Lumentum Holdings Inc (LITE) 2017 Q3 法說會逐字稿

Good day, ladies and gentlemen, and welcome to the Q3 Lumentum Holdings Inc. Earnings Conference Call. (Operator Instructions) As a reminder, this conference is being recorded.

I would now like to introduce your host for today's conference, Mr. Chris Coldren, Vice President, Strategy and Corporate Development. Sir, you may begin.

Thank you, Bruce. Good morning, and welcome to Lumentum's Third Quarter Fiscal 2017 Earnings Call. This is Chris Coldren, Vice President of Strategy and Corporate Development. Joining me on today's call are Alan Lowe, President and Chief Executive Officer; and Aaron Tachibana, Chief Financial Officer.

This call will include forward-looking statements, including statements regarding the markets in which we operate, including trends and expectations for products and technology, purchasing trends, Lumentum's expected financial performance, expenses and positions in our markets. These statements are subject to risks and uncertainties that could cause actual results to differ materially from our current expectations. We encourage you to review our most recent filings with the SEC, particularly the Risk Factors described in our 10-Q filing for our fiscal third quarter ended April 1, 2017, which will be filed today.

The forward-looking statements we provide during this call, including projections for future performance, are based on our reasonable beliefs and expectations as of today. Lumentum undertakes no obligation to update these statements except as required by applicable law. Please also note, unless otherwise stated, all results and projections are non-GAAP. Non-GAAP financials should not be considered as a substitute for or superior to financials prepared in accordance with GAAP. Our press release with our third quarter fiscal 2017 results is available on our website, www.lumentum.com, under the Investors section and includes additional details about our non-GAAP financial measures and a reconciliation between our GAAP and non-GAAP results. Our website also has our latest SEC filings, which we encourage you to review, and supplementary slides relating to today's earnings release. Finally, a recording of today's call will be available by 11:30 a.m. Pacific time this morning on our website.

Now I would like to turn the call over to Alan for his comments and third quarter business highlights.

Thank you, Chris. At Lumentum, we are focused on accelerating the speed and scale of cloud, networking, advanced manufacturing and next-generation 3D sensing applications with our photonics technologies. Our strategy to invest in growing markets, develop the best products and technology and foster the closest relationship with our market-leading customers is succeeding. Since our last call, we have made very good progress on our strategic goals in each of our markets. Our revenue from 100G QSFP28 transceivers nearly quadrupled in the third quarter. We grew our ROADM revenue and shipped record levels of next generation TrueFlex ROADMs.

Our commercial lasers business returned to growth, and we achieved major milestones on 3D sensing products for mobile device applications. To answer the inevitable question and not to keep you in suspense, yes, we have received volume 3D sensing production orders for delivery in the September quarter. However, like many in our industry, we have some near-term headwinds. Late in the third quarter, we saw demand from China slow significantly. Before last quarter's call, we anticipated some softness in China. We expected demand to pick up late in the quarter and increase into the June quarter. This did not happen. Demand from our Chinese customers continues to be at significantly lower levels than we saw in the first half of our fiscal year.

Third quarter revenues were $255.8 million, up 11% relative to a year ago. Optical Communications revenue grew 10% year-on-year. Commercial lasers returned to growth with revenues of $39.7 million, up 20% year-on-year, driven by increases in both fiber and micromachining lasers. During the third quarter, we started shipping our new higher-performance Gen 3 fiber laser and recognized our first revenue.

Telecom revenue increased 3% quarter-on-quarter despite seasonal factors and softer-than-anticipated demand from China late in the quarter. We saw strong demand outside of China, particularly in North America, driven by metro and datacenter interconnect applications. ROADM revenue increased 5% sequentially and our newer TrueFlex ROADMs achieved new record levels. TrueFlex ROADMs represented more than 80% of total ROADM revenue. We had a higher mix of blade-level ROADMs products in the third quarter compared to the second quarter due to the increased metro demand.

Nearly all major telecom product lines were up quarter-on-quarter. Pump laser revenue was down slightly due to increased ROADM blade shipments, but this is because we continued to be constrained by our pump laser capacity in the third quarter.

Submarine revenue, which is project based and tends to be lumpy, also dipped in the third quarter. Datacom revenue decreased 43% quarter-on-quarter, primarily due to slower demand from China for our CFP2 products. Demand is strong for our QSFP28 transceivers, which led to a nearly fourfold increase in QSFP28 revenues. QSFP28 products share the same manufacturing capacity as our CFP2 products. Growth was capped by capacity limitations and our ability to pivot our supply chain in favor of QSFP28 during the quarter.

Given the significant average selling price difference between CFP2 and QSFP28 transceivers, 100G transceiver revenue declined despite an increase in unit shipments. Also lower-speed 10G and 40G Datacom revenue continued to decline as expected. Industrial and consumer revenue, which consists primarily of revenue from 3D sensing and laser diodes for fiber lasers was up $3.8 million or 45%, sequentially. Our next-generation 3D sensing products are tracking very well to our plans. We have already been approved to begin volume manufacturing for our first consumer mobile applications, and we have received volume production orders for delivery within the September quarter.

These are very exciting milestones and our 3D sensing team has done a great job in getting us into this leadership position. We continue to have strong engagement from a broad base of 3D sensing customers, both in the consumer electronics and automotive markets. We expect 3D sensing to be a significant growth driver over the coming years.

Looking to the fourth quarter, we expect continued soft demand from China, which will cause a decline in telecom revenue. While we are hearing about some modest tenders related to ROADM networks in China being awarded recently, we believe that there is sufficient inventory at our customers to satisfy any short-term surge in demand during the fourth quarter. We expect our Datacom revenue to be up in the fourth quarter driven by strong demand for QSFP28 transceivers and capacity additions. Despite the near-term softness in demand from China, we remain confident about our longer-term outlook. While we don't provide guidance more than 1 quarter at a time, we expect the demand from China will increase in the first half of our fiscal 2018 and 3D sensing should add meaningful revenue in our next fiscal year. It is an exciting time at Lumentum as our photonics technology is increasingly a critical enabler of leading-edge communication, industrial and consumer applications. Our investment in new products and technology positions us well for future trends.

Demand continues to grow for bandwidth across the world's datacenters and communication networks, including those in China. New submarine cables are being built and lit up around the world. In addition to traditional communication service providers, cloud operators are increasingly driving new deployments of optical networks and submarine cables. Hyperscale datacenters have started to transition to 100G.

New ROADM-based networks -- network deployments are in their early stages globally. China is taking their first steps in their initial ROADM deployments this year. We are working with all of our customers on even newer platforms that incorporate higher performance, smaller and more cost-effective ROADM solutions. We believe that we have the best ROADM products available today and new products -- and a new product pipeline that will further our leadership position over time.

Manufacturers around the world are increasingly using advanced laser-based techniques. Leaders in next-generation consumer electronics, virtual and augmented reality as well as the automobile industry are looking to laser-based 3D sensing to enhance capabilities and enable new applications.

I will now hand it over to Aaron for more details on our financial results and our guidance for the fourth quarter fiscal 2017.

Thank you, Alan. Net revenue for the third quarter was $255.8 million, which increased 11% compared with the same period last year, and decreased 3.5% sequentially due to softer demand from China and seasonal effects from annual price reduction. GAAP gross margin was 32.1% and decreased 70 basis points quarter-on-quarter. GAAP operating margin was 5.3% and GAAP diluted net loss per share was $0.92. The GAAP diluted net loss per share includes the impact of a $0.94 loss per share from a noncash unrealized loss on the derivative liabilities for both Series A preferred stock and convertible notes, and also the noncash effective interest on the convertible notes.

Our third quarter non-GAAP gross margin was 34.4%, which increased 220 basis points compared with the same period last year. The year-over-year increase was driven by a 330 basis-point improvement in Optical Communications gross margin.

Non-GAAP operating margin for the third quarter was 12.6%. Our operating margin increased 380 basis points year-over-year from the gross margin improvement and operating leverage from higher volume. Non-GAAP net income was $30.8 million, an increase of 56% compared with the same period last year. Non-GAAP earnings per share was $0.49 based on a fully diluted share count of 63.4 million compared with earnings per share of $0.32 last year.

Third quarter earnings included $500,000 of other expense and $900,000 of tax expense.

Now for some additional details. Optical Communications revenue was $216.1 million, an increase of 10% over the same period last year, driven by a $24.9 million or 18% increase in telecom revenue, a $500,000 or 4% increase in industrial and consumer, partially offset by a $6.5 million or 14% decrease in Datacom revenue. Although Datacom revenue declined 14% year-over-year, our 100G transceiver revenue increased 29% and our revenue from QSFP28 transceivers grew over 560% over the same period last year.

Optical Communications gross margin at 33.1% increased 330 basis points compared with the same period last year, as a result of higher volume and mix of products. Commercial Lasers revenue was $39.7 million, an increase of 20% compared with the same period last year. Third quarter Commercial Lasers gross margin was 41.3%.

Operating expenses totaled $55.7 million or 21.8% of revenue compared with the same period last year of $53.9 million or 23.4% of revenue. R&D expense was $34.2 million and SG&A expense was $21.5 million.

In the third quarter, we raised $450 million with an issuance of convertible notes. The notes have a coupon of 25 basis points, a 32.5% premium to our closing price prior to the date of issuance and a 7-year maturity. The additional cash strengthens our balance sheet and we will use the proceeds for general corporate purposes, which may include capital expenditures, including -- manufacturing capacity expansion and working capital. Our cash balance was $577.9 million at the end of the third quarter, and represented an increase of $422 million sequentially.

Capital equipment additions were approximately $40 million in the third quarter. During the quarter, we purchased a facility in Thailand for $10 million. This facility will provide Lumentum with additional capacity to support long-term growth initiatives. The outsource model using contract manufacturing partners continues to serve us well. Going forward, we intend to utilize a combination of both external partners and internal capability.

Now on to our guidance for the fourth quarter of fiscal 2017, noting again that all projections are on a non-GAAP basis. We project net revenue for the fourth quarter to be in the range of $220 million to $235 million, with operating margin in the range of 9% to 11% and earnings per share to be in the range of $0.30 to $0.40.

Now I will turn the call back over to Chris to begin the Q&A session.

Thank you, Aaron. I would like to ask everyone to limit discussion to one question and one follow-up. Bruce, let's begin the question-and-answer session.

(Operator Instructions) And our first question comes from Patrick Newton from Stifel.

Alan and Aaron. I guess, just 2 questions. First on the China side. Given kind of lack of visibility now, do you have any thoughts on when China trends could turn positive? And if we think about the current China revenue that's baked into June quarter guidance, and assume no rebound, would you anticipate a return to sequential growth in your Optical business, excluding lasers and 3D sensing in the September quarter?

Patrick, I think the visibility in China is pretty tough right now. Like I said in the prepared remarks, we are seeing 2 small tenders for ROADM deployments in the provincial networks. But I would expect that the first half of our fiscal '18 to rebound from where we are today. The question in my mind, does that happen in the September quarter or is there a continued bleed-off of inventory during that quarter, and then a rebound in December quarter. And at this point in time, it's too hard to say.

And then the potential to grow if China remains flattish from current levels in the September quarter with the rest of your optical business?

Oh, certainly, we'd expect so. Yes.

Great. And then on 3D sensing, it's great to hear that you have volume production orders in hand. I'm curious if you could help us understand the potential revenue contribution in the September quarter, or any color on the magnitude of revenue potential in FY '18. And then I think that capacity is somewhat of a concern. I'm sure you don't want to talk about specifics. But if the ramp of this product is successful over the next 12 months and given you're talking about multiple OEMs and automotive, can you help us think through how we -- how capacity needs to come on relative to current levels? Does it need to increase by a factor of 2x or 3x? Any color there would be appreciated.

Sure. I mean, we're at very low levels this quarter. The orders that we received for the September quarter are multimillion units. We expect that to continue to grow. And we're bringing on capacity at our partners this quarter and into early next quarter that we believe should suffice to be able to meet the initial ramp. And then we'll reevaluate as we make progress with other of our customers. I think our partners and we have invested sufficiently for the existing ramp and are bringing capacity online as fast as we possibly can to meet the ramp.

And any commentary around revenue potential?

Yes. At this time, Patrick, I think it's tough to say. We'll certainly give guidance in our -- 3 months from now for our September quarter. And we can give you some color directionally, but as I said in the prepared remarks, we expect significant revenues in fiscal '18 from 3D sensing. But it's too early to quantify.

And our next question comes from Simon Leopold from Raymond James.

I just wanted to see if we could clarify what happens in the mix in the June quarter in the Optical Communications segments, Datacom, Telecom. Given that you've already had a very significant sequential step down in the Datacom, I'm just wondering how to think about the components in the June quarter relative to the guidance. Maybe if you can give us a little more color on CFP2 versus ROADM, what you're expecting?

I would say that Datacom is expected to be up as we've brought on more capacity to meet the needs and demand for QSFP28. So I expect the June quarter Datacom to be up. Telecom will be down mainly due to the China slowdown that we saw start really in March and continue through the beginning of this quarter. Lasers flattish, but pretty exciting times as we've brought on our new Gen 3 fiber laser and start to ramp that into a meaningful way this quarter. And then 3D sensing up slightly, but not significant revenue from the June quarter.

Great. And then just my follow-up on 3D sensing. Is there any thoughts you have on the competitive landscape? I think what you've said in the past is that, one key aspect of your advantage is -- working through the partnership fab is your ability to ramp volume. If you could provide us maybe some thoughts on your competitive position in 3D sensing?

Sure, Simon. As I've said all along, I think we have selected partners that know how to manufacture at high volume and the combination of our partnerships and our design capability to design the kind of products that our customers need, really put us in a leadership position. I think we're off and running, and I think we're significantly ahead of our competitors, so we should be able to capture a good share of the business as it comes on in the September and December quarters.

And our next question comes from Alex Henderson from Needham & Company.

So if we were to ignore China altogether and just take that out of the mix, can you give us some sense of what the rate of growth in Telecom might look like sequentially outside of China, so that we can at least have a better gauge of what the ambient demand conditions look like in rest of the world?

I haven't studied that in detail, Alex. It's a good question. But I think we're seeing normal growth demand from North America as expected. And in fact, if you look at different countries in Asia, like India, for instance, is driving very strong demand. Europe is looking okay to a little bit better-than-expected. So I think, outside of China, business is good for Telecom. And then you couple that with the Datacom hyperscale growth in demand as the 100G transition is happening, things looks pretty good outside of China.

And on the pump laser side, I assume that you're adding capacity pretty aggressively there. Can you give us some sense of what the capacity adds looks like to address the tightness that you're experiencing in supply? And the assumption that ROADM's ramp as Telecom ramps in China eventually kicks in?

Yes. So pumps, we've been adding capacity over the last 1.5 years. We have more capacity coming online this quarter. If you look at it from the December quarter to the June quarter, I'd say it's probably another 20% to 25% of capacity going online. And we think that, that puts us in a pretty good position to be -- have flexibility to meet our customers' demands. We are seeing additional demand from Raman pumps, which is unique and I think positions us pretty well. So I think, overall, the terrestrial pump business is good. And what happened in last quarter, maybe it wasn't clear, was we used a lot more of our pump capacity for internal blade production, and that constrained us for external revenue on pumps. I think we're to the point now that while we're still tight on capacity, as we bring on capacity through this quarter, we should be in pretty good shape.

If I could just sneak one last thing in. Any change in the status of the ROADM qualification that your competitors -- that you can see evidence of?

I think it's a better question to ask them. But we're focused on making sure we meet our customers' demands, and I think we're doing a very good job of that. I haven't seen anything that would indicate that we're losing any share at either of our major blade customers.

And our next question comes from Rod Hall from JPMorgan.

And I appreciate the clarity on 3D sensing. I've just got a few follow-up questions on that and also, obviously, on China. I'd like to just understand, Alan, you're talking about multimillion unit orders for the September quarter. When do you start delivery on that? Can you give us any idea? Is that representative -- what you're talking about in September and what you'll eventually guide for, is that going to be representative of kind of full-volume capacity? Or are we talking about 0.5 quarters' worth? Just help us understand when that volume ramp starts. And then, are we at full-volume output in the December quarter? So that's the first question. The second question on 3D sensing is, could you help us understand what you think your share position is now? I know it's hard to know, but given you now have orders, do you have a feeling for where your share position stands on that? And then on China, just coming back to that, I just want to understand how conservative you think you've been on China in your guidance for June. Particularly, and I don't know if you want to comment specifically on this, what you're assuming for Huawei. And you say you've seen these tenders starting to come through. Do you think there's a little bit of unlocking in terms of the Chinese movement? Any other color that you can give us from the carriers there would be helpful.

Sure. Well, starting with the 3D sensing, these multimillion-unit order -- orders are just the initial orders. We expect to be receiving orders on a weekly basis moving forward to cover our lead times. As I said in the prepared remarks, we have been approved to start volume production. And so we are ramping production to be able to achieve the volumes that our customers need in the short term. And you can imagine the cycle time of 3D sensing for us is a few months. And so given that we're in production today, you can determine when the production ramp starts, and given a 13-week lead -- cycle time. So we'll be ramping through the quarter, and in the December quarter, we'll be at full-volume production capability, assuming the program goes as we expect. As far as share is concerned, it's hard to tell. But I think given the attention that we're getting and the volume orders that we're receiving, I think we're in a very, very good position relative to our competitors and are getting more than the majority of the share of the business at the initial ramp. Does that answer your question on 3D sensing?

Yes, that's helpful.

Okay. And then on China, how conservative are we? I don't know that you can be too conservative, given that it's -- the visibility is pretty tough. So I -- I've always anticipated being surprised by China. And I think we're trying to take an approach that even if business were to come back, I don't think that that's going to drive incremental volume for us over our existing plan in the June quarter. It could drive business in the September quarter over and above the current levels, but I think it's too early to state that at this point in time.

And our next question comes from Troy Jensen from Piper Jaffray.

Congrats on the 3D sensing order. I guess, my first question, I'd like to know for the volume 3D sensing, is it multiple flavors, as in different power VCSELs, or is it just one?

Troy, I think we are under all sorts of confidentiality agreements with our customers. I'd prefer not to say what the number of different chips we're providing or the applications. Suffice it to say that we're very well positioned with this lead customer. And I think the fact that we've got volume production orders for production ramp in September places us pretty good. So I'd prefer to stay away from any of the details.

All right. Totally understand, I just had to ask it. And how about just a follow-up on the QSFP28. I'd just be -- wanted to know, I'd like to know, what percent of the business you think is going into Web 2.0 versus the network equipment companies? And then maybe the number of hyperscale customers that you have for QSFP28?

Troy, this is Aaron. So in terms of the third quarter, it was less than half. And here in the near term in the June quarter, it's still going to be a little bit less than half. But it's going to -- it's continuing -- continuously growing in terms of what's going to hyperscale.

And the number of hyperscale wins?

We haven't said the exact number, but there are several.

And we're engaged with all of them. So whether we're in production today or about to be qualified, I think it's just a matter of time.

And our next question comes from Michael Genovese from MKM Partners.

First question, I just want to ask about ROADMs. It seems like in the March quarter, you had a strong demand in both the U.S. and China, but that China maybe built some inventory, so they're -- that's going to fall off in June. But specifically in the U.S., are you expecting growth in ROADMs in June? Or will it be like before, where the customer kind of shifts to transmission lines for a while before coming back to ROADMs again?

From the China perspective, I think you've categorized it accurately in that we ship ROADMs by quarter, don't expect to ship very many ROADMs this quarter. As far as U.S. is concerned, we put the ROADM blades in the BMI hubs and they get pulled or they don't get pulled. I haven't looked exactly at where we expect that to go, but I think there's nothing that says it's going to be highly lumpy this quarter compared to last quarter. So I think constant growth would be my expectation.

Okay. And then my follow-up is back to 3D. I mean, I think the last time you guys did a call like this, there was some discussion, I can't remember if it was publicly or in the callbacks later. But there was some discussion about a potential $50 million revenue level for 3D sensing in the December quarter in this project. So I guess, my question would be has there been any change to that particular outlook?

Michael, I don't know that we gave any specific numbers like that. I think what I did say was that we believe the 3D sensing market for optical components will achieve a $1 billion annual-type market over the next coming years, and that over the next year or so, one can imagine a market that drives $100 million a quarter kind of business. Our share of that's yet to be determined. So I think we'll continue to update you on quarterly calls, but I don't think we've ever said what we expect the December quarter revenues to be.

And our next question comes from Tim Savageaux from Northland Capital Markets.

I have a question on your comments about internal manufacturing capacity additions, and I'm -- wonder, especially in light of kind of current demand conditions, where that capacity is sort of focused. If there is certain pain points on the comm side or across industrial or even 3D sensing, where that capacity might be focused, and what the timing of that might be, on the one hand. Then in a potentially related follow-up, I don't know if you said -- obviously, you saw some strong uptick on the laser side, the industrial laser side, whether you expect that to continue into June relative to your guidance for the overall revenue line?

Sure. Tim, in terms of the manufacturing, so we did buy a building in Thailand for $10 million. We're currently in the process of fitting the building up and in terms of getting it ready, with infrastructure, meaning IT systems, networking systems, capability, it take several months to do that. So some of that has been in the works in terms of the preparation. In terms of which specific products. So 3D sensing, I think Alan -- we've noted we're using an outsourced foundry model. This facility in Thailand is going to be primarily used for our base business. In terms of the specific products, we're not in a position to really state which products will be the first to go in. Those are things in the plans, we're working through them at this point in time. In terms of the timing of readiness, we expect to be shipping some initial units out of the factory from a production standpoint early in our fiscal 2018 period, which is the September through the December quarter. In terms of your second question on commercial lasers. So we did achieve $39.7 million in the March quarter. Looking into the June quarter, we believe that the laser volume could be flattish and there could be some variability depending upon customer inventory levels. So it could vary a couple of percent either way in terms of the top line.

Operator, next question, please.

Our next question comes from Richard Shannon from Craig-Hallum.

Maybe I'll ask a first question on Datacom. Wondering if you can give us a sense of the mix within that group between the CFP form factors collectively, and QSFP28. I'm assuming QSFP28 is a still well below that. Kind of curious if you can give us a thought process of how long it takes before that crosses over, where QSFP28 is larger.

Thanks, Richard. So in terms of the QSFP28 volume, it was almost 50% of the volume in the March quarter. Into the June quarter here, so as Alan had mentioned on prepared remarks, we are ramping QSFP28 form factor, especially for hyperscale. So it's going to grow to be the majority during the June quarter.

Okay. Great. I must have missed that but thanks for the recollection there. Second question would be on 3D sensing. Any sense of how we can think about what gross margins to expect from this revenue stream either initially, and then how we go over the next several quarters would be great to hear, please.

In terms of the gross margin on 3D sensing, so -- what we've been saying is it's significantly north of our corporate gross margin average. And it's going to be -- it should be over 50%.

And I'd say that we're extremely happy with the current existing yields as we start up the ramp. So there's no reason to believe that the initial ramp gross margin would be anything different than that.

And our next question comes from Doug Clark from Goldman Sachs.

I want to come back to Datacom first, actually. I know you talked a little bit about China being a drag on the March quarter. But could you give us a little bit more detail into kind of what led to the, I think, 43% sequential decline in that business?

Doug, in terms of the sequential decline, the bulk of that was due to China. So our China business was down close to 40%, sequentially. And most of the CFP2 volume was going into China. So that's the bulk of the decline. We do see recovery here. Not from CFP, but we do see Datacom recovering in the June quarter, primarily because of the transition to QSFP28 for hyperscale. In the June quarter, both CFP2 and QSFP28 share the same capacity, or manufacturing capacity. We were not able to really shift all of that capacity in the March quarter so we weren't able to fulfill a lot of QSFP28 orders. We're doing that now. And we expect QSFP28 to ramp significantly in the June quarter.

And as a follow-up to that, how long does it take to convert that capacity from one product line to the other?

Yes. So the [tester] conversion is not a big deal. There's components as well as a different wafer that we produce for the QSFP28 to meet the power requirements. So I think we've already pivoted the supply chain and the wafer fab to focus primarily on QSFP28 and the different varieties of short and medium and long reach. So I think we're positioned pretty well to really capitalize on the growth in demand for QSFP28 in the June and September quarters.

Okay. And then I also have one question on 3D sensing, kind of a clarification to a comment prior. I think you said the initial order was for multimillions of units in the September quarter. Does that satisfy the entirety of the September quarter demand? Or do you expect to kind of get additional orders for that period between now and then as well, which could take it into tens of millions of units?

Yes, we're not going to quantify exactly how much it is, but I'd say that the orders that we have today that we categorized as multimillion units is just the very, very start. We expect orders to be streaming in through this quarter in a meaningful way.

And our next question comes from James Kisner from Jefferies.

I want to drill into Datacom for a moment again. I mean, you said the majority -- or that you've given volume comments on QSFP28. I wonder if you could talk about proportion of revenue, is QSFP28 over $20 million at this point? How much of the business is still legacy? I think it was below 20% last quarter -- 10 and 40. I just -- longer-term on CFP2, [Clear] has made some comments here that they think this market is really not that big going forward. What do you think the opportunities for CFP2 at this point? Is it kind of on the wane at this point or could there be some more growth there as it recovers?

So -- Aaron, you want to take the 10 and 40?

Sure. Yes. In terms of the percentage, James, so faster speed, meaning 40 gig and above is 77% in the March quarter. Last quarter it was 83% just to calibrate, but we had more CFP2 volume last quarter. So legacy, 10 gig and below was 23% in the March quarter. In terms of the [dock orders] so our 100-gig Datacom volume was about $25 million-or-so in the March quarter. So $12.5 million was QSFP28.

And as far as the long-term outlook for QSFP28, CFP2, I think, CFP2, first of all, has a long tail. I'd say that it's a matter of inventory buildup in China that created the issue. And there is still a lot of CFPs being deployed in the world, and that transition from CFP to CFP2 is well underway. And I'd say that the demand for QSFP28 is very strong both from that 300- and 500-meter short-reach products in the hyperscale datacenters as well as the 2-kilometer CWDM4. And the transition from CFP2 to QSFP28 LR4 is underway as well, but that transition is going to take a little bit longer. But we're positioned very well. Qualifications are happening and have happened, so I think our broad portfolio of QSFP28 that we showed at OFC has really positioned us well across the portfolio from short reach all the way to 10-kilometer kind of products.

Great. And just a follow-up. Just commercial lasers, it's nice to see that recover, but your gross margin is not what I would have expected, just given the revenue level. Is there start-up costs in there? Is this kind of what you'd expect going forward in terms of gross margins for commercial lasers or might we see that kind of get to the high levels again?

So James, in terms of commercial lasers gross margins, we're at 41%. In terms of where we had been in the past, we were close to 50% in the past. And there's no reason we can't get back there. It's highly mix-dependent. So we have had added capacity capability as we've brought new generations of fiber lasers to market. Unfortunately, it requires investment in terms of some fixed costs that have been put in place. So there is a little bit of a drag on gross margins from that perspective. As Alan had mentioned in the prepared remarks, micromachining lasers have come back in this quarter and represented close to 40% of our mix or $16 million. Those products have very high gross margins, which are higher than our average commercial lasers. So if that can continue -- and most of those lasers go into semiconductor for slicing and dicing. And if that continues and it continues to grow back to where it once was in terms of over $20 million per quarter, there is no reason our gross margin can't get back into the high 40s and close to 50%. So the way we look at it, it could be 3 to 5 quarters out that we get back there.

Again, I'd say just one other dynamics is that in the March quarter, most of the fiber laser business was Gen 2 and that's been in production for quite some time. So the margins as we end of life a product like that, are lower than the new products. So as we bring on the Gen 3 that we started revenue last quarter, and that becomes a bigger percentage of our fiber laser, that will help the overall gross margin of both the fiber laser as well as the overall commercial lasers gross margin.

Great. That's really helpful. I'm going to sneak in just one more. Did you guys comment on OpEx sequentially in your guidance?

No. We didn't break out gross margin or OpEx in our guidance. But in terms of the operating margin and the midpoint there, where we have guided 10%-or-so, if you go look back 4 to 5 quarters ago, we were below 9% in terms of operating margin. So in terms of some of the near-term headwinds, we are keeping a close eye on discretionary spending, so to speak, and keeping expenses under control so we can maximize our operating margin. In terms of fixed costs over those 5 quarters, we've put a lot of fixed cost in place, both on the operational side as well as within OpEx. And you can see that our operating margins are still expanding by at least 100 basis points, which means our variable margins have increased over that period of time, which translates into -- or translates from a richer mix of products, newer products with higher margins.

And our next question comes from Meta Marshall from Morgan Stanley.

Just wanted to circle back on China and just see overall pricing environment. Is there any kind of going back and renegotiating pricing as they work through inventory? And then second, I don't think it's been touched upon, around ACO timing. That had been kind of planned for the second half of the year. Is that still on track? Or is some of that going into China and so there is a little bit of slowdown in getting that product out?

Yes. So, for China pricing, we set pricing annually with our Chinese customers, and there is no renegotiating because demand is softening. I would say that as far as ACO is concerned, we've made tremendous progress over the last 3 months. We've been shipping beta units now, not just evaluation units but beta units to Tier 1s, including in China. But I'd say that we weren't expecting, nor was, I think, the industry expecting China to be consuming a lot of CFP2 ACOs even in this calendar year. That's really more of an end of calendar '17 and really ramp into '18. So I think that's our expectations for China on ACO. Rest of the world, I think, we're still on track over the next couple of quarters to start volume productions as we get qualified outside of China.

And our next question comes from Joseph Wolf from Barclays.

You made a comment about the relative value on the QSFP28 in terms of the revenues and the capacity in terms of the margins. Given the shift right now and the movement, can we see any -- how much room is there for margin improvement in the Datacom business as you get to really high levels of QSFP28 in the mix, 28?

Yes, the majority of QSFP28 in the March quarter was more of the 2-kilometer CWDM4 and then comparing that to a 10-kilometer CFP2, the price is -- average selling price is a factor of 50% lower plus-or-minus a little bit. So I do think that as we introduce new generations of QSFP28s, and we have 2 new products coming out that are focused primarily on cost reduction, as we ramp up volume, we should see continued improvement in gross margin in our QSFP28 Datacom business, even with the hyperscale pricing pressures that we see in the industry. So I think from that perspective, we're in very good position. I do think that CFP2 will pick back up, but we're not counting that in the short term.

And then, just as you look -- if you take this, I guess, relative softness in China and think about what that means in terms of the components that were on allocation where you had supply constraint, how is that shaping up for the rest of the calendar year as you look across those components? What kind of opportunities do have in terms of additional capacity and getting the supply chain set for the pickup in China?

Well, we've added $100 million of CapEx over the last 3 quarters. So that capacity is coming online and becoming productive this quarter and next. And so I think we're very well positioned for when China comes back. We still are constrained on some of our products, but I'd say that we are focused on making sure that when the demand comes back, we have the capacity to supply what our customers need. So the question in my mind is, is it in the September quarter or is it in the December quarter, not is it ever. So I'd say that we're confident in our ability to meet the demand as the demand comes back in China.

So this hasn't changed any of the build plans?

Oh, well yes. So in terms of capacity that Alan was alluding to, it typically takes 6 to 7 months from the time you launch orders and approve it to when it comes -- starts to come online. There's a lot of qualification that takes place. So a lot of the CapEx that landed in the March quarter was launched 6 -- several months ago. Same thing -- as of last quarter, there's some things we're still ramping. For example, pump lasers are still tight. So we're continuously adding capacity there. 100-gig transceivers, we've been capacity-constrained in that regard. So we've added a lot of capacity and more is coming online. In terms of the next couple of quarters, some of the traditional or base products have some near-term demand challenges from China. And it could take a [couple years] to rebound. But we think when it does come back, we're going to be well positioned to take advantage of it because we will have the capacity in place. We've been trying to stay rational as well and not get too far ahead of ourselves in terms of capacity, but unfortunately, it does take 6 to 7 months to get it online.

At this time, there are no further questions. I would now like to turn the call back over to Alan Lowe for any closing remarks.

Thank you, Bruce. I want to thank our customers for their business and our employees for all of their hard work in growing our business, putting us in an excellent position in the market. I believe we are in a worldwide bandwidth expansion, and photonics are increasingly becoming critical to manufacturing in next-generation, high-volume and 3D sensing applications, making the future bright at Lumentum.

We regularly discuss our business at Investor Relations events. These events are listed on our website in the Investor Relations section and are regularly updated.

This concludes our call for today. We would like to thank everyone for attending. And we look forward to talking with you again in another 3 months.

Ladies and gentlemen, thank you for your participation in today's conference. And this does concludes the program. You may all disconnect. Everyone, have a great day.