NeoPhotonics Corporation
Q3 2017 Earnings Call Transcript

Published: 7 Nov 2017

[Operator:]

Welcome to the NeoPhotonics2017 Third Quarter Conference Call. This call is being webcast live on the NeoPhotonics event calendar webpage at www.neophotonics.com. This call is the property of NeoPhotonics and any recording, reproduction or transmission of this call without the express written consent of NeoPhotonics is prohibited. The webcast will be available on the event calendar page of the NeoPhotonics website. I would now like to turn the call over to Erica Mannion at Sapphire Investor Relations.

[Erica Mannion:]

Good afternoon. Thank you for joining us to discuss NeoPhotonics operating results for the third quarter of 2017andour outlook for the fourth quarter of 2017. With me today are Tim Jenks, Chairman and CEO, and Beth Eby, Chief Financial Officer. Tim will begin with a review of the third quarter. Beth will then provide financial results for the third quarter and the outlook for the fourth quarter of fiscal 2017 before turning it back to Tim to provide color on the market and business drivers before opening the call for questions. The company's press release and management's statements during this call include discussions of certain Non-GAAP financial measures and information, including all income statement and balance sheet amounts and percentages other than revenue, unless otherwise noted. These Non-GAAP financial measures are not prepared in accordance with GAAP and should not be considered as a substitute for or superior to measures of financial performance prepared in accordance with GAAP. These financial measures and a reconciliation of GAAP to Non-GAAP results are provided in the company's press release and related Form 8-K, being filed today with the SEC, and can be found at the Investor Relations section of the NeoPhotonics website. Material contained in the webcast is the sole property and copyright of NeoPhotonics, with all rights reserved. Certain statements in this conference call, which are not historical facts, may be considered forward-looking statements that involve risks and uncertainties and include statements regarding future business results, levels of sales and profitability, subsequent events, product and technology development, capital needs and availability, customer demand, inventory levels and economic and industry projections. Various factors could cause actual results to differ materially. Some of these risk factors have been set forth in our press release dated November 6, 2017 and are described at length in our annual and quarterly SEC filings. Now, I will turn the call over to CEO, Tim Jenks.

[Timothy Jenks:]

Thank you for joining us today. Before we discuss our third quarter results, I would like to take a moment to introduce Beth Eby, who was appointed in August as our Senior Vice President and Chief Financial Officer, succeeding interim CFO Sandra Waechter. Beth joins us with more than 25 years of finance experience at Intel, where she most recently served as Vice President of Finance for their Internet of Things Group. Beth’s experience in the semiconductor industry and deep expertise in running finance organizations with complex, best-in-class global fab operations in the US and Asia is of great value to us and has already begun to have an impact. We are delighted to have Beth on board. Overall in the market we are seeing continued traffic growth in China and customer inventory levels returning to normal ranges, and we are seeing success in our 400G product introductions as well as continued growth in our laser products. At this point we are optimistic about our position and our ability to return to top line growth in 2018 and profitability by mid-year. Moving on to our third quarter results, in line with the revised outlook we provided on October 5th, we generated revenue of $71.1million.Our business progressed steadily in the quarter, with varying levels of contribution from our geographic regions. China was again up modestly with 5% sequential growth and the Americas grew 14%, offset by declines in EMEA and Japan. In line with the overall dynamics within China, our largest customer, Huawei, was up 2% sequentially in the third quarter, while Ciena was down sequentially after strong growth in the prior quarter. Beth will provide more details on revenues shortly. Demand in China has remained muted with lingering impact of excess customer inventory. Customers are suggesting that inventories are less than one quarter for most products at this point. While it is our expectation that recent tender awards such as those by China Mobile and China Unicom should create a more normalized demand environment in 2018; our discussions with China carriers indicate that network traffic continues to grow at a robust pace, necessitating continued network capacity adds. While China is soft, our mid- to long-term opportunities within the high speed optical market continue to grow. We have initiated a series of actions to focus our strategic direction on high speed components and coherent solutions, recognizing the explosion of optics in the cloud, together with transitioning our product development to focus on some newer markets and opportunities that include next gen components for data center interconnects, the expanding market in cloud and converged edge and the impact of disaggregation, especially in data center and cloud. At the same time, a very bright spot for us is the growth of our laser business, which I will highlight. We completed the divestiture of our low speed transceiver assets in the first quarter and we initiated certain restructuring actions to improve near-term profitability. We focused our developments on key growth opportunities within and between datacenters, front and backhaul for 5G wireless, traditional Metro and long haul telecom, plus emerging cable TV and edge network applications. The two largest technology markets we address both serve the cloud and converged edge. For coherent transport from DCI through long-haul, we continue to be a vertically integrated designer and manufacturer of the highest speed optical components which we sell to the merchant market and incorporate into our own module level products. For applications inside the datacenter, we are focused on supplying the highest value laser and optical IC solutions to the merchant market. As disaggregation of optical equipment continues to gain momentum, high performance merchant components capability will be an increasingly important differentiator. For Coherent transport we are driving down the cost per bit by increasing the bandwidth on a single wavelength through higher baud rates and higher order modulation schemes. Our 64 Gbaud Coherent Receivers and our ultra-narrow line width tunable lasers, plus our new Indium Phosphide 64 Gbaud Micro-Modulator, which is optimized for these higher data speeds, have been chosen by four major systems companies thus far for applications that will put 400G on a single wavelength over telecom distances, or600G on a wavelength, or 1.2 Terabits per second in a single transponder using two wavelengths for data center distances to 80 km. Our ultra-narrow line width lasers are a critical element in achieving these speeds because they have the narrowest line width and minimal phase noise for the higher order modulation schemes – or data coding –that are required. When more bits are encoded on a given wavelength, there is risk of signal degradation unless very pure lasers are used. Our laser, receiver and modulator optimally work together to enable this highest performance and we are now supplying all three together to customers for whom the highest performance and data rates matter. We also serve certain market segments by supplying complete coherent modules. Our DCO product offering uses our high volume 32 Gbaud optical component suite in a CFP form factor together with our internally developed optical ICs. We are in the final stages of qualification with several customers and will be shipping for revenue in Q4. Our Multicast Switch platform, which enables colorless, directionless and contentionless networks (or “CDC”), is qualifying with a major new customer in Q4 and will begin production shipments in early 2018. These switches are optimized for use with coherent transceivers to drive performance, data rate and reach while lowering amplification requirements, and therefore cost. Inside the datacenter we serve both short reach and long reach segments by supplying lasers and optical ICs. For short reach Silicon Photonics based 100G transceivers, we supply the light sources that generate the signals that are then modulated using Silicon Photonics. Our light sources are reliable in low cost non-hermetic packages, are custom designed for a customer’s application and provide a single wavelength for parallel signals in different fibers, or PSM4 architectures. We are qualified at multiple customers and are beginning to ship commercial volumes. For longer reaches and higher speeds inside the datacenter we produce electro-absorption modulated lasers (EMLs) which combine the laser with an Indium Phosphide modulator on a single chip. We are shipping to 100 Gbps applications in volume and we recently introduced a low power 28 Gbaud EML with an integrated driver. For even higher speeds, we have 56 Gbaud EMLs with integrated drivers for use in single wavelength 100G PAM4 and 4x100 PAM4 400G data center applications. We supply optical ICs such as drivers and amplifiers based on our Gallium Arsenide or Silicon Germanium platforms. As I noted, I want to highlight NeoPhotonics laser solution suite, which is an important business for us. Let me place this product area in perspective. Despite a challenging market due to the China inventory overhang, our suite of laser solutions, including transport, inter-data center and intra-datacenter solutions, has grown steadily over the year. Our revenue from lasers is up 25% over the same period last year and now represents almost half of our total revenue. We have been investing in our leading laser capability to have compelling new devices in development, including silicon photonics-based tunable laser solutions and laser array solutions for silicon photonics-based PSM4 and CWDM architectures. We have added considerable manufacturing capacity for these products and we are well positioned to lead the market going forward. We continue to see strong demand in regions outside of China driven by North American players, with strength in Metro and in DCI deployments. Within China, it is still too early to predict the timing of major demand changes. In the short-term, while the overall activity levels of our China customers may overshadow Metro growth and 400G wins, we believe that the mid- and long-term market drivers for our business are compelling, as China continues to build out national backbone, provincial and Metro networks, and as they advance in data center deployments and prepare for 5G wireless. With the continued traffic growth in China and now that customer inventory levels are approaching normal ranges, together with our success in 400G product introductions, we anticipate returning to top line growth in 2018 and profitability by mid-year. I will now turn the call over to Beth to provide further detail on our financial performance.

[Elizabeth Eby:]

Thank you, Tim, and good afternoon. I’m excited to be part of NeoPhotonics. I believe in the long term growth prospects in this market and that NeoPhotonics has a unique position in high speed optics. My goal is to make the company as well managed financially as it is in technology development by focusing on cash, cash flow and profitability. Highlighted in our October 5th press release and as Tim previously mentioned, as a part of our continuing actions to improve profitability and cash flow, in the third quarter we implemented restructuring actions to focus on key growth opportunities and to lower our breakeven revenue. Our actions included a reduction in force, real estate consolidation, a write down of inventory for certain programs and a write-down of idle assets. With these actions we have reduced risk to the balance sheet and lowered our operating expenses going forward. We believe our actions reduce quarterly operating expenses to achieve an approximately two-million-dollar quarterly OpEx reduction and an approximately $600 thousand reduction in manufacturing spending (or under-absorption)when fully realized at the start of fiscal year 2018. The costs to implement these actions are expected to be approximately $5.8 million, with $2.3 million in real-estate consolidation and fixed asset-write off, $2.0 million of inventory write-downs related to end of life projects and $1.5 million in severance costs. The Company incurred approximately $5.1 million of these costs in the third quarter with the remainder to be incurred in the fourth quarter. As we discuss our operating results for the third quarter, these costs have been excluded from our Non-GAAP results. Our actions balance the need for a sustainable level of operating expenses with long term growth goals. There is more work to be done and in process to balance our infrastructure and operations against the joint goals of growth and profitability. Moving on to our results for the third quarter, revenue totaled $71.1 million, with strength in North America and sequential growth in China offset by softness in other regions, as Tim mentioned. Our High Speed products were 84% of total revenue. Huawei Technologies and their affiliate HiSilicon Technologies have consistently been our largest customer, which was the case again this quarter as they accounted for approximately 39% of the Company’s revenue. It is important to note that we have several large customers in addition to Huawei, which may or may not exceed 10% in any given quarter. Ciena and Fiberhome Technologies, were 14% and 11% of revenue, respectively. In future quarters we will talk about our top five customers as a group, and to avoid undue emphasis on these fluctuations and respecting the privacy of our customers, we will talk to the number of customers and percentages of our greater than 10% customers, but will not discuss the individual names. In the third quarter our next four largest customers after Huawei were an additional 41% of revenue. For comparison, the next four customers represented 41% and 35% of total revenue in the prior quarter and the first quarter of this year, respectively. During 2016 the next four customers represented 33% of total revenue. Our Non-GAAP gross margin in the third quarter was 19%, down 5% from the second quarter of 2017. This decline was driven by

[Timothy Jenks:]

Thank you, Beth. We are focused on operational execution including cost containment, managing working capital and lowering our revenue break-even level. And I have described our key growth initiatives including data center and cloud focused developments. We feel confident about the position we are in and with the numerous levers we have at our disposal for further improvement in our financial performance, even if China demand remains at muted levels. Looking beyond the near-term, we remain enthusiastic about our mid- and long-term prospects within the optical networking space. NeoPhotonics has been focused on delivering industry leading performance components and modules. We are now the first to deliver a suite of 64 Gbaud component solutions, including Indium Phosphide lasers, modulators and receivers for speeds of 400G, 600G and beyond. Growth drivers for our telecom, data center and cloud markets include

[Operator:]

[Operator Instructions]. And we will go to Paul Silverstein, Cowen and Company.

[Paul Silverstein:]

Tim, a couple of questions if I may. First off, I apologize, because I heard the commentary about China. Also ask you in terms of your confidence level well with respect to comments about there now being one quarter, I think you actually could less than one quarter of inventory, and [I assume you phrase] specifically to your key components et cetera. Can you, how much of that is truly informed insight and how much of is hoped in wishful thinking?

[Timothy Jenks:]

Thanks for the question. It is informed thinking because it's information directly from customers, but the fact is yes, I'm referring to our situation not others broadly. And I don't necessarily have knowledge of others, but this is referring to our situation.

[Paul Silverstein:]

And your [indiscernible] Huawei HiSilicon Fiberhome is well through carriers. Is that the basis for this statement?

[Timothy Jenks:]

Generally, it is with our direct customers, rather than the customers of our customers.

[Paul Silverstein:]

Okay. From a product perspective, and I heard your comments at the end about the key drivers key metric, China data, but if we think from a product perspective, I've got to ask, given the cash situation and the performance of the business is that crumping your ability to -- are you have to make decisions in terms of supporting DCO, supporting the traditional components. Any hard choices you have made or need to make?

[Timothy Jenks:]

When we entered the year, we were planning on a higher level of business. And as we talked about when we announced our first quarter, we talked about our fourth quarter results towards the end of the first quarter. We reported that our largest customer was not taking anywhere near the volume that were suggested. And so, 2017 has been at a significantly lower level that was originally planned. So, through the course of the year, and in the changes, that we've done during the year, those have reflected decisions that we've made. And so, when business is lower than planned, it does lead to requirements for decisions. I think we've taken those decisions in timely ways. And we've lowered our breakeven as Beth explained, and we've lowered our operating and manufacturing cost as well. So, yes and yes.

[Operator:]

We'll next go to Simon Leopold with Raymond James.

[Simon Leopold:]

Hey thank you for taking my question. Just first wanted to squeeze in a housekeeping item. I don't think you gave us the geographic breakdown that you've given in the past. Just wanted to see if we could fill that in?

[Timothy Jenks:]

Sure, the third quarter it was China was 57%, Americas were 23%, EMEA 18% and Japan the balance.

[Simon Leopold:]

Great thank you. And then I just wanted to verify, you were kind enough to give us some commentary beyond the next question in terms of suggesting a return or normal seasonal pattern. But looking over my past model particularly with the exit of the access business, seasonality seems to be all over the place in terms of particularly the March quarter where some years you had March 2016 the high speed was up 25% sequentially yet in March of 2017 it was down 26% sequentially. So, I'm frankly not sure we really know what normal is. Wonder if we could get a little bit more finer detail at least what your perception is of what you consider normal seasonality.

[Timothy Jenks:]

Thanks for that Simon. It perhaps is worth trying to clarify. Let's talk first about the normal activities that happen in the fourth quarter and first quarter. Throughout our industry customers, the network equipment manufacturers generally do hold negotiations with their vendors during the fourth quarter. And in most cases those pricing decisions impact the first quarter. So generally speaking pricing is lower. And so, as Beth commented, we would expect in general to see higher volumes offset by lower prices. Higher volumes driven by tenders and general business trends and pricing is offsetting. So, in a general, I think if we look over a longer period of time and just the last two first quarters. We would see an overall decline in the first quarter relative to the fourth quarter. But let’s talk about the two that you cited. You will recall that, you mentioned March 2016. You will recall that in the third quarter, the third calendar quarter of 2015, there was expectation for significant growth and business in China, which did not occur. And at the time, we guided softly for that third quarter and suggested that it was a delay and that delay then came back in large quantities for the fourth quarter of 2015 and the first quarter of 2016. And so that was driven by a special circumstance i.e. very strong China demand with major tenders for March 2016. If you fast forward one year later, this was a point in time where, we saw the first reports of the inventory overhang in China. We and many of our competitors saw very strong business through the course of 2016. And as we, back to the end of the first quarter of 2017, it appeared that there was a meaningful inventory overhang that has been hanging over us for the remainder of the year. So, the normal situation would be over time, a steadily trending up volume offset by step-wise decreases in pricing happening in the first quarter, that would be the normalized situation. And then you have to overly on that in the two quarters you said very strong tender work at the end of 2015 and the absence of that at the end of 2016. Does that address your question?

[Simon Leopold:]

Yes. No, I really appreciate, you going into that level of detail, I think that’s helpful. And when I look over the longer-term trend, I think what you’re suggesting is, if we think about normal price declines of 10% to 15% and we think about a volume ramp. It should net us to probably a kind of mid to high-single-digit sequentially decline is my sort of back of envelope math. Am I crazy?

[Elizabeth Eby:]

Well, I expect the volume to go up offset by some level of pricing decline that we’re still in negotiations on what that level of pricing decline will be.

[Timothy Jenks:]

So, the way you are thinking about it though is perfectly in line.

[Simon Leopold:]

I wanted to just ask one more question that’s maybe a little bit nuance, but really trying to get my hands around what pressures the gross margin and I think what I had anticipated seeing in this quarter would be a drop-in inventory that I imagined you had factory underutilization that would result in drawing down inventory. We didn't really see that. And your commentary about lasers puzzles me more, because I would think with very strong laser business that I would see stronger gross margins. So, I'm puzzled on two fronts why we didn't see inventory coming down with depressed gross margin and why the laser contribution isn't driving gross margin better. And that's all I had. Thank you.

[Timothy Jenks:]

Yeah, okay great. Simon let me make comment on the laser situation and then Beth can comment more quantitatively subsequently. But our laser business as I noted has a couple of different pieces. There is a tunable laser business, there is an [even now] laser business and then there is a DFP and [railways] business that adds up to the total. And some of our laser fabrication is not in our own factory. So, we do use some CM support for part of our laser business. So, it has grown as we have added capacity and we have gained share we believe. And that has been beneficial, but as the loading in our internal factories has been light as we've been managing the inventory. Beth do you want to comment further?

[Elizabeth Eby:]

Yes. So, I think Sandra laid into some of this in the comments that she made in the last quarter. But we set factory loadings in September with the anticipation that we would have slightly stronger revenue in Q3 and stronger yet in Q4. I think she mentioned that the under-absorption charges would be fairly low in the third quarter. So, as we went through the quarter and continue to watch monitor inventory levels and we continue to drop factory loading to limit inventory and preserve cash.

[Operator:]

[Operator Instructions]. And we'll next go to Alex Henderson Needham & Company.

[Alex Henderson:]

Thank you very much. Tim, there has been a lot of discussion about China from pretty much everybody in the industry at this point. So, I figured most people have at least some sense of what's going on over there. But you've got a little bit of different mix of business, you've got a little bit of different viewpoint from the way you play across the OEMs. I mean Oclaro they said that they said that they thought it wouldn't be a recovery to the back half of next year which seems to be pretty draconian. Could you give us what your sense of the timing might be and realize that always a shot in the dark to try to pin down a timing. But if you have to play it out what do you think the most likely scenario is?

[Timothy Jenks:]

Well let's see, there are a number of things to consider here starting with the tender levels and then product mix. And who the network equipment company in question is, you highlighted in your comment Oclaro and certainly we have heard of Oclaro and [indiscernible] both who had similar patterns this quarter talk about their business. and I think in both cases ZTE is a more prominent customer perhaps than it has been historically for us. For us historically, Huawei has been more prominent. So essentially, as we look at 2018, we’re looking for information from both the network equipment companies as well as the carriers about tender timing and quantity. And as of this point, they’re certainly is no reason to believe in any delays beyond the second half of ’18. But by the same token, we can’t say that there is material increases in orders for the first half. So, I think the comments of companies regarding the second half versus the first half just reflects that reality. As we go forward, we know that we’ve seen China Mobile and China Unicom publish recent tenders. There has been some variation in which leading systems companies have won those respective tenders and it doesn’t necessarily portend who’s going to win the future based on what the current or past is. So, at the moment, what we heard from carriers and systems companies alike is a meaningful increase in overall volume but offset by pricing as we talked about in the prior set of questions. I think those are the mix of factors that affect our future outlook.

[Alex Henderson:]

So, are you saying that you don’t have any visibility for recovery, or don’t think that this going to happen in the first half?

[Timothy Jenks:]

I would say that I don’t have visibility for a rebound. I would expect things to trend up rather than jump up. And until we have tenders to suggest otherwise, our visibility is non-specific. And so, we can’t really say. We don’t really want to try and forecast the future that we don’t know.

[Alex Henderson:]

So, if I would look at the China Mobile tender that was led mid-September, the contract calls for 244,000 ports going to Huawei, 59,000 ZTE and 51,000 ports to FiberHome. I would think that that 69% loading, which is much larger than most people had expected will be favorable to your customer base. Any sense of when that might actually start to be implemented again?

[Timothy Jenks:]

Well, I think, the thing I would point out about that is its non-specific on the data rate. So, what you’re highlighting here is you’re highlighting approximately 350,000 ports, but there is a mixture of 10G and 100G and a lot of it is in fact 10G. And so, given that a lot of it is 10G, it doesn’t really have benefit or detriment to us, though our customers would see benefit in proportion to those numbers. But without being specific on 10G and 100G, it’s not an indicator to us.

[Alex Henderson:]

Clearly, it was an OTN build. So, the decision that underneath that, even if it was a small percentage 10% or 15%, it would still be a fairly significant coherent order. Again, the question is, do you have any sense of when that might actually be implemented?

[Timothy Jenks:]

Well, we did in the quarter ship 39% of our $71 million or almost $30 million of business to Huawei. And so, I think we're seeing meaningful business each and every quarter, but it's difficult to connect individual deployments more granually than that. it's also worth saying I think that what we are seeing I believe in that particular tender that you noted for China Mobile it was centrally announced. But I believe it was prudentially deployed. And so, the central announcement gives us an idea of the volume, but the actual procurement is deployment and the procurement is prudential which is more decentralized and need not necessarily follow precisely the breakdown that we've just suggested. So, it's pretty difficult for me to connect the dots on how that directly impacts us. It varies by province and it varies by systems company.

[Alex Henderson:]

If I could ask one more question, as we look out into the Datacom and some of the other segments that you haven't really competed aggressively in or not been focused on since you've been predominantly heavily on coherent in the past. How do you see that progressing, will that increase significantly as a percentage of sales in 2018, is it reasonable to think about percentage moving up? And can you gauge the size where we are in '17 and how large it might be in '18 relative to your base?

[Timothy Jenks:]

In the last conference call we talked about that being 10% to 15% of our revenue. And in fact, in my prepared remarks today, I did highlight that we provide both components and module solutions for coherent which is from datacenter interconnect to transport and long haul. And then when it comes to client and datacenter modules, we are providing primarily components and optical ICs. And so, while the growth rate maybe higher than our average, it is in the 10% to 15% range and it is primarily at a component level. So, I would expect that the percentage of our total revenue remains in that range or close that range. I don't think it will be breaking out wildly from that level.

[Operator:]

We'll go next to Richard Shannon of Craig-Hallum.

[Richard Shannon:]

Well Tim, thanks for taking my questions and Beth welcome to NeoPhotonics, looks forward to working with you. I guess my first question maybe also over to Beth financial question here. So, I'm doing my math right. Catching your comments to do my math right, you talked about a breakeven model kind a down into the mid-80s. And if we assume that OpEx doesn't grow much from your forecasted levels in the third quarter it sounds like you expect that to happen with gross margins around 28%. How close am I on that math?

[Elizabeth Eby:]

I think directionally correct.

[Richard Shannon:]

Directionally correct. How close to optimal utilization would you be at that breakeven level. I assuming when you nearly maxed out but I guess I'm curious after some of the restructuring going on there, how much room you have beyond that breakeven level to keep moving without having too much excess capacity or more capacity, I guess?

[Elizabeth Eby:]

What we said is that our current capacity levels, we’ve got about $500 million -- we’ve got capacity to support about $500 million of revenue that is been, that’s what we put in place. So, we got a lot of room.

[Richard Shannon:]

Okay, perfect. I missed that one. So, thanks for reminding that one. Tim, maybe one back to you and Alex a few minutes ago had a question on one specific of your kind of next generation products growth here, specifically in data center. I guess, I wanted to kind of more broaden this out to look at all of your growth drivers. So, you’ve mentioned if you have like 400G coherent. Again, data centers some multicast switches and you also specifically announced or talked about DCO product you’re starting to ship in the fourth quarter. As we look into 2018. Is there one of those or a subset of those, you expect to be the dominant driver beyond your current 100G coherent business?

[Timothy Jenks:]

Well, let’s see the 100G coherent business is without a doubt, the largest piece of our business. And that overtime will add in the 400G domain. But we would expect material volumes and revenue contribution from 400G to be more 2019 event than a 2018 event. So that causes us to then focus on components and ICs, DCOs multicast switches et cetera. Of those, they’re actually is a pretty good balance across products. I did say that we are starting to ship for revenue on our DCO, we have several customers for that. So, we’re optimistic about that. I mentioned that in multicast switch, we added another key customer and we would expect that to be shipping in 2018. But both of multicast switches and the DCOs are small numbers at the beginning. So, I can’t say that they are the biggest driver. So, it really is a fairly tight balance across these products.

[Richard Shannon:]

One quick follow-up and I’ll jump out of line, Tim. Specifically on the DCO, would you characterize your wins here second sources are these into primarily new designs maybe with features and performance maybe above what kind of the standard DCO been shipping in the market thus far?

[Timothy Jenks:]

Well, let’s see DCOs overall are an interesting product segment where we anticipate DCO to be a more important and more product strategic solution for system companies overtime. They enable some companies to compete in new segments, it allows connecting, it can be used as a replacement if you will for engineered line cards. It can also be used for connecting routers and switches. And so, it has a broad range of applicability and we would expect overtime to see meaningful growth in the DCO space. DCOs are currently provided in the market, our largest customer makes some of their own and certainly, Acacia is an incumbent supplier that has done well in creating the segment. So, an answer to your question, for the most part of the second source activity as opposed to new opening, but I will say there are some new opening possibilities in the mix.

[Operator:]

[Operator Instructions]. And we'll go to Mark Kelleher, D. A. Davidson.

[Mark Kelleher:]

I was wondering if you could give us more detail around the reduction in force. What line items were that from, how many people would that involve?

[Timothy Jenks:]

So, let me lead off with the functions and then Beth can comment on the cost. Essentially the reduction in force that we've done have been across the company have affected each of the operating areas from sales and marketing to G&A, some in R&D and in manufacturing organization. So, it’s not restricted to one area. And that has resulted in our being more focused on some of the things we're doing with respect to the products and programs that we're supporting and with respect to how we go forward. We haven't detailed numbers of heads but we have detailed costs. And Beth do you want to comment on those?

[Elizabeth Eby:]

Sure, a lot of the reduction and a lot of the restructuring was non-risk related. But we did do a substantive risk. We aren't public on the numbers but the total cost of the severance, split between Q3 and Q4 will be about $1.5 million. From a geographic standpoint we did the U.S. in Q3 and China will be the Q4 charges, because we did not quite have that in place by the end of Q3. And we did a write down of our building that we went out of asset write-offs of a couple of the programs that we have stopped and some related inventory write-offs that were not officially restructuring from an accounting standpoint but more related to that.

[Mark Kelleher:]

Can you tell me the number of employees that you have right now total?

[Timothy Jenks:]

I think we have approximately at the end of the quarter we have between 1800 and 1900 employees at the end of the quarter.

[Mark Kelleher:]

And are you still in the process of the [rift] or that completed in just similar cost associated with that or into the next quarter?

[Timothy Jenks:]

Some of the cost rolled over into the fourth quarter. Implications on individuals I think is inappropriate to comment on.

[Mark Kelleher:]

Beth, could you just talk about the -- just reiterate what you're saying on the debt capacity, how much is available? And I think you said you're going to add $20 million of that next quarter?

[Elizabeth Eby:]

We have not made specific decisions, but I did say we would. We have $10 million that we expect of cash from operations negative in Q4 and another $10 million of CapEx expenditures. We do expect to partially fund that with debt. Our capacity that remains available is we have drawn $30 million out of the total of $50 million credit line with Wells Fargo Bank. We do have to keep as is public. We do have to keep $5 million of availability at all times on that one. So, we have about 15 million remaining there. We also have a $20 million from Shanghai Pudong Bank that we could draw against. So, we will draw some of that in Q4 and we expect our cash balance to go down a little bit in Q4.

[Operator:]

And it appears there are no further questions at this time. I would now like to turn it back over to Tim Jenks for any additional or closing remarks.

[Timothy Jenks:]

Thank you, Malinda. I appreciate everyone taking the time and interest in NeoPhotonics today. We look forward to talking to you on our next call. Good bye.

[Operator:]

That does conclude today’s conference call. We thank all for joining us.