祥茂光電 (AAOI) 2017 Q3 法說會逐字稿

Good afternoon. I will be your conference operator. At this time, I would like to welcome everyone to Applied Optoelectronics Third Quarter 2017 Earnings Conference Call. (Operator Instructions) Please note that this call is being recorded.

I will now turn the call over to Maria Riley, Investor Relations for AOI. Ms. Riley, you may begin.

Thank you. I'm Maria Riley, Applied Optoelectronics Investor Relations, and I'm pleased to welcome you to AOI's Third Quarter 2017 Financial Results conference call.

After the market closed today, AOI issued a press release announcing its third quarter 2017 financial results and provided its outlook for the fourth quarter of 2017. The release is also available on the company's website at ao-inc.com. This call is being recorded and the webcast live. A link to that recording can be found on the Investor Relations page of the AOI website and will be archived for 1 year.

Joining us on today's call is Dr. Thompson Lin, AOI's Founder, Chairman and CEO; and Dr. Stefan Murry, AOI's Chief Financial Officer and Chief Strategy Officer. Thompson will give an overview of AOI's Q3 results, and Stefan will provide financial details and the outlook for the fourth quarter. A question-and-answer session will follow our prepared remarks.

Before we begin, I would like to remind you to review AOI's safe harbor statement. On today's call, management will make forward-looking statements. These forward-looking statements involve risks and uncertainties as well as assumptions and current expectations, which could cause the company's actual results to differ materially from those anticipated in such forward-looking statements. You can identify forward-looking statements by terminology such as may, will, should, expects, plans, anticipates, believes or estimates and by other similar expressions. Except as required by law, we assume no obligation to update forward-looking statements for any reason after the date of this earnings call to conform these statements to actual results or to changes in the company's expectations. More information about other risks that may impact the company's business are set forth in the Risk Factors section of the company's reports on file with the SEC.

Also with the exception of revenue, all financial numbers discussed today are on a non-GAAP basis, unless specifically noted otherwise. Non-GAAP financial measures are not intended to be considered in isolation or as a substitute for results prepared in accordance with GAAP. A reconciliation between our GAAP and our non-GAAP measures as well as discussions of why we present non-GAAP financial measures are included in our earnings press release that is available on our website.

Before moving to the financial results, I'd like to announce that AOI management will attend a Needham Networking & Security Conference on November 14, the Craig-Hallum Alpha Select Conference on November 16, the Raymond James Technology Investors Conference on December 5 and the Cowen Networking & Cybersecurity Summit in mid-December. We hope to have the opportunity to see many of you there.

Now I would like to turn the call over to Dr. Thompson Lin, Applied Optoelectronics Founder, Chairman and CEO. Thompson?

Thank you, Maria. Thank you for joining us today to discuss our third quarter results. In summarizing our performance in the quarter.

Our total revenue increased 27% year-over-year to $88.9 million. Revenue was below our expectations due to a slowdown in demand from a large customer. This resulted in softer-than-expected datacenter revenue of $65.8 million. Despite this shortfall, our datacenter revenue grew 24% over Q3 of this year.

In CATV, I'm very pleased to report that we increased our revenue to an all-time record $18.9 million, representing 47% growth over Q3 of last year. Beyond revenue, we maintained a strong gross margin of 44.4% and delivered an operating margin of 23.2%. This led to net income of $22 million, or $1.08 per diluted share, an increase of 184% over the $0.38 reported in Q3 of last year.

While we are disappointed with our third quarter revenue, we remain confident in our leadership positon in advanced optics. We are also encouraged by the customer traction we are generating with our 100G products, specially our CWDM full MSA spec 100G transceivers. We secured nine design wins in the quarter, including three 100G products and we have approximately forty 100G and 200G qualification efforts underway with various customers, many of whom are outside of our core hyperscale customer base.

In CATV, we continue to expect improved demand as cable MSOs upgrade their new architecture with DOCSIS 3.1. We are working hard to diversify our customer base and are encouraged with the customer response. We also continue to make progress on developing new innovative products and expanding our vertical integration to further extend the gap between AOI and the competition.

With that, I will turn the call over to Stefan to review the details of our Q3 performance and outlook for Q4. Stefan?

Thank you, Thompson. Total revenue for the third quarter was $88.9 million, up 27% year-over-year. Our datacenter revenue was $65.8 million, up 24% year-over-year compared with $52.9 million in Q3 of last year.

As discussed on our pre-announcement call, we saw lower demand overall from one of our large customers. And our revenue visibility in the quarter was impacted by the vendor-owned inventory management model that we employ with this customer. As previously discussed, this VOI program allows the customer to pull inventory from a hub that AOI manages, and revenue is recorded at the time the inventory is pulled. This arrangement can make revenue prediction difficult, but it allows us to maximize sales by ensuring that AOI products are available for customers when needed. For example, AOI has, in the past, been able to meet unexpected demand surges because we had available inventory in the hub for our customer to purchase with little to no lead time.

We continue to have ongoing discussions with this customer and based on those conversations, we believe the disruption in order flow is related to the ongoing transition from 40G to 100G and not specific to AOI. We believe there was some inventory buildup during the transition and based on conversations with this customer, we believe that inventory conditions will normalize within the first half of next year.

Outside of this customer, we continue to experience good demand with our other top datacenter customers. As previously anticipated, we saw a crossover during the quarter between revenue generated from 100G datacenter products, which represented 56% of our datacenter revenue and 40G products, which generated 41% of our datacenter revenue.

Revenue from 100G CWDM products increased by 21% compared to Q2, while 100G PSM products decreased sequentially. Our CWDM full MSA spec 100G transceiver was our fastest-growing datacenter product line in the quarter. And we are shipping volume quantities of full MSA spec 100G CWDM transceivers to several datacenter customers with additional qualification efforts underway.

As a reminder, our 100G product offerings include a full MSA spec CWDM transceiver as well as a lower-cost, lower-performance CWDM light version to provide customers with multiple options to address their cost and performance requirements.

We believe that one of the strengths of our business model is our ability to flexibly adjust manufacturing to adapt to changing customer requirements, while maintaining industry-leading gross margin. As you know, we serve a customer base that is composed of some of the most dynamic, rapidly evolving companies in the world. As their needs change, we think our ability to adapt along with them gives us a long-term sustainable advantage.

We remain in active qualification for our 100G and 200G products. We secured 9 design wins in the quarter, including 3 for 100G products. These include 3 new customers of varying scale as well as wins with existing customers. As a reminder, we only record a design win when the customer has fully tested and qualified our product and we have begun to receive orders from the customer for the product.

Looking ahead, we have approximately 40 100G and 200G qualification efforts underway with various customers, many of whom are outside of our core hyperscale customer base.

As we look ahead to 400G and higher-data rates, we believe that expertise in both high-speed directly modulated lasers, or DMLs, and Electro- absorption Modulated Lasers, or EMLs, becomes increasingly critical for success. Transceivers built on DML and EML devices are typically less expensive to produce than transceivers that use external modulation, including silicon modulators. And we believe that having strong, in-house capabilities for producing these DML and EML devices will allow us to attain class leadership at 400G, as we have developed at 40G and 100G data rates.

As we announced in August, AOI has developed a 50G EML laser that is well suited for 200G and 400G transceiver applications. And we look forward to announcing other innovative laser devices for next-generation solutions, as we reach new development milestones.

Turning to our cable television market. Revenue from CATV products increased 47% year-over-year to reach an all-time record $18.9 million compared with $12.9 million in Q3 of last year. This increase in demand was driven by ongoing upgrade projects primarily related to DOCSIS 3.1. And we continue to anticipate improved demand as cable MSOs evolve and transition to next-generation node and head-end architectures. As a pioneer of innovative fiber-optic access products for the CATV market, we believe our digital forward Remote-PHY product will play a significant role in this transition. We recently developed a new FPGA-based Remote-PHY product in collaboration with Intel and showcased this latest design at the SCTE Cable Tech Expo.

We are encouraged with the customer response we are seeing so far and believe we are well positioned to capture leading market share as this network architecture evolves.

Our telecom products delivered revenue of $3.5 million compared with $3.4 million in Q3 of last year. For the quarter, 74% of our revenue is from datacenter products, 21% from CATV products with the remaining 5% from FTTH, telecom and other.

In the third quarter, we had 3 10% or greater customers in the datacenter business that contributed 37%, 24% and 10% of total revenue, respectively. Based on current orders and forecast from our customers, we continue to expect that we will have 3 hyperscale datacenter customers representing 10% or more of our revenue for the full year 2017.

Moving down the income statement. In the quarter, we maintained a strong gross margin of 44.4%, which was at the high end of the 41% to 45% range that we view as sustainable and also represents an increase of 1,130 basis points compared with the 33% reported in Q3 of last year. The increase in our Q3 gross margin reflects a greater mix of 100G products and improvements in our manufacturing costs. Our strong gross margin is a testament to our deep manufacturing know-how and cost leadership position in the market.

Total operating expenses in the quarter were $18.9 million or 21% of revenue compared with $16.5 million or 14% of revenue in the prior quarter. The sequential increase was mostly due to higher R&D expense as we brought on more engineers to develop our next generation of datacenter and CATV products as well as design certain optical components that we plan to bring in-house to further reduce our costs.

Operating income in Q3 was $20.6 million, up from operating income of $8.1 million in Q3 of last year. Our operating margin in the quarter increased to 23.2%, up from the 11.5% reported in Q3 of last year. Net income after tax for the third quarter was $22 million, up from a net income of $7 million in Q3 of last year. We reported net income of $1.08 per diluted share compared with $0.38 per diluted share reported in Q3 of last year. GAAP net income for Q3 was $19.4 million or $0.95 per diluted share compared with GAAP net income of $17.7 million or $0.97 per diluted share in Q3 of last year. The Q3 weighted average fully diluted share count was approximately 20.4 million shares.

Turning now to the balance sheet. We ended Q3 with $72 million in total cash, cash equivalents, short-term investments and restricted cash compared with $75.9 million at the end of the previous quarter. As of September 30, we have $74.6 million in inventory, an increase of $14.9 million from Q2. Our cash generated from operations totaled $8.1 million in Q3, bringing our year-to-date cash generated from operations to $57.5 million. We made a total of $20.4 million in capital investments in the quarter, including $17.5 million in production equipment and machinery and $1.5 million on construction and building improvements. This brings our total capital investments year-to-date to $46.5 million. We now expect to spend roughly $75 million in capital investments this year compared with our projection last quarter of $85 million.

Moving now to our Q4 outlook. We expect Q4 revenue to be between $81 million and $90 million. We expect Q4 gross margin to be in the range of 41% to 43%. Net income is expected to be in the range of $16.6 million to $19.5 million and EPS between $0.82 per share and $0.96 per share using a weighted average fully diluted share count of approximately 20.3 million shares.

With that, I will turn it back over to the operator for the Q&A session. Operator?

(Operator Instructions) And our first question comes from Simon Leopold with Raymond James.

Got a bunch. I'll try to be focused and give other folks a shot. But just couple of things I want to make sure we clarify. One, is the top customers in the quarter, I'm assuming that the largest contributor, that 37% customer was a 10% customer in the second quarter as well as the third quarter, is that a correct assumption?

Yes.

Great. That's very helpful. Pleased to hear that. The other thing is -- and I'm assuming I've entered formulas correctly, so apology if I botch this. But I think the 100 gig business was down sequentially. So given that the primary explanation is a pause of 40 to 100 gig, I would have imagined you'd sell every piece of 100-gig gear. So maybe help us understand if: one, I did the math correctly; and two, if so, why would 100 gig be down in your third quarter.

Well, I think we saw an overall decline -- as we mentioned in our prepared remarks, we saw an overall decline in business from one customer. So that included both 40 gig and 100 gig. On the other hand, the other customers that we had, were actually up for 100 gig, so that didn't quite balance out the overall decline from the one customer.

Okay. And then maybe the last one for me then is in terms of the forecast you've offered, I generally expect that there's some seasonal decline in the cable spending, which, to me, suggests the datacom should be up a little sequentially. But within this, particularly in light of some of the bare arguments about losing additional customer position, what's your assumption about the customer mix that you expect in the fourth quarter in the datacom business?

So I can't -- we have, obviously, nondisclosure agreements with our customers, so I can't really talk very specifically about any particular customer. We do think that we are -- we remain a major supplier to all of our major customers for their long-reach transceiver needs and intra-datacenter applications. And I'll kind of leave it at that as far as customer positioning. Obviously, we have ongoing discussions with all of our customers regarding their needs. We remain engaged with all those customers. We have orders from all the customers that -- all the major customers that we have. And based on all that, we've -- we think we've built in an appropriate forecast for them for the fourth quarter.

So is it safe to say that you expect a somewhat similar mix in the fourth quarter as the third quarter? Is that a reasonable assumption?

It's reasonable. I can't comment on the exact mix. I mean, it's reasonable to assume that it's not vastly different. In other words, I mean, there's no customer that was -- let's put it this way, there was no customer that is sizable in Q3 that is, say, significantly lower in Q4 or lower to the point of nonexistence or something in Q4.

Our next question comes from Paul Silverstein with Cowen and Company.

I'm going to also ask for your patience. I've got a number of questions. First off, Stefan, I apologize. I know you gave some of the details on the call, but my ability to type, is -- does not keep up with your voice speed. Can you give some of the numbers of the 100G and 40G splits with respect to CWDM4 and PSM4?

Yes. I think I may...

I think I heard you say that 100G was up -- 100G CWDM4 was up 21% and you said that the PSM4 declined sequentially, but didn't give a number. Can I trouble you for that number?

So 100G CWDM increased by 21% compared to Q2. And yes, we did not give a specific number for PSM in the quarter. We did say that 100G datacenter products overall were 56% of our datacenter revenue and 40G products were 41% of our datacenter revenue, which is consistent with what we've said for several quarters that we expected those 2 products to cross over in terms of revenue during the quarter.

And Stefan, in terms of not quantifying their client, PSM4, I would love for you to give us the number, but any reason why if you're not going to give it any reason why?

Those particular reasons.

And you don't want to give it.

No, I don't, Paul. Thanks.

All right. Let me ask you about visibility. So you've got VMI with what I presume is -- what has been your largest customer. How much -- in terms of the quality of your insight, it's always hard to predict the exact timing, et cetera. But in terms of the quality of your insight and the confidence you have in your positioning relative to where your customers are going, with their deployment plans? Can you revisit with us qualitatively what is the quality of that insight? Why do you have the confidence you do? What, if anything, have you learned from, perhaps, some of the miscues along the way, the recent miscues, if there is anything to be learned.

Well, I think it's fair to point out that while we're obviously disappointed in this quarter, we've had 5 straight really good quarters in a row. And if you go, kind of, look under the hood a little bit, which I'm sure you'll do once you have a chance to look at our numbers, I mean, what you're going to find is that compared to last year, let's say, for the 3 quarters in early -- leading up to this, in other words, Q1 through Q3 of 2017, our datacenter customers, besides what had been our largest customer, are up 180% year-over-year. So what we're looking at is a more diversified customer base than what we have before. And that helps us in terms of giving visibility. Now obviously, we still have a customer -- several customers that represent a sizable portion of our revenue. But to the extent that we can diversify that revenue more among different customers, that helps our visibility somewhat. So we're no longer as tied to a particular customer in terms of living and dying by that visibilty. In terms of kind of looking at it in a little more detail, we continue to have discussions with all of our major customers. As I said earlier, we continue to get orders from these customers. We've said all along that it's very difficult to predict, on any given quarter, the exact mix between customers and products and things like that. We do believe that we remain a major player for our hyperscale datacenter customers with their long-reach transceiver needs. And we believe that those needs are going to continue in the future. The problem is that they don't always continue in a linear fashion for each customer individually. And sometimes, we get caught in a situation like we did in this quarter with the transition from one technology to another.

And Stefan, with respect to each of the Amazon, Facebook and Microsoft, I assume you and your colleagues are speaking directly to counterparts within those organizations as opposed to a third party purchasing on behalf of those entities. Whether or not you're speaking to those third parties, I assume your conversations are directly with each of the web (inaudible) companies in question and that's the basis for the statements that you've made and are making about your relationships with each.

Yes. So we have direct conversations with our large hyperscale datacenter customers.

All right. And finally on this and I'll take the rest offline. But my final question is blind, now that Intel has come in and some others, your view of your competitive positioning in 100G, in CWDM4, in PSM4. And obviously, PSM4 was down. You're not giving us the number. But your view of your competitive position with respect to where Facebook, Amazon and Microsoft are going relative to Intel and the others, you're still confident that you're going to see meaningful growth as you go forward. I mean, the question is what is -- can you give us any insight in terms of the outlook there? I know there's sensitivity with each one of those customers, but maybe without giving the customer names, you can revisit what your outlook is and I'm not talking just one quarter.

Well, what I could say is this. I've said a couple of times already that we believe that we continue to be a major supplier for all of our hyperscale customers with their long-reach transceiver needs. We know that those needs -- or we believe that those needs are going to continue based on the discussions that we've had. We've said all along that 100 gig is going to be a competitive business. We've never indicated that we expected to be a sole supplier for any of these things. And I think that we continue to be pleased with our competitive positioning. What really matters in this business is being the low-cost supplier. And I think we've engineered our company and we've talked about this for several quarters to be the low-cost supplier. I think you can see from our gross margin, for example, that the flexibility that we built into our manufacturing processes allows us to adapt to customer changes, and we highlighted that in our prepared remarks, as well. So yes, I continue to believe we're well positioned here for the hyperscale business that we have.

Our next question comes from Alex Henderson with Needham & Company.

I was hoping you could talk a little bit about the build in your inventory in the quarter. What's that composed of? Is it mostly 40 gig components? Are those components built in a somewhat customized way for a particular customer would make it difficult to sell to other customers? How do you plan on working that inventory back down? And does that have any impact on your margins in forward periods?

So the inventory is a mix. There's various different components that are in there. It is predominantly 40 gig. And we -- in the conversations that we've had with this customer and other customers, we believe that, that 40 gig will be consumed over time. And we indicated that, that time frame extends out into the first half of next year.

So do you think that you'll have to discount it in order to move it or you think you can move it at normal margins?

We don't anticipate having to specifically discount the inventory below pricing that we've offered for similar products already.

Okay. And then, if I could, when you're looking at the customer weakness in the quarter, could you talk a little bit about whether you think that it's a function of them not being ready to deploy 100 gig and, therefore, having a stutter in overall demand or whether there was displacement at the customer or whether this was a function of them qualifying a system that might have not been prior qualified that caused slowdown in their decision? How would you characterize what's caused that weakness?

Alex, I really can't. I mean, we have nondisclosure agreements with all of our customers and I cannot comment on customer-specific issues like that. I'm sorry.

All right. Have you seen any of your customers slow down 100 gig as a result of the need to requalify the Arista products as a result of the patent workarounds? Did that have any factor in any of your customers during the quarter?

I haven't heard that issue come out, but I couldn't say that it didn't happen. I mean, that's not something we would ordinarily be apprised of necessarily.

And Alex, this is Thompson. They need to clarify. When we got order from the customer for VOI, that's hub order, then the price is set. So we are not -- when the customer will pull the inventory into revenue, the price will not change. Okay?

All right. So as long as that same customer eventually pulls that inventory through that hub.

Yes.

I see. Are you expecting -- can you give us any sense of what you're expect rate of decline in overall pricing to be over the next couple of quarters? Any sense of industry pricing trajectory?

Yes. We don't really give any kind of forward pricing guidance. I'm quite certain our customers and our competitors would be very interested in our viewpoint on that. And I'd rather keep that little closer to the vest.

(Operator Instructions) And our next question comes from Richard Shannon with Craig-Hallum.

I guess, a few for me. I'm going to follow up on one of the topics Alex has brought up regarding inventory. How fast do you expect the overall inventory levels to come down through the fourth quarter? And when do you get back to kind of normal levels?

We think we'll be back to normal levels in the first half of next year.

Okay. And I would assume, given what seemed like a fairly rapid change during your third quarter that you brought utilization in your internal factories is down a little bit. Has that had any noticeable impact on your gross margins?

Yes. We have brought the capacity down a little bit. And it did impact our gross margins a little bit. There's 2 impacts. I mean, as you could see our gross margin, we guided down slightly from Q3 in Q4 and part of that is due to capacity utilization and part of it is due to price reduction.

Okay. So as I look at your guidance for the fourth quarter here, your revenue level at the midpoint is, if I look back here, in history, is roughly comparable to the fourth quarter of last year. And your mix in that quarter of datacenters is probably somewhat [slower] what I perceive you're modeling for this quarter as well. And that your gross margin is, I think, at midpoint roughly 500 basis points higher with a little bit of lower than normal utilization, it seems here. What can you describe in terms of the reasons for that much improved gross margin outlook? Is that, meaningfully better cost structure? Any sort of mix underneath there? Anything you can describe?

No. I mean, I think we've said pretty consistently, ever since, at least as far back as our analyst day last summer, that is the summer of 2016 that we've made great strides in adding automation, reducing costs, enhancing our flexibility of our manufacturing processes and being able to really adapt our process. Now in previous quarters, we've mainly been talking about it in the context of -- as revenue grows and as we increase and improve our outlook, how fast we can adapt. But it also works the same way on the downside, too, that we have a very flexible and adaptable manufacturing infrastructure that allows us to adapt up or down. And so we've done that. And we can do that, obviously, while maintaining gross margins. While they're down a little bit from our peak are certainly healthy and historically very competitive for the industry.

Well, certainly, in the context of the issues here in the near term that they look pretty good. One last quick question from me. The -- you talked about CWDM4 growth in the third quarter, 21%. Does that include both full spec and the light version?

That does. It includes both the full spec and the light version, although as we pointed out, the fastest-growing product that we had in the quarter overall was our full MSA spec 100G transceiver, which we are shipping in volume to several customers, several hyperscale customers.

Right. I'll probably follow up online on that last part there, Stefan. Was CWDM4 full spec a meaningful part of third quarter contribution?

Yes.

And our next question is a follow-up from Paul Silverstein with Cowen and Company.

Stefan, I apologize if this came up from a previous question. But we -- assuming you sell the inventory, that big jump up in inventory, is that going to adversely impact gross margin? And why not?

Well, Thompson discussed already. And I've said the gross margin on these products is pretty much already decided. So we understand what the gross margin is going to be and we've built that into the guidance that we've given in Q4.

Our last question comes -- is a follow-up from Alex Henderson with Needham & Company.

When you put stuff in a hub, a lot of companies have clauses that if it's not pulled by x amount of time that it automatically is sold. Is that something that's in any of the clauses you have with any of your customers? Or is that something that is not part of your modus operandi?

Broadly speaking, I could say we do have arrangements like that with some customers, but I cannot comment on any customer specifics or which customers have what type of arrangements, obviously. Again, nondisclosure agreements are pretty clear on that.

Second question. Can you explain why the tax line went negative in the quarter? And what's your thinking on tax line for the December quarter?

Yes. So there's some discreet items in the quarter that gave us a tax benefit there. And broadly speaking, what I would expect is about a 22% tax rate, excluding those discreet items.

Yes. So when you -- when I'm looking at your guidance for the December quarter, are you -- I assume you're not assuming a 20% -- 22% tax line in that quarter. Is that correct? You might have additional discreets.

Yes. We expect -- we've taken some estimates on what we expect the discreets to be in the quarter and we've considered...

If those are totally unforecastable externally, can you give us some guidance of what that might look like? I mean, we talked about a negative number again in the December quarter.

I don't have that number in front of me. I'd have to take offline.

It's totally unforecastable externally, so I mean, it's -- it -- I would assume that there has to be some adjustments there. Or asked another way, could you give us some sense of what your OpEx is going to be like sequentially?

Broadly speaking, on the taxes, I would expect it will be somewhat similar to Q3. It varies -- it is hard to predict even for us because it varies with things like stock price and other things. But very broadly, I wouldn't expect a change too much from Q3. And your second question was, I'm sorry, about OpEx. We don't give guidance for OpEx either.

Yes. So I guess, what I'm trying to figure out is that when I plug in either the high number on your revenues and the high margin, I'm coming out with numbers that are quite different than what your guidance is at the bottom line, and I'm trying to reconcile that. Is it reasonable to think that your OpEx spending would actually come in somewhat in the December quarter? Or is that something that you're going to continue to grow and spend -- spending growth should persist?

Well, we are going to continue to spend. I would suggest that -- I mean, if your model isn't working out, I would suggest that the tax might be a place for where you need to take a little bit of a harder look on there because I don't think there's -- I mean, there is going to continue to be increases in our OpEx related to (inaudible) compliance and other things, continuing R&D spend and things like that. So if you're coming out with a result that's strange, maybe the tax is a good place to look.

Right. So I should assume continued OpEx, at least, at the current level and probably some growth is what I think I'm hearing you say.

Yes. Broadly speaking, I think that's correct.

All right. Okay. And then just going back to the architectural challenges here. Are you seeing any of your customers move radically to a PSM architecture where PSM is pushing past 500 meters into the 2-kilometer distances or changing the architecture in the environment in any meaningful way that would alter your competitive position relative to your bias towards CWDM4?

I'm not aware of any sort of fundamental or major structural changes to the architectures that our customers are employing. Broadly speaking, again, I can't go into specifics, but there's nothing large and fundamental like that, that I would attribute...

But actually for [1 eco start] DM transceivers since the price is going to be more and more tricky. So actually, more customer are moving into CWDM, not PSM4.

And then similarly, at the upper end of the spectrum, I've heard that there's a bias towards taking CWDM4 further than the 2-kilometer level up into the LR space because, obviously, there's a much more attractive price point. Are you seeing use cases where your CWDM4 is pushing out into the LR space?

Yes because the price is more attractive, so some customers are talking about to use CWDM to even up to 10 kilometer.

Our last question comes from Simon Leopold with Raymond James.

I wanted to follow up on the -- you talked about design wins. I think 9 design wins, 3 of, which were 100 gig. And so that, in my mind, sort of begs the question of what are the other 6? Are they cable-oriented, 200 gig, 40 gig? I'm just curious why you didn't highlight the breakdown beyond the 100 gig, first question. Second one is I just wanted to see if we can get a little bit more color on traction and opportunity of landing business with OEMs, not necessarily specifically, but this would be datacenter sales to the likes of Cisco, Juniper, Arista, if you could help us understand sort of the process and time line for those opportunities.

Yes. Thanks, Simon. So we do not have any design wins in this quarter with 200 gig. That's obviously a technology that we do have under qualification, as we mentioned with other customers, but we have not logged a design win on that yet. The other products that were -- the 6 other products that were not 40 gig or 100 gig are a variety of other products, some transceiver products for telecom, some for cable TVs you mentioned and other things. As far as the ongoing activity that we have that we talked about, we do have datacenter products with under qualification now with 40 -- or 40 different products under qualification with various different customers. And that's the result of some very diligent work that we've done to try to expand our customer base and to diversify that customer base. And obviously, we'll be reporting back on that as the results from that -- those qualification efforts come in.

Okay. And one last one. In terms of your expectations for the fourth quarter, we've seen this nice steady trend of the mix shift towards 100 gig. So this quarter, you talk about 56% of the datacenter segment coming from 100 gig. How do you expect that to trend in the fourth quarter and in 2018?

Well, we can't give specific guidance. I mean, I think it's clear -- it should be clear to everybody that 100 gig is going to continue to grow. I mean, that's definitely the future. As far as what's going to happen in each particular quarter, we can't really give that level of granularity at this point. But clearly, the long-term trend is for 100 gig to predominate.

So directionally, we should assume that it trends higher in the fourth quarter and in 2018. Is that fair?

Like I said, I can't give you specifics on sort of quarter by quarter. But I think the trend clearly is for more 100 gig than 40 gig into the long-term future.

And at the same time, we believe our overall market share in long-reach transceiver will increase next year in datacenter market.

At this time, we have no further questions. And I'd like to turn the call over to Dr. Thompson Lin for closing remarks.

Okay, and thank you for joining us today. As always, we thank our investors, customers and employees for your continued support.

The conference is now concluded. Thank you for attending today's presentation. You may now disconnect.