on a year-on-year basis. For the first half of 2017, our organic long-haul and metro business grew 36% versus first half of 2016. Our 64 gigabaud, 16 QAM and 64 QAM, TiAs and drivers have garnered Tier 1 designs at large system OEMs and module partners. These new TiAs and drivers support data rates up to 600 gig. We've also started sampling a new 100-gig and 200-gig coherent DSP from the ClariPhy acquisition, which due to this improved power and performance, is gaining traction in the metro and long haul markets. Second, for the between data center market, Inphi has a differentiated solution with ColorZ for cloud customers. ColorZ delivers 100-gig low-power solutions today with a roadmap to 400-gig. In Q2, we crossed the 10,000 cumulative unit shipment, ahead of expectations, more than doubling revenue from Q1. Our lead customer continues to transition from a few large mega data center architecture to multiple smaller data center architecture linked together with ColorZ into virtual, regional data centers. We believe that we continue on track to have a second customer in Q4 and gain further U.S. customers in 2018. We also just returned from China and saw significant interest there as well. Several China system OEMs and cloud providers are now evaluating ColorZ for potential applications in their market. We expect this could represent a significant market expansion for ColorZ in 2018. Inphi also recently won a major industry endorsement from the Optical Internetworking Forum or OIF, the standard body for our Forward Error Correction or FEC for the 400 gig ZR market. ZR stands for 120 kilometer distances between data centers. We expect more ColorZ design wins to ramp this year into the future due to our solution's clear cost, power, density and reduced latency benefit, as well as the scalability from a 100 gig to 400 gig ZR-enabled architecture. Customers can start today with our 100 gig ColorZ product and preserve their investment in line systems, firmware and ecosystem support were a strong roadmap to 400 gig ZR endorsed by OIF. We expect to increase unit shipments significantly by the end of 2017 and into 2018. Finally, in our inside data center business, the market is currently transitioning from 40 gig to 100 gig NRZ, QSFP28 solution. We play in the NRZ market with both a 100 gig Gearbox and a 100 gig CDR Retimer. And while we have reasonable market share, we believe the opportunity will shine for Inphi in the second half of 2018 when we expect the market to transition from NRZ to PAM while we have a clear lead in both DSP and linear TiA and driver technologies. PAM DSP is what enabled ColorZ to go up to 80 kilometer distances for between data center solutions. There has been a lot of noise around PAM with conflicting data points for inside data centers. Inphi has been investing in PAM since 2012. We have the industry most comprehensive portfolio of both 50 gig and 100 gig PAM based optical PHYs and re-timers. We have increased our design wins significantly and continued to gain share with module partners and networking system OEMs serving cloud and enterprise customers. This momentum holds for both our DSPs inside the optical modules and for our retimers' online card. We expect the cloud market to transition to 50 gig and 100 gig PAM DSPs with extra shade of linear TiAs and drivers in the second half of 2018. And that transition should translate into incremental Inphi revenue. We expect earlier revenue from 50 gig PAM in the first half of 2018, driven by 200 gig Infiniband HDR and other specific cloud applications. We also expect our 50 gig retimers to ramp in mid-2018, driven by early introduction of our customer latest switch systems. In summary, we remain confident about the road ahead. We're pleased with the progress of ColorZ in the ZR or between data center market. The introduction of our new coherent DSP for metro and our revenue diversification effort in the cloud data center market. Our revenue and design win funnel is deep, wide and growing. This funnel includes win in all three market segments we serve. Inside data center, between data centers and of course in the long haul and metro markets. As a direct result of our innovative in-house R&D effort and our ClariPhy acquisition, today, we have some of the largest and best-known cloud and service provider customers working closely with us, several for the first time. These close customer interactions, the continued migration to cloud architectures, our leading-edge solutions including ColorZ, PAM and our coherent TiA, driver and DSP solutions and our talented employee pool put us in a strong position to drive a bright future for our company and our shareholders in 2018 and beyond. With that, let me turn over the call to John to give you details about the financials for the past quarter and our guidance. John?

[John S. Edmunds:]

Thanks, Ford. Now, let me recap the financial results. In the second quarter of 2017, Inphi reported revenues of $84.4 million. Revenues were up 39% year-over-year and down sequentially 9.8% or $9.2 million in revenues. Core communications products such as amplifiers, drivers, 10-gig, 40-gig and 100-gig physical interface products, coherent DSP and ColorZ were 83.9% of total revenue, which was down from 88.4% of revenue in Q1. The delta represented a decline of $11.9 million net, primarily from a $25 million decline in amplifiers and drivers, being offset by $13 million of growth in ColorZ and coherent DSP products. The transport and legacy products were 16% of revenues and they were up $2.7 million in Q2, but are expected to decline again in Q3. Overall, we expect transport and legacy to decline only about 5% this year due to end-of-life programs as opposed to previous years when the business has declined more in the 15% to 20% range for the year. Gross margins on a non-GAAP basis in Q2 came in at 70.3% which, as expected, was down about 100 basis points from the 71.3% posted in Q1. This decline was not as large as the approximately 130 basis point decline we had originally forecasted for Q2. Generally, the decline was driven by the mix of product, primarily due to fewer amplifiers shipping as part of the overall mix. In Q3 2017, we expect the gross margins to be up approximately 60 basis points to 160 basis points based on a more favorable mix of products overall. All in, we estimate non-GAAP gross margins for Q3 to be in the range of 70.9% to 71.9%. GAAP gross margins were 56.7%, but include purchase accounting adjustments and stock compensation adjustments which, when excluded, brought the cash based non-GAAP gross margin up to 70.3% for Q2. Q2 GAAP net loss from continuing operations was $15 million, which compares to non-GAAP net income for the second quarter of 2017 of $15.6 million. This $30.6 million improvement for non-GAAP reporting is derived by adding back the following recurring non-GAAP adjustments, stock compensation expense of $11.2 million, purchase accounting adjustments of $13.6 million, acquisition-related expenses of $0.7 million, non-cash convertible debt cost amortization of $6.1 million and then also recognizing an additional tax charge for non-GAAP of approximately $1 million. All of these adjustments then totaled $30.6 million, being added to GAAP net loss in Q2, arriving at non-GAAP net income for Q2 2017. Non-GAAP operating expenses for Q2 totaled $41.8 million, which was down from Q1 or $45.1 million, due primarily from a $3 million NRE fee earned in Q2, which represented customer co-funding of R&D expense. This was the final payment in a particular contract and so will not recur in Q3. Overall, in Q2, we were able to deliver non-GAAP operating margin of 20.8%, which is down 2.3% from the 23.1% we recorded in Q1. The delta coming from a revenue base that was lower by $9.2 million, 100 basis points less in gross margin, offset by $3.3 million in lower operating expense. Interest income totaled $0.96 million in the quarter, which offset the cash cost of the convertible debt of $1.3 million for the quarter, making the net carry cost of the convertible debt about $0.3 million for the quarter. The convertible debt has a blended annual coupon rate of 0.972%. We had other non-operating expense of $0.3 million, which was primarily a year-to-date foreign currency accounting adjustment and combined with the $0.3 million carry cost, amounted to $0.6 million in non-GAAP interest and other income in the quarter. Using an effective tax rate of approximately 8.2% for the year, we booked $1.4 million for the Q2 non-GAAP tax provision. We continue to think that an 8.2% rate is sustainable for the balance of 2017, but with higher volumes should revert back to 8% for 2018. For GAAP income tax expense, we booked a $371,000 tax expense against a loss, implying a negative effective tax rate. This is because we are a taxpayer in Singapore, but have losses in the U.S. and we cannot currently benefit from the U.S. tax losses. We are currently forecasting the GAAP rate for the year to be roughly negative 5.6%. However, this rate generally changes as we move through the year. Cash income taxes paid in Q2 2017 was $161,000, up from the $50,000 we had paid in Q2 2016. Non-GAAP net income for the second quarter of 2017 was $15.6 million or $0.35 per diluted share, which was down sequentially from the $19.5 million or $0.44 per diluted share we reported in Q1 due primarily to the reduction of business coming from China. Now turning to the balance sheet, overall cash was $381 million, down from the March 31 balance of $393 million. We had cash flow from operations of $11.1 million in Q2, compared to $18.5 million in Q1 and compared to $11.6 million in Q2 one year ago. The $7.4 million sequential decrease in cash flow from operations from Q1 to Q2 stems from less income and not being able to be as efficient with working capital while we deal with the inventory accumulation and the relative slowdown in the market. DSO expanded from 51 days at the end of March to 67 days at the end of June, due primarily to the timing of shipments in the quarter. Inventory on the face of the balance sheet increased by $2.2 million in the quarter. However, this was including a change in the respective step-up value from purchase accounting adjustments of $5.1 million at the end of March, becoming $1.5 million at the end of June. When you exclude the change in step-up value, the inventories really increased by $5.8 million. As a result, inventory days excluding the respective step up adjustments was 91 days at the end of March and 119 days at the end of June. The $5.8 million rise in inventory is due primarily to the slowdown on our business stemming from customers in China. Capital expenditures were $6.9 million in the quarter and an additional $5.3 million in software and IP license used for development. Overall, we also paid out a net $13 million in the quarter, driven by buying restricted stock back from employees to satisfy payroll tax withholding associated with vesting in the stock. Now, let me recap the business outlook for Q3, 2017. I'll remind everyone again that the following statements are based on current expectations as of today and include forward-looking statements. Actually results may differ materially. We do not plan to update, nor do we take on any obligation to update this outlook in the future. Due to the continuing inventory accumulation issues in China, revenue is forecasted to be flat in Q3, plus or minus $2 million or (22

[Operator:]

Thank you. And our first question will come from the line of Quinn Bolton with Needham. Your line is now open.

[Quinn Bolton:]

Hey, guys. Just wanted to follow up, first a clarification. John, did you say the ColorZ and ClariPhy business was up $13 million sequentially and the core TA drivers were down $25 million sequentially?

[John S. Edmunds:]

Yes, that's correct.

[Quinn Bolton:]

Okay. Would that change, you guys talked in the script about sort of a more cautious second half outlook, I think in the past, you've given us some sense of where you think the core communications business might grow sort of on an annual basis. Do you have an updated thought now with the third quarter outlook for that core comms business for the full year?

[John S. Edmunds:]

Well, basically, we've obviously guided to something at flat – at the midpoint for the total business and core now makes up 84% overall of that business and we've said that legacy would probably decline a bit. So you can figure that core may grow a little bit. But generally, we think that amplifiers and drivers will remain relatively flat and the ColorZ and DSP will provide some additional growth in the Q3 quarter. So that just leaves us back to flat overall for the company.

[Quinn Bolton:]

Just for the third quarter though?

[John S. Edmunds:]

Just for the third quarter. That's right.

[Quinn Bolton:]

Got it, got it. And you guys roughly 90 days ago on the last conference call, I think, had been concerned about some capacity constraints for ColorZ. Can you give us an update on how you're managing the capacity ramp of ColorZ? It certainly looks like, I assume the majority of the $13 million increase from ColorZ and ClariPhy came from the ColorZ product. So kind of feels like you're tracking now well ahead of your prior target of $40 million to $45 million for ColorZ. But can you give us any updated thoughts on the capacity ramp and perhaps what a reasonable target for ColorZ revenue this year might be?

[Ford G. Tamer:]

Yes. Hi, Quinn. This is Ford. We have been able to resolve the capacity issue for ColorZ. There's still work in progress to even add additional capacity. We're tracking well ahead of expectations for ColorZ and do expect ColorZ to continue to grow in both the Q3 and Q4 quarters and there's no reason to believe that we wouldn't exceed $50 million for the year in ColorZ.

[Quinn Bolton:]

Great. I'll just say congratulations and turn it over to the next caller.

[Operator:]

Thank you. And the next question will come from the line of Ross Seymore with Deutsche. Your line is now open.

[Ross C. Seymore:]

Hi, guys. Thanks for letting me ask the question. Really just want to dig into the China situation. I know visibility is limited on that and Ford, you went through it in significant detail earlier. But wondered if you could just talk about the demand side of the equation versus the inventory side of the equation. I know they're interlinked in half of the business, you've talked about the inventory has pretty much burned down to a normal level. But are there big changes on the demand side or is it just the inventory is too high and basically what we're getting at I think all of us are beating around the bush deciding, is the fourth quarter going to rise when demand comes back or is it going to be pushed out further than that?

[Ford G. Tamer:]

Okay. I think to understand the demand, it's probably helpful to break it up into the trends. So there is a migration from the traditional MSA module to the lower power CFP-DCO in regional and metro. And the CFP-DCO are using linear driver that are placing the limiting driver in these MSA modules and that's really what's growing big time in China and still on track to grow and we are growing our share as a result of that CFP-DCO growth. So that, for us, has been the major growth vector. Just to give you a feel, the ratio right now, our view of – our estimate of the ratio between DCO and ACO was about 3 to 1. So ACO is still growing in North America and CFP-DCO is primarily used in China. And together, if you add up both the DCO and ACO, we feel that's account for a bit – a little less than half of the total coherent port. So, we're seeing a growth in both the CFP-DCO and ACO, so both of these markets CFP-DCO in China and the CFP2-ACO in North America are both growing, and about a ratio of 3 DCO to 1 ACO and together represent about half the market. The inventory situation is mostly being driven by the other non-pluggable, non-DCO, non-ACO type of offering, which is more line card or MSA. And this is where it is still a bit more inventory, Ross, okay, as well as a bit of inventory on the receiver TiA side of the world. As I said, on the DCO or the linear driver, SMT linear driver that goes inside that DCO, the inventory situation has pretty much resolved itself. On ACO, we're growing and really the two places where we still have inventory would be the TiA and the driver that goes on the line card or MSA modules and those would take probably until the end of the year or so to carry itself out. Hopefully that's a good view of where we are, Ross.

[Ross C. Seymore:]

Yes. That's very helpful. And then if we switch gears over to the, probably, the more pleasant side to talk about, on the ColorZ and the ClariPhy side of things, talk a little bit about the magnitude of the upside, I know, John, you said $13 million. If you can just describe some of the dynamics split between what you saw on the ColorZ side and the ClariPhy side? And then looking forward into those businesses, you've given some annual totals where you thought the businesses could get up to, it seems like they're both running ahead of plan. So if you could just give a little color on your longer-term outlook for those that would be helpful as well.

[John S. Edmunds:]

Right. So, ColorZ, we talked on the script that we doubled revenue from Q1 to Q2 and we expect to still significantly grow the revenue into Q3 and Q4. ClariPhy was a nice growth from Q1 to Q2, but we expect this to flatten somewhat into Q3 and Q4 on the average, if you look at the average in Q3 and Q4. So the expectation for ClariPhy for the year is still about the same. The expectation for ColorZ for the year is ahead of expectation. And then, ClariPhy, going into next year, is not going to be immune to some of the metro and long haul issues that we're seeing. So we're probably a bit more cautious on ClariPhy next year. On the other hand, ColorZ next year could be again ahead of expectation given the strong ramp we're currently seeing from the first customer, the continued progress on the second customer and the continued interest in China.

[Ross C. Seymore:]

Perfect. Thank you.

[Operator:]

Thank you. And your next question comes from the line of Harlan Sur with JPMorgan. Your line is now open.

[Harlan Sur:]

Hey, good afternoon. Thanks for taking my question. If we assume the China provision (31

[John S. Edmunds:]

So, Harlan, I think we can do that as you've suggested, but really the issues you have to separate out here are the inventory burn in China completing at one level and then it's a question of what's the natural uptake of business in China; that's the one that's a little harder to get a feel for, but it seems to be improving based on what we've heard about business from Huawei from Q1 to Q2. And this four-year Congress that gets held every year is a little bit of a monkey wrench because they have to stop – the carriers have to stop doing installations during that four-year period – four-week period in the fourth quarter. So that's the thing that people will wrap up projects before that and then wait till after that to begin projects and – but there is some speculation that there can be additional CapEx that will have to spent in the fourth quarter for the carriers and that there's opportunity for revenue growth naturally in that market. But irrespective of that, I think as you laid out, if we are flat into Q3 and flat or even small growth into Q4, with growth in the other parts of our business, we'll be able to show growth in Q4, some modest growth in the low-single-digit percentage kind of range.

[Harlan Sur:]

Great. Thanks for the insights there. And then good to see the margin expansion here in Q3, about 110 basis points, even without growth in the metro long haul business. Can you guys just help us understand the components of the higher margin trajectory? I know your coherent DSP products are ramping here. I know those carry very high gross margins, but anything else contributing to the better margin profile?

[John S. Edmunds:]

Well. I think it's a combination of – as you recognize coherent – ColorZ carries a decent gross margin as well. And so I think overall, the positives that we have, even the transport and legacy is a small growth factor, but also comes with very good gross margins. So, it's really a mix of business in general that's leading to this at this particular time.

[Harlan Sur:]

Great. Thank you.

[Operator:]

Thank you. The next question comes from the line of Joseph Moore with Morgan Stanley. Your line is now open.

[Vinayak Rao:]

Hi. This is Vinayak calling in for Joe. I just want to go back to your long haul and metro business. Can you guys touch upon how big is the non-China part of the business and what kind of growth are you seeing in that part of the business?

[John S. Edmunds:]

Yeah. Vinay, this is John. You'll see in, for the total company, 10-Q will be filed tomorrow, I believe, and I think the shipping to China is 27% or 28% versus having had been 40% previously. So for the company, we're doing about 70% – a little bit better than 70% of the business outside of China. So quite a bit for us goes on outside of China and China is a very big market. So we hope that that will grow and get back into a growth phase itself, but that's the ratio between the two.

[Vinayak Rao:]

Got it. For my follow-up, just want to touch upon your 100G PAM4 within the data center opportunity and Ford did touch upon that and gave us some color like what kind of growth you're expecting? But overall, when you see this market opportunity, like what is the total addressable market that you're targeting there as you ramp up OpEx and R&D towards that opportunity and what kind of competitive profile do you expect for that product over the next couple of years?

[Ford G. Tamer:]

So the best way to look at the market sizing is to look at the transition to 100 Gig QSFP28. And we are not directly in that business. We play in the Gearbox and retimer more for router space as opposed to directly into the 100 Gig for switches today. But the ecosystem players are playing into the 100 Gig switches today are reporting millions of port being shipped into those segments. And as the switches transition to the next-generation PAM-enabled switches, that would provide a 12.8 terabit front panel density, then you'll see the optical module transition to PAM. And there I think you'll have a different cloud provider and different enterprise players picking both 50 gig PAM and 100 gig PAM depending on different architectures. So we have confidence that we have a strong design win position because we offer three different product lines. We offer a 50 gig PAM optical PHY that goes inside the module. We offer a 100 gig PAM optical PHY that goes inside the module and we offer a 50 gig PAM retimer that goes on the line card. So it's pretty unique from the breadths and depths of this offering. It's also pretty unique by the fact that our PAM has been worked on since 2012. So all the analog pieces are very mature. Its introduction with ColorZ at very high volume, powering very major cloud architectures today. And so it's probably the most stable, the most mature, the highest performance, lowest power PAM offering on the market today. So as these PAM switches get introduced to market in the second half of 2018, Inphi should gain significant revenue and market share.

[Vinayak Rao:]

Got it, very helpful.

[Operator:]

Thank you. And the next question will come from the line of Tore Svanberg with Stifel. Your line is now open.

[Tore Svanberg:]

Yes. Thank you. I guess China will happen when it happens. So I'll focus on Chinese question. So, first of all, you mentioned the second customer production by the end of the year. Will you be able to announce that customer once in production or will we have to wait longer for that?

[Ford G. Tamer:]

We don't know, it's not up to us, it would be up to the customer. So we've asked and they're not sure they will. So, maybe they're still trying to decide. So if it does get announced, it'd probably be OFC or post-OFC of next year.

[Tore Svanberg:]

And Ford, just for us to understand the opportunity with that customer, would that be like a similar ramp to what you've had with your first customer or is this smaller volumes initially that become bigger over time?

[Ford G. Tamer:]

I think this is a different application that this second customer is looking into, so this could be a smaller ramp that becomes bigger over time, but the key is introducing that ZR architecture into that customer will be very significant because of a very large U.S. cloud customer. So it opens up the 400 gig ZR that we're working on.

[Tore Svanberg:]

Very good. And on ClariPhy, you mentioned probably being a little bit more conservative for 2018. Can you just elaborate a little bit on that? Are you basically seeing customers outsourcing their DSPs slower or is it just a matter of CapEx being less than 2018 than what you had expected before?

[Ford G. Tamer:]

No. We have some very strong design wins for that new product. We're just sampling a new product for metro that has good power and performance for 100 gig (40

[Tore Svanberg:]

Understood. Last question. Could you just give us an update where you stand on single-lambda PAM4, when you expect to have a product available?

[John S. Edmunds:]

We believe we're going to be extremely competitive, power, availability, performance, linear TiA driver, ecosystem maturity and we'll surprise a lot of people, we don't want to announce. So we'll announce in due course, but all I could tell you is we're getting significant amount of design wins for that product.

[Tore Svanberg:]

Sounds good. Thank you very much.

[Operator:]

Thank you. And the next question will come from the line of Vivek Arya with Bank of America. Your line is now open.

[Vivek Arya:]

Thanks for taking my question. Ford, first question on ColorZ, you mentioned it could be about $50 million or so this year, certainly ahead of where you started. How should we think about this opportunity going into next year as you start adding new customers? Does your first flagship customer continue to deploy and all the new customers, are they additive? On top of that, is there a back and forth? Just what is the right way to size what the ColorZ opportunity is over the next one to two years?

[Ford G. Tamer:]

Good question, Vivek. I think we definitely will see an increase in ColorZ next year, that's something we're confident about. The amount of that increase is currently being worked on. I think we'd have a better handle on it when the second customer goes into production and then when we get a better feel for the incremental both U.S. customers and China customers that are currently under evaluation. So we have a number of cloud customer as well as OEM customers that are looking at private labeling opportunities for these products and so we'd be in a better position to give you more guidance on this probably in the next earnings call, but a nice increase in ColorZ next year.

[Vivek Arya:]

Got it. And then Ford, in terms of the competitive landscape, some of your competitors such as MACOM are coming at the table with ownership of lasers and module making capability and so forth. And I think you have adopted more of sort of a best of breed approach. Can you help us contrast these two approaches? Are you at any kind of disadvantage or advantage in your sort of best of breed approach versus some of the more integration, right, module type approaches that some of your competitors might have?

[Ford G. Tamer:]

Thanks, Vivek. We don't believe we're at disadvantage because we're working very closely with module partners that too have their own lasers and – or can source their lasers. So as far as we're concerned, ownership of the laser is not a requirement. We're going to be providing the – all the different electronics that go into the module. And working very closely with the module partner to go to market. As you take a step back and look at the different optics that are being supported, there are four different types of optics. There is the traditional VCSEL optics for multimode fiber, and there is three different type of optics for single mode fiber, namely direct modulation laser or DML, external modulation laser or EML and then silicon photonics. So you've got four different type of optics that are being brought to the front as far as this inside data center market. And we've got our hands full with providing the PAM, DSP, TiA, driver and in some cases silicon photonics partner, so we – on the laser – we will work directly with a partner or they could source the laser from outside to get the offering to market.

[Vivek Arya:]

Okay. Thank you.

[Operator:]

Thank you. And the next question will come from the line of Brian Alger with ROTH Capital Partners. Your line is now open.

[Brian Alger:]

Good afternoon, guys. Obviously, lot of the questions have been asked. Want to kind of go back just to the big picture in terms of guiding to a flat quarter-on-quarter. Is all that attributable to the inventory situation in China or is there something else that on the margin is also dragging the growth that seems to be coming certainly from ClariPhy and the other businesses?

[John S. Edmunds:]

Thanks, Brian. No, there is no other factor. ClariPhy is growing, ColorZ is growing and on the TiA, driver and optical PHY, we have had some decrease as John explained in his script and you could see the decrease is quite significant and is all attributable to the China situation or inventory situation. And it manifests itself not only at a major OEM customer, but also it flickers down to the different module customers we have. So we get hit twice from the OEM customers having to bring the forecast down. But also from a module customer, they're selling into that same customer base. So, as I said, we expect this situation to clear itself up in the year and then hopefully resume stronger into next year.

[Brian Alger:]

As we look at China specifically, is what's going on in terms of there having been an inventory buildup and the work-down that's going on right now. Is this purely just market forces, typical cyclicality of the industry? Or is there more to it that may be politically inclined? And I guess where I'm getting at is if some of the rumors with regards to a trade war between our two countries emerges, which way would that play out for Inphi in terms of as a supplier into those Chinese OEMs?

[John S. Edmunds:]

So, a lot of ground to cover there, Brian. But I think again – what's helpful I think for people is you want to break out having the inventory burn run its course and curtail here first. And that's happening with some components, other components will take longer. We wish more of those happening sooner, but it's just not enough to turn the tide toward growth in Q3 yet, at least not from what we can see today. So I think that's the main – major thing and then you want to layer in on top of that what's the consumption rate, the natural consumption rate. Again, Huawei's business reportedly was down 19% in Q1. It's come back apparently in Q2 and you can see that overall for the business, Huawei announced that they were up 15% in the first half of 2017 versus the first half of 2016 and they were up in each of their businesses. So there's some improvement there in Q2 versus Q1 for them. Whether that's going to now presumably carry through to Q3 and Q4, I don't know that anybody necessary needs the China tenders to take place for the provincial to drive growth in the back half. But certainly if more tenders happen, that's going to fuel and feed into that. Beyond that, it's – I think it's difficult, I don't know if it's even productive to think in terms of whether a trade war could happen or not happen and what the implications would be.

[Brian Alger:]

Okay. Fair enough, I appreciate the detail there though, John. Thanks a lot.

[Operator:]

Thank you. And the next question comes from the line of Hans Mosesmann with Rosenblatt Securities. Your line is now open.

[Hans Mosesmann:]

Great, thanks. Hey, Ford, can you size the opportunity in PAM4 inside the data center? And also, can you also clarify on the timing in second half of 2018, is that an acceleration, is that being pushed out or just a little qualitative commentary there would be helpful?

[Ford G. Tamer:]

Thank, Hans. So the opportunity in PAM4 should be very similar to the revenue that's being enjoyed today by the players in the market; so you could just take a look at the numbers that are being posted by the semiconductor companies supplying retimers, 100 gig retimers and TiA and drivers into those data center market today and the PAM4 opportunity would be very similar and significant. As far as the timing of it, I think we're still on track. We depend on the PAM4 switches being taken to production and we expect those to go to production in the second half of next year. It's still consistent to where we've been staying in the past.

[Hans Mosesmann:]

Okay. Very helpful. Thanks.

[Operator:]

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

[Richard Cutts Shannon:]

Hi, Ford and John, thanks for taking my questions as well. Maybe first one or two questions on ColorZ, just to clarify or specify here, the $50 million number you said for this year, is that primarily or solely coming from your first customer, or is that expecting any contribution beyond the first one?

[Ford G. Tamer:]

It's primarily from the first one, but it does provide revenue from the second, we already have revenue from the second one in this quarter. So in the Q2 quarter, we had revenue from other customers and we expect those other customers to grow.

[Richard Cutts Shannon:]

Okay. Perfect. And Ford, obviously the ColorZ is addressing a DCI market that's experiencing some pretty good growth here. And when you first talked about the OFC OFC about 18 months ago, I think you evoked a competitive landscape largely looking at OEMs. Have you seen any other competitive offerings unique to the DCI market that come out since then that could close the competitive gap in any way?

[Ford G. Tamer:]

So, as far as the ColorZ competitive positioning for 100 gig, we're still extremely differentiated and unique. I think as you saw from the OIF vote from about a week ago, the standard body have endorsed into our probable FEC for 400 gig. And that should open up the market to a larger market having a dual source and potentially more sources off these DSPs, which is good in a sense that the OEMs and the cloud providers would feel safer about adopting the solution under its multi- source, and only that part is that we'd have to divide the market between us and other competitors. But that's fine as long as the market grows. A rising tide will help all boats. So we're in good shape on 100 gig and I think with a strong roadmap on 400 gig, where the Inphi FEC has been endorsed by OIF as a standard.

[Richard Cutts Shannon:]

Okay, perfect. Thanks for that detail. Maybe one last question from me. Ford, you talked about some strong design win activity for your components going into DCO. Curious if you can segment out where your wins are coming out from OEMs versus module makers and how you see the trends developing between that over the next few quarters.

[Ford G. Tamer:]

Yes. One of the biggest module maker is a captive module maker at one of the largest OEMs in China. And so that's one of the major wins that we have. We also have that SMT driver go into a number of other module maker offerings, okay. So that SMT driver is mostly for module makers. So that's the primary application for it because it goes inside the module as compared to the more gold box traditional driver that used to go on the line card.

[Richard Cutts Shannon:]

Okay, perfect. That's all the questions from me guys. Thank you.

[Operator:]

Thank you. And the next question comes from the line of Dave Kang with B. Riley. Your line is now open.

[Dave Kang:]

Thank you. Good afternoon. So, you've talked about ClariPhy DSP, obviously you'd be a little bit cautious going into 2018. I think previously, you talked about $70 million, $75 million. So what should we think about – what should our expectation would be for next year now?

[Ford G. Tamer:]

Yes. And again – hi, Dave.

[Dave Kang:]

Hi.

[Ford G. Tamer:]

This is Ford. Again, just to be clear, our view would be, we're still as positive about the ClariPhy design wins and momentum as we were before. We're just tempering it because of taking the overall number of coherent port down for the market. So this number of coherent port would influence not only the TiA and driver business, but also impact the coherent DSP revenue, right. And so the number right now should be slightly lower than the number we just mentioned. So, we may be 10% more conservative going into 2018 than we used to be on the prior call.

[Dave Kang:]

Got it. And then you talked about China, 27%, what do you think that number will be exiting third quarter or maybe fourth quarter for that matter?

[John S. Edmunds:]

I'm sorry, Dave, the question again, what do we think?

[Dave Kang:]

Yeah. So you just said China, shipments to China was like – were 27% in the second quarter. So I'm just wondering if that number is going to change in second half based on your cautious outlook for China in second half.

[John S. Edmunds:]

Well, we said that we thought the business would be – for that part of the market would be flat in Q3 and we really haven't guided at all for Q4. So we'll have to wait and see if a recovery does happen. There's signs that it could happen and of course we're optimistic, but I think we need to wait till we get there to call it.

[Dave Kang:]

What about a similar number for North America rough estimate is what like 50%, 60% for drivers and amplifiers?

[John S. Edmunds:]

No. I don't have it broken down...

[Dave Kang:]

Oh, okay. Maybe just total North America then?

[John S. Edmunds:]

Total, what you'll see in the 10-Q tomorrow is 22% is United States.

[Dave Kang:]

Okay. And shouldn't that segment, I mean, your competitors or customers are saying North America is pretty strong. So shouldn't that segment at least grow, so maybe if China is flat, then there should be some kind of a sequential increase in second half or is that a legacy offsetting that?

[Ford G. Tamer:]

Yes, certainly with our amplifiers and CFP2 ACO types of modules that will grow in the quarter, but I think we have to look at the business holistically all the way around and we've left room for China to even get a little weaker before it gets stronger. So I think all things considered, we think flat was the right guide for Q3 and then let's wait and see how – before we try, wait.

[Dave Kang:]

Okay. And just CapEx what was it and what's the budget for this year?

[John S. Edmunds:]

CapEx came in I believe at $6 million and I think we're shooting for about $45 million and $50 million this year in total CapEx.

[Dave Kang:]

Got it. Thank you.

[John S. Edmunds:]

Okay.

[Operator:]

Thank you. And I'm showing no further questions at this time. I would now like to turn the call back over to Mr. John Edmunds for closing remarks.

[John S. Edmunds:]

Thanks, Sabrina, and thanks everybody for joining us today. Inphi plans on attending the Oppenheimer Conference this week on Wednesday on August 9 in Boston, the Jefferies event in Chicago on August 29, the ROTH Conference in San Francisco on September 6, the Deutsche Bank Conference in Nevada on September 13. And again, Ford and I would look forward to joining you and seeing you at these conferences and speaking with you again in the future.

[Operator:]

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