Good day. I will be your conference operator. At this time, I would like to welcome everyone to Applied Optoelectronics second quarter 2018 earnings conference call. All lines have been placed on mute to prevent any background noise. After the speaker's remarks, there will be a question and answer session. To ask a question, you may press star then one on your touch-tone phone. To withdraw your question, please press star then two. Please note that this event is being recorded. I now will 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 second quarter 2018 financial results conference call. After the market closed today, AOI issued a press release announcing its second quarter 2018 financial results and provided its outlook for the third quarter of 2018. The release is also available on the company's website at ao-inc.com. This call is being recorded and webcast live. A link to that recording can be found on the investor relations page of the AOI website and will be archived for one 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 Q2 results, and Stefan will provide financial details and the outlook for the third quarter of 2018.
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. 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 non-GAAP measures, as well as a discussion 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 note the date of our third quarter 2018 earnings call is currently scheduled for Wednesday, November 7th. I would like to turn the call over to Dr. Thompson Lin, Applied Optoelectronics founder, chairman, and CEO. Thompson?
Thank you, Maria. Good afternoon, everyone. Thank you for joining us today. AOI delivered a strong second quarter with results exceeding our expectations, driven by increased demand for our market-leading data center products. In the quarter, we generated revenue of $87.8 million, gross margin of 40.4%, net income of $12.9 million, and $0.64 per diluted share. Data center was a bright spot in the quarter as we achieved record revenue for our 100G products. In CATV, the demand environment continued to improve, and we remain encouraged by the customer activity and engagements we see in the market. We are pleased with the success we have achieved with our marquee customers, still continue to work to diversify our customer base. In the quarter, we secured seven design wins, including one with a large data center operator in China.
We are pleased with our results and continued solid execution in the quarter as the demand environment improved. Looking ahead, we do remain encouraged by the trend we see in the market. We continue to expect 100G volume will more than double in the second half of this year over the first half, which is based largely on the committed order we announced in Q1 of this year. Additionally, we expect 100G volume to double again next year over this year as data traffic continue to grow, requiring data center operator to expand their data centers and upgrade their infrastructure to keep up with bandwidth demand. In CATV, we continue to expect additional Remote PHY and fiber-deep sales that will drive growth in this segment later this year. We remain confident in our growth prospects and competitive position.
AOI has developed and invested in technology and a manufacturing platform that allow us to be the cost leader in this evolving industry. Our proprietary optical platform goes back a number of years and was specifically designed to be manufactured at scale with high degree of automation. Our platform has been applied to data rate from 40G to 200G, and we expect it to be used in our 400G products as well. We also continue to innovate across our optical platform and expand our vertical integration and optimize the cost structure of our transceiver products. For example, between now and the end of this year, we plan to insource an additional 15% of the bill of materials for our 100G CWDM transceiver products.
We believe our platform, proprietary manufacturing process, and vertical integration are keys to our success in the market and remain focused on building on this strong foundation to position AOI for further success. With that, I will turn the call over to Stefan to review the details of our Q2 performance and outlook for Q3. Stefan.
Thank you, Thompson. Total revenue for the second quarter was $87.8 million, which was above our guidance range of $75 million to $81 million. As Thompson mentioned, the upside in the quarter was driven by increased demand for our market-leading data center products. Our data center revenue came in at $69 million, compared with $99.3 million reported in Q2 of last year. Our performance this quarter was driven by record revenue for our 100G products. In the quarter, 58% of our data center revenue was derived from our 100G transceiver products, and 39% was from our 40G products. We are encouraged by the customer traction we are generating with our 100G products. Looking ahead, we remain confident in our opportunities for growth.
As Thompson mentioned, we continue to expect 100G volumes to more than double in the second half of this year over the first half, which is based largely on the committed orders we announced in Q1 of this year. We also expect volumes to double next year over this year. We believe this strong growth is a reflection of accelerating data traffic trends that will require data center operators to expand their data centers and upgrade their infrastructure to keep up with the higher bandwidth demands. We believe the new innovative technologies and techniques that we have developed position us well to build on our momentum. The cost advantage, time to market, and flexibility afforded us through our vertical integration is a significant factor in our success and sets us apart from the competition.
As we continue to innovate, including our plan to insource an additional 15% of the bill of materials for our 100G CWDM transceiver products by the end of the year. We remain focused on providing our customers with the best product at the best cost and have a roadmap for continued cost reductions by additional manufacturing efficiency improvements and increasing the extent of our vertical integration. We are also adding additional testing steps that are required by certain customers. In the short term, these additional steps will constrain our manufacturing throughput somewhat, but we expect to once again have sufficient capacity to meet demand in Q4 of this year. As Thompson mentioned, our proprietary optical platform is also a key factor in optimizing the cost structure of our data center transceivers.
The platform was specifically designed to be manufactured in an automated way that not only provides us with the levers to reduce cost, but also provides a manufacturing process that can be leveraged across many product generations. The same platform has been leveraged across 40G, 100G, as well as 200G, which we just started shipping in volume last quarter. We expect to leverage this mature, high quality, and low-cost platform for many product generations to come. We also continue to maintain focus on diversifying our customer base, and in the quarter had seven design wins, including a 100G design win with a large data center operator in China. We believe our cost leadership and track record of innovation will allow us to be successful in many of these new customer engagements.
Turning to our cable television market, we generated revenue of $14.2 million, up 34% sequentially, and similar to the $14.4 million we generated in Q2 of last year. We are pleased with the improving trends we are seeing in CATV and continue to expect growth in this market, especially later this year when demand picks up for our Remote PHY product. We have three different customers moving into field trials with our Remote PHY product and an additional customer who is in the demonstration phase with the technology. The broad-based interest in Remote PHY by the MSO community is encouraging. Our product has already been demonstrated to interoperate with several CMTS vendors. We believe that we lead our competitors in terms of time to market for this technology.
Our telecom products delivered a record $4.2 million in revenue, representing 35% growth year-over-year, with demand coming from ongoing deployments of advanced mobile telecom networks around the world. For the quarter, 79% of our revenue was from data center products, 16% from CATV products, with the remaining 5% from FTTH, telecom, and other. In the second quarter, we had three 10% or greater customers in the data center business that contributed 52%, 16%, and 10% of total revenue, respectively. Moving beyond revenue, we generated a gross margin of 40.4%, which represents an increase of 40 basis points compared with the 40% reported last quarter. Our gross margin was in line with expectations. I'm pleased with our ability to continue to generate strong gross margins compared to the industry.
Total operating expenses in the quarter were $20.8 million, or 23.7% of revenue, compared with $20.1 million, or 30.8% of revenue, in the prior quarter. The sequential increase was mostly due to higher R&D expense. As a reminder, we expect R&D to remain at this level over the next few quarters while we invest in new production technologies that will enable further cost reduction on our transceiver products, as well as our next-gen data center and CATV products. Operating income in Q2 was $14.7 million, compared with operating income of $6 million in Q1 of 2018. Our operating margin increased in the quarter to 16.8%, compared with 9.2% reported last quarter. Non-GAAP net income after tax for the second quarter was $12.9 million, or $0.64 per diluted share, compared with $5.6 million or $0.28 per diluted share in Q1 of 2018.
GAAP net income for Q2 was $8 million, or $0.40 per diluted share, compared with GAAP net income of $2.1 million, or $0.11 per diluted share last quarter. The Q2 weighted average fully diluted share count was approximately 20.1 million shares. We recognized approximately $0.3 million in tax benefit from employee options that were exercised and restricted stock that vested during the quarter. Turning now to the balance sheet. We ended Q2 with $77.9 million in total cash equivalents, short-term investments, and restricted cash, compared with $83.3 million at the end of the previous quarter. As of June 30, we had $93.3 million in inventory, a slight increase from $92.6 million in Q1. We made a total of $21.2 million in capital investments in the quarter, including $9.9 million in production equipment and machinery and $9.5 million on construction and building improvements.
This brings our total capital investments year-to-date to $30.9 million. We now expect our capital expenditures in 2018 to approach $90 million. The lower revised CapEx forecast for the year reflects more efficient utilization of existing manufacturing equipment and does not reflect a reduction in expected production capacity. This figure also includes the beginning of construction of our new factory in Ningbo, still currently on schedule to be completed in early 2020. Lastly, I would like to make a few comments on the tariff situation with China. As you know, AOI operates three different manufacturing locations, only one of which is in China. All three locations are capable of manufacturing transceivers, with Taiwan and China both capable of manufacturing these products in high volume. Because of our vertical integration strategy, we also manufacture many of the components and subassemblies that are used in these modules.
The diversity of our manufacturing operations, both geographically and in terms of the types of products manufactured, gives us significant flexibility in adapting our production location in order to maximize cost efficiency. As political conditions change, we believe that we are well-positioned to adapt and will continue to plan for such contingencies. Moving now to our Q3 outlook. We expect Q3 revenue to be between $82 million and $92 million. Non-GAAP gross margin is expected to be in the range of 40%-41.5%. Non-GAAP net income is expected to be in the range of $11.1 million-$15.2 million, and non-GAAP EPS between $0.54 per share and $0.75 per share using a weighted average fully diluted share count of approximately 20.4 million shares. We expect our Q3 effective tax rate on our non-GAAP net income to be between 6% and 12%.
With that, I will turn it back over to the operator for the Q&A session. Operator?
Yes. Thank you. We will now begin the question-and-answer session. To ask a question, you may press star, then one on your touch-tone phone. If you are using speakerphone, please pick up your handset before pressing the keys. If at any time your question has been addressed and you would like to withdraw it, please press star then two. At this time, we will pause momentarily to assemble the roster. The first question comes from Simon Leopold with Raymond James.
Great. Thank you for taking the question here. I wanted to just follow up on your commentary around tariffs, in that I understand you've got some flexibility in terms of where you manufacture, but it does appear you are investing in adding capacity in the China facility in a scenario that we now see in place, I think the $34 billion or so of existing tariffs. Do you have exposure to those, or do you simply have exposure to proposed tariffs above and beyond that? I'm looking for a little bit more detail and an understanding of if tariffs come in place, could you manufacture what you need out of Taiwan and the U.S. without having a penalty?
Hey, Simon. The initial $34 billion tariff list, none of our products are on there, at least none of our major products. We're still doing analysis on all the smaller ones, but there's basically no impact from the initial tariff list. Some of our products are on the subsequent larger $200 billion tariff list. The point of what we were trying to say in the prepared remarks is that we do have capability to manufacture those products, in both locations. We think there would be minimal impact from those tariffs on AOI's business.
Okay.
If any.
Minimal impact. I think that's the additional color I was looking for. I also was hoping you could update us on the competitive landscape and maybe some color in terms of at your top customers, are you the primary supplier, a secondary supplier, does a secondary supplier always exist, and how has that changed over the last quarter or two? Thank you.
Yeah. As has been for the last several quarters, we believe we continue to maintain a leadership position with our major customers. I don't think there's been any major changes in our competitive positioning relative to the competition. I do think that the tariff situation, since you brought that up, I think that customers are aware of those tariffs. It's certainly something that they are concerned about because anything that adds cost to their total cost of ownership for their equipment is something they're concerned about. I think AOI's flexibility is also something that they find attractive.
You've talked about a customer who in the past was your largest customer, probably isn't today. That customer, I believe, was absorbing inventory and had also some architectural shifts. I think the adoption of SmartNICs was one of the factors affecting some of the demand for transceivers. Can you give us a snapshot of where you see your previous top customer's position today in terms of inventory and purchasing patterns, the outlook? Thank you.
Yeah. As we said, last quarter, I think we expected the inventory to normalize with that particular customer. That's what we've seen. The inventory is down to what I would consider to be a normal level consistent with the amount of business that we see and expect to see from that customer.
Great. One last one, and this may be difficult to answer, but maybe some blue sky thoughts. When you look at these hyperscale operators and their CapEx investment, when they build a new data center, acquire the land, do the construction, and have a shell of a building, how long, what's the period of time that you've observed it takes them to then fill the building with electronics? How long is the tail for AOI once the structure is built? Thank you.
Well, Simon, I think that's not something that I can easily comment on. I think it probably depends customer to customer. To be clear, we don't always have a lot of insight into which specific data center and what the building stage it is and that sort of thing. It's not really something I can speculate on. I know that the customers, as anybody that's trying to make a large capital investment, the customers are very eager to get the data centers turned on as quickly as possible. Oftentimes there's a lot of pressure on us to deliver quickly in situations where they're building a lot of new data centers. I know that there's a need to get that CapEx deployed and operational, as quickly as possible. As to how long that takes, I couldn't speculate an exact length of time there.
Yeah. Understood. Thank you for taking my questions.
Thanks, Simon.
Thank you. The next question comes from Paul Silverstein with Cowen.
Thanks, guys. First off, I assume the significant EPS spread reflects the concentration of your business and the lack of some better way to put it, lack of significant visibility, or is there something else in that spread? I've got a couple of follow-ups.
No. I think you're referring to the spread between the lower end and the higher end of the guidance range being a little bit wider this quarter. I think it just reflects our desire to be prudent and make sure that, given the uncertainties that we see, that we're encompassing all the range of possibility.
Nothing extraordinary, Stefan?
No.
Maybe other than you being more conservative than usual.
I don't know if necessarily more I guess you could say it's more conservative than usual. I think we're intending to try to encompass the range of possibilities that we see out there, and I guess you could say that's more conservative. Sure.
All right. Then two related questions. First off, pricing environment, both 100G, 40G. Any update you can provide us with to the extent there is an update? On the other side of the equation, I heard your comments about cost reduction in general. I'm hoping you can give us some quantification of the recent or reminder of the cost reduction to date and any quantification of cost reduction going forward in the future.
The pricing environment maintains, it's consistent with what we expected. I wouldn't say there's anything surprising in the pricing environment that we've seen or what we're seeing in terms of our expectations for the pricing going forward. In terms of the cost reduction, I think what we said is that historically, we've obviously been able to reduce the cost at a rate that's very close to the price reduction, and we expect to continue to be able to do so.
One final one from me. With respect to the seven design wins, especially the big one in China, I know it's hard predicting timing, but any sense you can give us for when you expect that to initial rollout and meaningful revenue? I know that's not necessarily 100%.
Just as a reminder, for us, a design win, when we announce a design win, it means we've already got orders. In fact, we're already shipping to this customer right now. As with any large new customer engagement, it doesn't happen overnight. It takes some time. We have meaningful revenue that we expect to get from this customer in this quarter and I would anticipate that over time they'll continue to grow.
I appreciate it. I'll pass it on. Thank you.
Thank you.
Thank you. The next question comes from Alex Henderson with Needham & Company.
Great. First question I'd like to ask is just around the 400G timeline. It seems like that it's gotten a lot of attention on the street, but it still seems like it's kind of back half of next year before that starts to become a meaningful factor. Is that consistent with your expectations, and how would you expect the 100G trajectory to manage through that initial ramp of 400G? I would assume that 400G would be too small to really undermine the 100G growth anytime, even in the 2019 timeframe. Is that right way to think about it?
Yeah, Alex. I think the 400G, it depends a little bit on which customer you talk to as to their urgency behind 400G. There are some customers who are pushing very hard for 400G. We've had customers in-house very recently evaluating our 400G solution, and I think that they're very interested in seeing that solution into production as quickly as possible. We have other customers that really, frankly, are still getting their feet wet in 100G and aren't likely to be pushing 400G anytime soon. I think there's a diversity among the different customers. I would say back half of next year is probably a reasonable timeframe to see any kind of meaningful revenue from that.
Although, again, some customers would like it earlier. I think realistically, as we saw with 100G, customer desire to have it doesn't necessarily translate into the entire ecosystem being ready to actually supply it in meaningful quantity. I think we're looking at back half of next year for some customers and longer than that for other customers. Which gets to your second question about the tail or the timeframe for shifting over. I think that 100G has a long road in front of it. I think there's a long ways to go before that gets meaningfully supplanted by 400G. Again, certain customers may move a little faster, certain customers may move a little slower into that direction, but I think there's quite a lot of growth continuing in 100G for the foreseeable future.
The second question I have for you is really around the pricing structure and the rate of decline last year in the back half versus what seems like somewhat of a moderation of that rate of decline in the first half. It sounds like you're suggesting that that moderation is going to persist into the back half, still declining, but at a more normalized rate. Have we seen some of the large players that have been trying to get into the market moderate the aggressiveness of their price, or has it been a challenge for some of those players to deliver the product? How do you see the competitive landscape relative to the ability to execute versus aggressiveness in pricing, if you could parse between those two?
Well, I guess probably the best way for me to put it is there seems to be a number of competitors who perhaps have been very aggressive on price to try to get business but really have had difficulties scaling up to the level of demand that's required. When you talk about price pressure, I think you have to take both the quoted price, if you will, from the competitor and their ability to actually deliver. They have to have both of those conditions in order for it to be a meaningful threat for us. I think AOI's scale, our demonstrated track record over multiple product generations in the data center, and the knowledge that we have in how to switch from one generation of product, say 40G to 100G, and manage through those transitions, I think that's something that our customers find value in.
Not all of our competitors, I think, have that level of experience. I think some of them are perhaps struggling to meet the demand that they've encountered by their pricing.
One last question, then I'll cede the floor. Any comments around PSM? Is that now pretty much a de minimis portion of your business at this point, or are you still involved with that?
We're still shipping PSM. As we've said for a number of quarters, CWDM, we feel like is a more attractive technology in most applications for most of our customers, and we have expected for some time that CWDM products would predominate over PSM and would probably take even greater share. That's what we're seeing, and I don't expect that to change.
Great. Thank you. Very helpful.
Thank you. The next question comes from Mark Kelleher with D.A. Davidson.
Great. Thanks for taking the questions. When you signed the supply contract with Facebook, the thinking, I think there with Facebook, was that there may be a situation where supply was in question. With the new deal you've got ramping in China, how close do you think you're going to be running to full capacity?
Well, as we noted in our prepared remarks, we're basically running at full capacity this quarter. Our intention is to run very close to capacity. We are investing in new production facilities. You noted in our prepared remarks as well that we continue to have a very aggressive capital expenditure plan for this year. Not quite as high as we had forecasted earlier, but that's because we think we can get more efficiency out of the existing expenditures that we're planning on making. But it's still a 40% increase or so over last year. In terms of the level of CapEx, last year was our previous record year. We're clearly continuing to invest in additional manufacturing, and our goal is to keep pace with the demand that we see as conditions improve.
The China facility isn't due online till the end of 2019. Is that correct?
Early 2020.
Early 2020. You're looking to double volumes in 2019 without the China facility.
That's correct.
Okay.
Basically, what we are doing is adding more equipment. More important is automation, so we can increase the throughput a lot, both in Taiwan and China. The investment is not only for China, we invest a lot of CapEx in Taiwan too, especially automation.
Okay. That leads me to my next question. Can you just run us through the vertical integration plan that you've got? I know you've talked about certain components you're going to be manufacturing internally. Can you just review the timing of bringing those inside?
Well, I don't want to get too specific on that. I'm sure many of our competitors would love to know our specific plans for how we're going to continue to reduce costs. I don't really want to provide that information publicly. What I can say is, look, there's a large amount of the bill of materials that we still do not manufacture in-house. We can bring more of that in-house. I think, as Thompson mentioned, in addition to bringing those components in-house, and we did talk about on the call the fact that we plan to bring about 15% more of the bill of materials in-house by the end of the year. In addition to just insourcing more parts, very important is continuing to improve the automation, improving the yields, and making the manufacturing process more efficient.
I think that's going to be an increasingly important part of our cost reduction strategy moving forward. It's one that I think we've spent a lot of time and a lot of effort fine-tuning over the last year or two.
All right. One last numbers question, just to check something. You said 58% of the data center revenue was 100G, 39%, 40G. I think that leaves 3%. Is that 3% 200G?
Some of it's 200G, and some of it is some 10G still remaining.
Okay. Thanks. That's it.
You're welcome.
Thank you. The next question comes from James Kisner with Loop Capital Markets.
I guess just quick housekeeping. Do you guys disclose a cash from operations figure?
It's in our 8-K. You can look at it. I can give you the number if you want.
I'll follow up.
Yeah. It was $22 million. Sorry, minus $4 million for the quarter.
Okay.
That was Q1, I'm sorry. Give you the wrong number. $22 million for the quarter, net cash used, provided by operations, yeah.
Okay. Just want to clarify something. You said multiple times now that 100G is expected to double half over half, and you've outperformed in this Q2, and it sounds like 100G was a big driver of that. I'm just wondering, given that now you perhaps have outperformed 100G volumes in the front half, does that mean that your sort of forecast for the second half has sort of gone up from what you thought it might be volume-wise for 100G, versus where it was a quarter ago?
No, I don't think our outlook has changed much from a quarter ago. I think the fact that we have a large portion of our business that's coming from, I would say, more secure contracts than what we've had in the past makes our visibility a little bit better, typically. I think nothing has changed really in our outlook from before.
Okay. Can you talk about the impact of currency on your margins, both gross and operating here, or I guess even net income margin, in this last quarter and the impact on your guidance?
Yeah. The currency consideration doesn't change margin for us too much. We have most of our raw materials and things that we purchase are in US dollars. We have, I wouldn't say exactly equal flow, but we have flow both directions between dollars and RMB and dollars and TWD. It doesn't have a great impact on our gross margin.
Okay. Last one. Okay.
Our operating expense or the labor cost in China.
Yes. Yeah.
Okay. I think I heard that.
Thompson was pointing out that the depreciating RMB lowers our labor costs, obviously, in China.
Okay. Also, this has been a topic in the past about your customers doing kind of end run going to the manufacturing partners and making their own "white box transceivers." In this quarter, obviously, I'm sure you saw that a competitor, another player in the space talked about that becoming a more serious effort again after kind of a head fake 18 months ago. I'm just wondering, do you have any updated thoughts on that? Is that something that you aren't really worried about at all? Just any updated thoughts on the kind of white box transceiver idea would be helpful. Thanks.
Well, I think making transceivers is a very technically difficult proposition. It's not like electronic manufacturing where it's kind of got a standardized process that virtually any factory can perform. In optical manufacturing, there are specialized processes, and in particular for us and for the designs of these very high-performance, high-speed modules, there's quite a lot of specialized equipment and know-how that goes into doing it. It's not as easy as it may seem to manufacture these transceivers. I think some of the competitors that have attempted to do this in the past have had, let's just say, mixed results. I'm very confident in the work that we've put into cost reduction, automating our production, really fine-tuning those things, and there's still more work ongoing in that respect. I think we're in a good place relative to the competitive business strategies that we see out there.
I can see something missing in this model. It's not clear who is responsible for the yield loss, for the automation, for manufacture process improvement. Who's responsible for the quality issue? That's not clear. I think there's still many issues in this model. I don't think it really work. All right? I don't believe the contract manufacturer, like people in Thailand, I don't believe they'll be responsible for this kind of job. Who will be responsible? Okay. Will it be the end customer responsible for this? I don't think really, do they have the right people to take care of all this issue? Because this will require a lot of manpower. It's not easy. AOI is very good because we spend many years. We know how to do automation, we know how to improve the process. That's very important.
The yield could be very different between company, from 72%-95%. There's a huge difference. Who is responsible? There's many questions need to be answered.
Thank you for that perspective. Thank you.
Thank you. The next question comes from Richard Shannon with Craig-Hallum.
Thompson and Stefan, thanks for taking my questions. Let's hear a couple from me. First of all, Stefan, I think in your prepared remarks, you talked about a headwind in third quarter due to testing going on. Can you help us understand what's going on there, and when does that headwind abate?
What we were saying in the prepared remarks was that we are implementing some additional testing procedures as we continue to diversify our customer base. Different customers have different requirements in terms of what they expect to see relative to design testing and so on and so forth. We're undertaking some additional testing steps in the manufacturing. What it does is it adds a few days to our manufacturing throughput. It takes a little bit longer to manufacture the transceivers. It's not going to meaningfully impact the cost of the modules, but it will lengthen the time that it takes to manufacture them. In this quarter, we expect that to constrain our production. In other words, there will be more demand than we can supply in this quarter.
As we said in our prepared remarks, by the fourth quarter, early in the fourth quarter, we expect that to normalize.
Okay. Is that something you're doing for a current customer or a new one? Is that with one specific SKU or across multiple ones?
It applies to multiple SKUs. It applies to multiple customers. It's data that we need to take to get a large enough sample size for some of the statistical analysis that we need to do for certain customers, including some existing and new customers. We're doing that across the board.
Got it. Okay. Makes sense. Another question, though, again, in data center. Your 40G revenues have stayed, I would say, largely flattish for a few quarters. You've talked about that product line eventually trailing off. Any updates on your thought process there? Is that continuing at current levels, or do you see any imminent decline there?
I think it's going to continue to tail off, but I don't see an imminent falling off of a cliff. I think we've said for some time that Actually, 40 Gig has been somewhat surprising in its resilience, but it's a testament, I think, to the fact that the economics of 40 Gig, in some instances for certain customers, are still very good. I wouldn't expect it to dramatically drop off anytime in the near future, but I would expect over time that it will continue to tail off gradually.
Okay. Helpful. Next question for me. You mentioned the expectation of doubling your volume in 100G in 2019 over 2018. When you say volume, is that sales or a unit comment?
That's units.
Unit. Okay. Follow-up on that specific topic is, do you expect this doubling, is that to happen over a broader base of your customers, or do you think it'd be more narrow profile?
Well, it's a combination of factors, right? We see some customers that I think are anticipating growth in their infrastructure, so that is purchasing more products from us next year. We also see new customers like the China data center customer that we talked about, that we expect to come on board that will contribute. It's not just one factor, it's a preponderance of multiple different things that we're looking at.
Okay. Last question for me. You mentioned 200G that you're shipping. Is this shipping into a design win for production volume or testing? What's your expectation of ramp of that over some period of time in the next two to four quarters or so?
It's for, I guess you could call it field trials or sort of advanced deployment testing. As far as the ramp up, at 200G, I think our viewpoint on 200G is that it's not going to be widely used. I think there's going to be a few customers who have interest in it. Many customers, it's not the direction that they're heading. 400G is clearly 400G shares wider acceptance, I would say, in the customer base. For us, 200G is important because it is a stepping stone for us to get to 400G, and I think that's probably the most significant part for us.
Got it. Okay. That makes sense. I think that's all the questions from me. Thanks for taking those.
You're welcome.
Thank you. The last question is a follow-up from Paul Silverstein with Cowen.
Hey, I appreciate you taking the follow-up. I just want to clarify, I hope you clarify, with respect to the comment about you being maxed out on volume, on supply. I know you've got new supply that's going to be coming online. Stefan, going back to your comments about China not being a meaningful risk and your ability to move manufacturing around, I'm just trying to reconcile, if you are maxed out at all of your facilities, what am I missing in terms of your room to move manufacturing around if need be?
Well, I don't want to get too specific on our plans, Paul, because look, there's a lot that's still up in the air here with respect to the tariff situation. As I mentioned, the tariffs aren't even passed yet, as you know. What I'm saying is, there's a number of different manufacturing steps that go into the manufacturing process for these transceivers, as you know. Currently, we do all of those steps in several locations. One possible thing that we could do, for example, is do a portion of that manufacturing in one location and then move it to a different location to do the rest of the manufacturing. We could also segregate it by customers. We do have significant customers in China and other parts of the world that are not subject to tariffs.
For example, many of our, even our U.S. customers, many of them have operations in Europe or other places that would not be subject to the tariff. We could potentially manufacture products for places, or customers that are not coming back to the U.S., in China, and do other customers or other geographical locations in Taiwan. There's a lot of things that we can do. I was merely pointing out.
The other point is that we still have empty space. If we use all the empty space in Taiwan and China, we can more than double our capacity for transceiver. With the 40G going down, we can convert the 40G transceiver manufacture capacity into 100G next year. I think Paul is thinking about is, do we use our space and the new China facility will not be ready until 2020, but that's not point. We can add more equipment in Taiwan, increase capacity in Taiwan.
Got it. Guys, one last clarification. Stefan, I heard your comment that you don't want to get into the details and the Bill of Materials in terms of insourcing, but I just want to clarify something. I think the last time you had publicly referenced the percentage of your BOM that you had brought in-house or were planning to, I thought the number was 60%-65%, maybe my memory is not serving me. The 15% you referenced today, is that new and different on top of what you previously announced, or is that part and parcel of what you had previously announced?
I don't have that statement in front of me either, Paul, but I think what we were talking about previously was that, we see ourselves being able to eventually in-source 60%-65%. This 15% is what we have plans to do between now and the end of the year.
15% additional on top of wherever you're at?
On top of where we are. Yeah. In other words, how much of the bill of materials, say we had a $100 bill of materials, right? At the end of Q2, say $50 of it was already insourced. Adding 15% additional to that would get us up to 65% of the bill of materials insourced, right? That's what I'm trying to say, by the end of the year.
All right. Are those the numbers, or you're just giving us a hypothetical?
Oh, no. It's a mathematical example. I'm just trying to provide you clarity into the math that I'm going through in the earnings call.
Understood. You clearly don't want to provide those exact numbers, is that correct?
Yes. I'm confident that there's a lot of people out there in our competitive universe that would love to know exactly what our cost is and what it'll be by the end of the year, but we're not going to give that information out.
Just wanted to give you the opportunity. All right, I appreciate it. Thanks, guys.
Thanks, Paul.
Thank you.
Thank you. This concludes our question and answer session. I would like to return the conference to Dr. Thompson Lin for any closing remarks.
Again, thank you for joining us today. As always, we thank our investors, customers, and employees for your continuous support.
Thank you. The conference is now concluded. Thank you for attending today's presentation. You may now disconnect your lines.