Coherent Conference Call: Orders scheduled until 2028, 1.6T optical modules are expanding "at an astonishing rate," and 6-inch indium phosphide production capacity has doubled ahead of schedule.

Question AI · What technological breakthroughs are behind the 1.6T optical modules ramp-up?

Driven by two forces—the booming demand for AI computing power and the concurrent evolution of network infrastructure toward higher bandwidth and lower energy consumption—“optical communications giant” Coherent delivered an unexpectedly strong performance. According to its financial report, in Q3 the company achieved record revenue of $180 million, up 21% year over year and 7% quarter over quarter; non-GAAP earnings per share (EPS) surged 55% year over year, and gross margin expanded to 39.6%.

However, as market analysts had previously warned, in the face of a “high-quality earnings report with no major surprises,” profit-taking sentiment took center stage, sending its after-hours share price down by more than 7%. Still, setting aside short-term stock price volatility, during the earnings call Coherent’s management delivered deeper signals about the underlying picture:

The demand side is no longer the problem; the real game is breaking through the capacity bottleneck and executing the future new story (CPO and thermal management).

During the call, Coherent CEO Jim Anderson was explicit: orders are experiencing “step-like” growth, backlog has hit a historical high, and customer scheduling has been extended to 2028 and even to the end of this decade. At the same time, its 6-inch indium phosphide capacity upgrade is progressing ahead of expectations: capacity will double a quarter earlier, and 1.6T optical modules are also accelerating mass volume ramp-up at a “stunning speed.”

Jim Anderson also said that three new technologies—CPO, Multi-Rail, and Thermadite thermal management materials—are poised to be unleashed. Together, their potential total addressable market exceeds $17 billion, giving the company fresh room for growth imagination.

Demand is growing in “step-like” fashion; orders are scheduled out to 2028

On the call, Coherent’s CEO Jim Anderson offered an extremely positive assessment of the current industry outlook. He used multiple highly impactful phrases to describe the ongoing order surge.

“We are at the center of an extraordinary expansion of optical network infrastructure, driven by the rapid growth of AI and rising demand for bandwidth and energy efficiency,” Anderson said plainly in his opening remarks. “This quarter, our order volume has experienced another step function increase, pushing backlog to a record level.”

For the sustainability question that the market is highly focused on, management gave investors reassurance. Anderson made it clear that there is no sign of any weakening in customer demand: “Our visibility continues to extend into the future. Our current orders are scheduled into the 2028 calendar year, and our long-term agreements (LTAs) even extend to the end of this decade.”

The data center and communications segment, which accounts for 75% of the company’s total revenue, is an absolute growth engine. In this quarter, revenue surged more than 40% year over year. Looking ahead to Q4 (the second quarter of the natural calendar year), the company guided that revenue would be between $191 million and $205 million, and gross margin would further improve to 39%—41%.

Breaking through core production capacity: 1.6T ramps out “at an incredible speed,” and 6-inch indium phosphide capacity doubles ahead of schedule

Against the backdrop of a major AI computing boom, the industry’s pain point is not that it lacks orders—it is that it cannot deliver shipments. For Coherent, indium phosphide (InP) capacity is the key bottleneck.

This is also the profit-margin driver that Wall Street cares about most. The upgrade from 3-inch production lines to 6-inch production lines is delivering immediate financial returns.

“Upgrading from 3-inch to 6-inch means the number of devices increases by more than 4 times, while cost is less than half.” Anderson revealed that yields on the 6-inch platform have already surpassed those of traditional 3-inch production lines, and in the third quarter for the first time the company shipped transceivers containing 6-inch components.

To meet “exceptionally strong” demand, the company is applying maximum pressure to accelerate its capacity expansion. “Our goal is to double our internal indium phosphide capacity this year. The good news is, based on how we’re executing, we now expect to achieve this milestone one quarter earlier (i.e., next quarter) than originally planned.” In addition, the company expects that by the end of 2027 calendar year it will double internal indium phosphide capacity again, accomplishing a fourfold expansion of capacity within two years.

When asked by analysts how to allocate these valuable capacities across new and legacy product lines, Anderson’s answer reflected the cold, rational logic of the capital markets: “Overall, our approach to capacity allocation is: we allocate indium phosphide capacity to whatever areas can bring the most and highest profit-margin dollars. Wherever we can generate the largest profit dollars for the company, that’s where we allocate capacity.”

In the highly watched optical module (Transceivers) space, demand is quickly iterating toward higher speeds. Anderson pointed out that 800G products will continue to grow year over year this year, but what deserves even more attention is the 1.6T explosion:

“1.6T optical modules are ramping at an incredibly rapid pace. In fact, their ramp rate is faster than what we had expected a year ago.”

Today, not only EML (electro-absorption modulated laser) solutions, but also 1.6T modules based on silicon photonics (SiPho) have entered mass production in parallel.

Opening up imagination: CPO, Multi-Rail, and the thermal-management “black tech” of “squeezing” computing power

If 800G and the 1.6T optical modules that are rapidly ramping are the foundations supporting the current valuation, then Coherent’s new technology blueprint depicted during the call is its core weapon to hedge future market concerns.

First is the photonic co-packaging (CPO) technology, viewed as a “transformational growth opportunity.” The company expects this to be an incremental market of over $15 billion, backed by Nvidia’s long-term procurement commitments worth billions. CPO revenue from scale-out is expected to start ramping in the second half of this year, while CPO revenue from scale-up is expected to begin in the second half of 2027.

Second, to address the increasingly severe energy consumption and thermal dissipation challenges in AI data centers, Coherent presented a highly compelling “cross-industry” story—transferring materials technology originally used in industrial applications to AI data centers.

Anderson highlighted its proprietary Thermadite material:

“Compared with the copper-based cooling solutions that are commonly used today, Thermadite—or other types of materials we provide—can offer 2 to 5 times better heat transfer than copper.”

He used an extremely vivid analogy to explain the commercial value of this thermal management technology:

“This is a huge win for our customers. Because it means it allows XPU or GPU to run at higher frequencies or higher utilization rates, since it can be cooled more effectively. It’s almost like getting more tokens out of the same CPU or GPU.”

In addition, the company disclosed a Multi-Rail technology under development, aimed at addressing the growing bandwidth demand between AI data centers. The company estimates the market size to be over $2 billion, with revenue expected to start contributing in the first half of 2027.

Full translation of the earnings call transcript is as follows (AI-assisted translation)

Coherent FY2026 Q3 earnings call

Meeting date: May 6, 2026

Company: Coherent

Event: Coherent FY2026 Q3 earnings call

Coherent announced record third-quarter revenue of $180 million, up 21% year over year and 7% quarter over quarter; non-GAAP EPS grew 55% year over year, and gross margin expanded to 39.6%.

Management’s guidance further accelerated, with Q4 revenue expected to be between $191 million and $205 million, gross margin between 39% and 41%, and continued strong visibility into demand through before 2028 supported by growth in AI data centers and communications.

Operator:

Hello, and welcome to the Coherent FY2026 Q3 earnings conference call. I’m pleased to introduce today’s host—Paul Silverstein, Senior Vice President of Investor Relations for Coherent. Please go ahead.

Host:

Thank you, Operator. Good afternoon, everyone. Joining me today are Coherent CEO Jim Anderson and Coherent CFO Sherri Luther. In today’s call, we will review our financial and business performance for Q3 FY2026 and provide our outlook for Q4 FY2026. Our earnings press release can be found in the Investors section of our company website, coherent.com. I’d like to remind everyone that during this call, we may make predictions about future events or future company financial performance, or other forward-looking statements. These statements are subject to certain major risks and uncertainties, and our actual results may differ materially.

Host:

For factors that could affect our future financial performance and business, please refer to the earnings press release for today, the related disclosures in our recently filed 10-K and 10-Q forms, and any 8-K reports we may file with the U.S. Securities and Exchange Commission. All of our statements are based on information available as of May 6, 2026. Except as required by law, we do not assume any obligation to update any such statements. During this call, we will discuss non-GAAP financial measures. You can find the reconciliation tables between these non-GAAP financial measures and the GAAP financial measures in our earnings press release and investor presentation, which are available on the Investors section of coherent.com. Now, I’ll turn the call over to our CEO, Jim Anderson.

Jim Anderson:

Thank you, Paul, and thanks to everyone for joining today’s call. Coherent is a global leader in photonic technology, which underpins AI data center performance and scalability, and is critical to many important industrial applications. We are at the center of a massive expansion of optical network infrastructure, driven by the rapid growth of AI and increasing demand for bandwidth and energy efficiency. As a result, we delivered another strong quarter, achieving accelerated growth, margin expansion, and continued improvement in profitability. Importantly, we see demand across the entire business continue to strengthen. This quarter, our order book once again experienced a stepwise increase, pushing our backlog to a record high.

Jim Anderson:

Customer demand remains exceptionally strong with no sign of weakening. Our visibility continues to extend into the future. Orders are now extending into the 2026 calendar year, and customer long-term agreements extend into the end of this decade. This demand is increasingly translating into near-term shipment and revenue opportunities, as we continue to expand capacity. Given the strength of both near-term and long-term demand, together with our ongoing expansion of production capacity, we expect a sustained period of strong revenue growth in the coming quarters. We expect the June quarter to deliver strong sequential revenue growth. We continue to expect the growth rate in fiscal year 2027 to be higher than in fiscal year 2026. Turning to our operating performance in Q3, on a pro forma basis, revenue grew 9% sequentially and 27% year over year, with the year-over-year growth rate acceleration further accelerating compared to the prior quarter.

Jim Anderson:

Non-GAAP gross margin expanded both sequentially and year over year. The combined effect of revenue growth, margin expansion, and operating leverage drove non-GAAP EPS to increase 55% year over year. The pace of our profitability growth continues to be significantly faster than the pace of revenue growth. We’re pleased with our ongoing execution, but we also see major forward opportunities because we are expanding our business scale to meet the current demand environment. Our data center and communications segment continues to be the primary driver of our growth, accounting for 75% of total company revenue in the third quarter. Growth in this segment once again accelerated this quarter, with revenue up more than 40% year over year. The segment performance is driven jointly by demand acceleration and strong execution across our entire product portfolio.

Jim Anderson:

In our data center business, revenue grew 13% sequentially and 37% year over year, marking the second consecutive quarter of double-digit sequential growth. We expect data center growth to accelerate further this quarter, thanks to exceptionally strong demand, improved supply, and continued progress in our capacity expansion. Demand in our data center business remains exceptionally strong and is broadly distributed across multiple customers and product categories. We expect this quarter’s acceleration to be driven jointly by transceivers and OCS systems. In transceivers, we expect growth to be driven by both 800G and 1.6T. Specifically, we expect 800G revenue to grow year over year in the 2026 calendar year, while 1.6T transceivers will ramp rapidly over the remainder of this calendar year and into next year as broad adoption of 1.6T accelerates among a wide customer base.

Jim Anderson:

Given the exceptionally strong demand environment and tight supply of indium phosphide across the industry, capacity expansion remains one of our top priorities. Importantly, we continue to make excellent progress in ramping 6-inch indium phosphide capacity. This is a key driver of our long-term capacity expansion and an important differentiated competitive advantage for Coherent. We are already seeing the benefits of this capacity ramp in revenue and gross margin and expect these benefits to increase further in the coming quarters. We still expect to meet the goal of doubling internal indium phosphide output capacity by the end of this calendar year. Based on our current execution, we now expect to reach this milestone one quarter earlier than originally planned. We also expect that by the end of the 2027 calendar year, internal indium phosphide capacity will double again.

Jim Anderson:

Our 6-inch platform is producing EML (electro-absorption modulated lasers), CW (continuous wave) lasers, and photodetectors. Yields for these three device categories continue to exceed those of our 3-inch production lines. This quarter, we shipped the first batch of transceivers containing components produced on the 6-inch production line. These shipments contributed to both sequential revenue growth and gross margin improvements. The initial contribution from 6-inch capacity comes from our factory in Sherman, Texas—the world’s most advanced indium phosphide production base—which will play an important role in expanding CW laser capacity for our CPO (co-packaged optics) solutions (including solutions supporting our collaboration with Nvidia). Given the success of the 6-inch capacity ramp to date, we also announced plans to start 6-inch indium phosphide production at a third site in Zurich. Overall, we are very satisfied with the execution of our production teams.

Jim Anderson:

As we continue to expand 6-inch capacity, we expect increasing benefits in both revenue and gross margin for our transceiver and CPO product lines in the coming quarters. We expect OCS revenue to grow this quarter as capacity expansion ramps to meet demand. We have raised our estimate of the OCS market opportunity to more than $4 billion, reflecting its expanding application scenarios in data center interconnects, horizontal and vertical expansion networks, and continued broadening customer engagement. We believe OCS will also expand our role in higher-value layers of AI network infrastructure. We recently addressed a production capacity bottleneck and are now rapidly scaling output at two production facilities. We expect strong sequential revenue growth in the coming quarters as improved production translates into higher shipments and conversion of backlog into orders.

Jim Anderson:

We are also making strong progress in co-packaged optics (CPO). We believe this represents one of Coherent’s most important long-term growth opportunities. As we discussed earlier, CPO expands our role in AI data center architectures, especially in the vertical extension portion of networks. Optical technology is expected to increasingly supplement and ultimately replace copper cabling. We believe CPO represents an incremental addressable market opportunity of over $15 billion. In March of this year, we announced a strategic partnership with Nvidia focused on multiple CPO-related products and solutions. This partnership includes Nvidia’s $2 billion equity investment in Coherent and a multi-year supply agreement that extends through the end of this decade. The agreement covers multiple CPO-related products, including our high-power CW lasers, and provides meaningful long-term visibility into future demand. More broadly, our CPO opportunity is supported by the breadth and depth of Coherent’s photonic technology platform.

Jim Anderson:

We believe that the breadth of our photonic technology and our manufacturing scale position us very favorably to support a wide range of customer needs across key optical components, subsystems, and higher-level components. We expect initial scale-out CPO revenue to begin climbing in the second half of this calendar year, while scale-up CPO revenue is expected to begin climbing in the second half of the 2027 calendar year. Beyond Nvidia, we are also actively engaging with multiple other customers regarding broad CPO and MPO opportunities. Overall, we believe CPO will become an important contributor to Coherent’s long-term revenue growth and gross margin expansion and will further strengthen our strategic position in AI data center infrastructure.

Jim Anderson:

Turning to our communications business, Q3 revenue growth accelerated sharply, with sequential revenue growth of 16% and year-over-year growth of 60%, driven by strong demand for data center interconnects, cross-center expansion, and traditional telecom applications. We expect this segment to deliver strong sequential growth again this quarter. Demand remains broadly distributed across customers, products, and end applications. We see strong momentum across our entire communications product portfolio, including components, modules, and systems, reflecting favorable market conditions and Coherent’s strong competitive positioning. In particular, we continue to see strong demand for our DCI (data center interconnect) solutions, including ZR and ZR+ transceivers, as well as strong demand for our broader transmission product portfolio. One additional growth driver that we are especially excited about is Multi-Rail. These solutions address the urgent need for more bandwidth and connectivity between AI data centers as workloads become increasingly dispersed across multiple locations.

Jim Anderson:

We believe Multi-Rail represents a major expansion of our addressable market opportunity in communications, and we expect initial revenue to start ramping in the first half of the 2027 calendar year. Overall, we believe our communications business is positioned for sustained strong growth, supported by current demand, our continually expanding product portfolio, and the ramp-up of important new platforms over time. In our data center and communications segments, the breadth and depth of Coherent’s photonic technology portfolio, together with our manufacturing scale, continues to resonate strongly with customers. Therefore, we have signed or are finalizing long-term supply agreements with multiple strategic customers. These agreements include multi-year demand commitments and upfront investments to support capacity expansion. Moving to our industrial segment, on a pro forma basis, revenue showed slight sequential and year-over-year declines, reflecting the ongoing softness in parts of the broader industrial market.

Jim Anderson:

However, we are seeing encouraging signs of improvement, particularly in semiconductor capital equipment, where order volumes have increased significantly. We expect this improving demand to begin contributing to revenue growth in this quarter and support further sequential improvements for the remainder of the calendar year. Over the long term, we see meaningful incremental growth opportunities for our industrial technologies in AI data center applications. At OFC (Optical Fiber Communication Conference and Exhibition), we focused on showcasing our data center XPU cooling solutions and thermoelectric generators—products designed to address the increasing thermal dissipation and power challenges brought by large AI data centers. Our proprietary Thermadite material can improve heat dissipation performance and help increase XPU efficiency, and our advanced materials used for thermoelectric generation can improve data center power efficiency by recovering waste heat. We are engaging with multiple strategic customers on these technologies and believe they represent meaningful expansions of our long-term market opportunities.

Jim Anderson:

We expect revenues from these products to begin ramping in the second half of the 2027 calendar year. Overall, although the industrial segment’s contribution to growth is currently smaller than the data center and communications segments, we believe it will become an increasingly important source of incremental revenue and diversification over time. In summary, we delivered another strong quarter, with accelerating revenue growth, margin expansion, and continued improvements in visibility into future demand. We are in a highly favorable demand environment driven by AI data center expansion, and we believe Coherent is uniquely positioned to capture this opportunity with the breadth of our photonic portfolio, our manufacturing scale, ongoing capacity expansion, and the accelerating conversion of backlog into revenue and shipments. I want to thank the entire Coherent team for their strong execution and continuous innovation. Now I’ll turn the call over to Sherri.

Sherri Luther:

Thank you, Jim. In our third quarter, we achieved accelerated double-digit year-over-year revenue growth and significant gross margin expansion, with profitability improving substantially. We strategically increased capital investments to expand internal capacity to support the rapidly growing demand in data centers and communications. In addition, we continued strengthening our balance sheet and reduced debt leverage to below 1x. Now, I will summarize our third-quarter performance. Third-quarter revenue reached a record high of $1.8 billion, up 7% sequentially from Q2 and up 21% year over year. The growth momentum was driven by increases in AI data center and communications demand. On a pro forma basis, revenue grew 9% sequentially and 27% year over year, excluding revenue from our aerospace and defense business and the Munich product division in Germany, both of which were sold in Q1 and Q3 respectively.

Sherri Luther:

Our non-GAAP gross margin for the third quarter was 39.6%, an increase of 57 basis points sequentially compared to last quarter and 105 basis points compared to the same period last year. We continue to execute our gross margin expansion strategy, achieving gross margin improvements both sequentially and year over year in our data center and communications segments. These improvements were mainly driven by lower product raw material costs, improved yields on 6-inch indium phosphide, and significant benefits from pricing optimization. Non-GAAP operating expenses in the third quarter were $348 million, compared with $321 million in the prior quarter and $297 million in the same period last year. R&D expenses as a percentage of revenue increased to 9.9% in the third quarter, compared with 9.4% in the prior quarter and 9.4% in the same period last year.

Sherri Luther:

The sequential and year-over-year increases in R&D expenses are mainly concentrated in the product roadmaps for the data center and communications segments. These investments focus on multiple near-term and long-term revenue growth drivers—transceivers and CPO (co-packaged optics), as well as OCS (optical circuit switching) and MultiRail, which are new high-margin, high-value system solutions with strong growth potential. We remain focused on investments with the highest return on investment to drive future growth. Selling, general, and administrative (SG&A) expenses as a percentage of revenue declined to 9.4% in the third quarter from 9.6% in the prior quarter and 10.4% in the same period last year, reflecting continued progress in improving efficiency and enhancing SG&A leverage. We have seen benefits from shared services initiatives in low-cost regions. This is reflected in administrative functions, as we streamline processes to achieve better leverage effects and efficiency.

Sherri Luther:

Additionally, our ERP integration project has made significant progress, and most of our operations are now running on a single ERP platform. We expect these initiatives to generate additional benefits in the fourth quarter and more significant benefits in fiscal year 2027. Our non-GAAP operating margin in the third quarter increased to 20.3%, from 19.9% in the prior quarter and 18.6% in the same period last year, driven by strong revenue growth and continued gross margin expansion. Non-GAAP diluted EPS in the third quarter was $1.41, up 9% sequentially and 55% year over year. The acceleration in profitable growth outpaced revenue growth thanks to strong revenue performance and gross margin expansion.

Sherri Luther:

Our cash balance increased from $1.5 billion in the prior quarter to $3.0 billion, mainly because of the $2 billion equity investment from Nvidia announced on March 2, 2026. In this quarter, our capital allocation prioritized investments to drive long-term revenue growth and profitability, including R&D product roadmap investments and capacity expansion investments in our data center and communications businesses. We also repaid $162 million of debt in the quarter, reducing debt leverage to 0.5x, below 1.7x in the second quarter and 2.1x in the same period last year. Our capital expenditures increased to $290 million, from $154 million in the prior quarter and $112 million in the same period last year.

Sherri Luther:

These investments focus on expanding our internal capacity to support strong demand in data centers and communications. Due to strong order volume and rapidly growing demand, we expect fourth-quarter capital expenditures to increase sequentially. We continue to execute our capacity expansion plan as scheduled. With a strong balance sheet and a continued focus on improving profitability, we are well prepared to rapidly scale production capacity through investments in order to meet unprecedented customer demand. Reminder: we completed the sale of the Munich product division at the end of January. For reference, over the past four quarters this business contributed an average of $25 million in revenue per quarter, and its gross margin is far below Coherent’s overall corporate gross margin. Our third-quarter performance includes $8 million of revenue from this business. Now I will move on to our guidance for Q4 FY2026.

Sherri Luther:

We expect revenue to be between $191 million and $205 million. We expect non-GAAP gross margin to be between 39% and 41%. We expect total non-GAAP operating expenses to be between $360 million and $380 million. We expect our non-GAAP tax rate for the quarter to be between 18% and 20%. We expect non-GAAP EPS to be between $1.52 and $1.72. With strong backlog and excellent visibility, we remain focused on rapidly expanding our internal capacity through investments that drive the company’s long-term growth and profitability. We will continue to allocate capital in a disciplined manner, execute our long-term financial target model, and create lasting shareholder value. That concludes my prepared remarks.

Sherri Luther:

Operator, please open the call for Q&A.

Operator:

Thank you. We will now take the first question from Samik Chatterjee of Morgan Stanley. Please go ahead.

Samik Chatterjee:

Hi. Thank you for taking my question, and congratulations on these strong performance metrics. Jim, maybe I can start with the guidance for the June quarter. This implies acceleration compared with Q3, especially when I look at the full year, where you managed to accelerate sequential revenue growth in each quarter. Maybe you could dig deeper into what demand-side factors helped you achieve this acceleration; perhaps you can also explain it from the supply side—how supply helped to achieve it. I have a follow-up as well. Thanks.

Jim Anderson:

Yes. Thank you for your question, Samik. Yes—if you look at the midpoint of the guidance for the June quarter, we do expect growth to accelerate compared with the previous quarter. If you also look at the year-over-year growth rate—we truly believe that the current June quarter represents a new inflection point in our future revenue growth rate. Growth is faster this quarter, and looking into FY2027 (starting July), we expect the FY2027 growth rate to be higher than FY2026. On the demand side, I would say demand looks very strong right now, both in terms of demand scale and in terms of our visibility into demand.

Jim Anderson:

If we only look at orders from the previous quarter, orders increased significantly compared with the quarter before, hitting a record high; backlog is tremendous, and we have already received orders extending into the 2028 calendar year. There is huge demand in front of us, and we also have excellent visibility into those demands. The demand is coming from the areas you would expect—data center growth, including transceivers, and some new growth vectors that we are launching, as well as growth in the communications space. Perhaps more importantly, on the supply side, that might be the primary focus. Demand looks very good. What we’re doing is raising supply very quickly. In this quarter and going forward, in the next few quarters we will increase capacity significantly.

Jim Anderson:

The best single example is the indium phosphide capacity coming online. Indium phosphide has been our key bottleneck for the last several quarters. That has also been the bottleneck across the whole industry. Our goal this year is to double indium phosphide capacity. The good news is, based on how things are executing now, we will achieve this goal one quarter earlier, in the next quarter. Looking into the calendar year after next, we expect indium phosphide capacity will double again. That means capacity will quadruple over two years. That sounds very good. I believe this really unlocks the acceleration in our future revenue growth. That is broadly on top of all existing businesses.

Jim Anderson:

In addition, some new growth areas and vectors are also coming online. OCS is scaling up; we expect it to contribute growth this quarter and achieve sequential growth. CPO revenue will begin in the second half of this year. We believe this is entirely incremental revenue. Our Multi-Rail system will begin contributing revenue in the first half of the next calendar year. We believe thermal solutions will begin generating revenue in the second half of the 2027 calendar year. We have these multiple growth vectors stacking on top of the growth in our existing businesses. We are very optimistic about growth and the acceleration ahead.

Samik Chatterjee:

Understood. Understood. Thank you, Jim. Perhaps just a follow-up in a similar direction. You mentioned the acceleration in indium phosphide capacity. Given that you slightly beat the 2x target in year one, how should we think about the potential upside or acceleration possibilities of the 2x target next year? As an investor, how should investors view the impact on gross margin? How important is this? When would it start to materially affect your gross margin trajectory? Thanks.

Jim Anderson:

Thank you, Samik. Actually, on the second part of your question about gross margin, we have already started to see the impact of the 6-inch indium phosphide capacity. The cost structure is much better. Compared with 3-inch, the chip count is over 4 times, and cost is less than half. We started seeing contributions to gross margin expansion in Q3 FY2026. As Sherri said in her prepared remarks, I think our guidance for the current June quarter shows gross margin will increase sequentially. Similarly, one of the reasons driving gross margin expansion is the 6-inch indium phosphide capacity, which gives us a better cost structure. Overall, I am very satisfied with how the 6-inch indium phosphide ramp is executing. There are two factors.

Jim Anderson:

One is the increase in raw capacity, but equally important is yield. The team has executed beyond plan in ramping up raw capacity, and we are also seeing very healthy yields. We are currently producing at scale three different types of devices: EML (electro-absorption modulated lasers), CW (continuous wave) lasers, and PD (photodetectors). The yields for all three device types on 6-inch are higher than our production yields on 3-inch. Texas is the first facility where we began ramping 6-inch. We are very satisfied with progress there. Because we started seeing such strong yields from Texas—this is the leading indium phosphide production facility globally—we began production in Sweden as well. Now we are announcing the third location that will start 6-inch indium phosphide production: Zurich.

Jim Anderson:

We will begin seeing output from that third site at the beginning of calendar 2027. This 6-inch ramp not only unlocks a large amount of additional growth, but as its share grows within our overall indium phosphide production capacity, it will certainly contribute to gross margin improvement as well.

Samik Chatterjee:

Perfect. Great. Thank you for answering my questions.

Operator:

Thank you. We will now take the next question from Simon Leopold of Raymond James. Please go ahead.

Simon Leopold:

Thank you very much for taking the question. The first thing I’d like you to respond to is that investors compare your company with major competitors’ forecasts, and there appears to be a clear gap in categories such as OCS and CPO. How do you explain this difference? Then I have a brief follow-up.

---

Jim Anderson:

I think, Simon, on these two new growth areas, we are very optimistic about future growth. Regarding OCS, you know that at OFC in the past few months, we doubled our market opportunity forecast for that market. The revenue growth rate—our guidance for sequential growth this quarter—includes part of the growth coming from OCS systems. We are very satisfied with the differentiation of our technology. It is highly differentiated, providing both higher reliability and better energy efficiency. We are very optimistic about both the long-term and short-term growth outlook for this product line. What we’re truly focused on is ramping capacity as quickly as possible.

Jim Anderson:

As I mentioned in my prepared remarks, over the past few months we have made breakthroughs in eliminating production capacity bottlenecks. This allows us to ramp output at a faster pace. We are expanding capacity simultaneously at two sites. We are optimistic about OCS, both in terms of long-term opportunities and near-term capacity ramp. For CPO, I believe it is a transformational growth opportunity for the company. We see this market exceeding $15 billion, and in the coming few years it could still be a conservative estimate. Our CPO revenue will begin in the second half of this calendar year. At that time, it will be early scale-out CPO revenue.

Jim Anderson:

We expect scale-up CPO revenue to begin appearing in the second half of the 2027 calendar year. We are working with multiple customers. Obviously, we have publicly announced a partnership with NVIDIA centered on CPO. This is a multi-billion-dollar agreement extending into the end of this decade. Importantly, it covers multiple different CPO solution types. If you look at what we can provide in CPO solutions—not just lasers, right? Of course, we provide high-power continuous wave (CW) lasers. In addition, we provide external laser source modules. We can provide Fiber Array Units, including micro-lens arrays, and also polarization-maintaining fiber. We have our own fiber, which will be provided in these solutions.

Jim Anderson:

Within that external laser source, we provide all components—not just the laser, but also isolators and thermoelectric coolers. We expect to have a lot of content in CPO, and I think this will become a major new growth area for the company. I think we are in a very, very strong position in CPO. As I said, the initial revenue will begin later this year, in the late part of this calendar year.

Simon Leopold:

Very good. As a follow-up, I understand you don’t want to micromanage the breakdown of each product detail, but I’d like you to confirm whether the revenue from the 1.6T transceivers in the March quarter exceeded $100 million. If not, when can we reach that milestone? Thanks.

Jim Anderson:

Yes. We don’t disclose separately the revenues for each data rate within the transceiver business. You know, we expect 800G to grow this year and it may grow again in the next calendar year. Based on that, 1.6T is ramping up at an extremely fast rate. In fact, as I think we shared previously, the ramp rate of 1.6T is actually faster than what we had expected about a year ago, and that makes us very happy. If you look at our incremental growth or sequential growth, in this quarter a large part was driven by 1.6T—that is, the ramp of 1.6T.

Jim Anderson:

We expect that 1.6T will not only contribute to sequential growth this quarter, but also continue to ramp at a very fast pace in the coming few quarters. I believe the real growth driver in the transceiver business is the combined contribution of 800G and 1.6T—not only this calendar year, but also next calendar year.

Simon Leopold:

Thank you.

Operator:

Thank you. We will now take the next question from Thomas O’Malley of Barclays. Please go ahead.

Thomas O’Malley:

Hi everyone. Thank you for taking my question. My first question is about gross margin. If I look at gross margin for March at 39.6%, and then look at last year’s gross margin at 38.5%, the year-over-year incremental is about 44%. Since then—what I mean is, you added 6-inch production, indium phosphide capacity is almost doubled, and you exited some businesses. In fact, for your data center business, you can assume that a certain portion of the communications business belongs to this area, and the growth there is also very strong. Why didn’t gross margin capture more incremental pass-through? Are there any headwinds you can point to that are preventing gross margin from breaking through?

Sherri Luther:

Thank you, Thomas. On gross margin, I want to highlight a few points: if you look back to the end of Q4 2025, in the past eight quarters, we have had gross margin increase sequentially in seven of those quarters. If we add the recent improvement of 57 basis points in the third quarter, the cumulative increase is about 530 basis points. If we add the midpoint number from our fourth-quarter guidance, the cumulative improvement reaches 570 basis points. I think that’s a pretty solid progress. My meaning is that we haven’t finished yet, but I’m satisfied with the progress we’ve made.

Sherri Luther:

You know, the investor day goals we set last year were to exceed 42%. We are very focused on ensuring that we meet that target. When you examine the drivers of our gross margin expansion strategy that we continue executing quarter after quarter—including cost reductions, yield improvements, and pricing optimization—each of these areas in our third quarter had a fairly significant increase versus the prior quarter. I briefly mentioned this in my prepared remarks. From the cost reduction perspective, we have seen improvements coming from 6-inch indium phosphide.

Sherri Luther:

We discussed the fact that after upgrading from 3-inch to 6-inch, costs are cut in half. We are already enjoying the benefits brought by 6-inch. We also discussed the yield improvement on 6-inch that we saw in the second quarter. As we continue to ramp, yields keep improving. We talked about advancing at two sites in parallel, and another site that is about to be deployed. I expect that as we put that additional site into operation and continue to increase 6-inch production, 6-inch will continue to improve, which will continue to benefit our gross margin. Other cost reduction areas we are seeing are actually mainly concentrated in our data center and communications business.

Sherri Luther:

The vast majority of our gross margin improvement does come from data center and communications. I am very pleased with this progress. We’re also seeing benefits from pricing optimization—this shows significant improvement both sequentially and year over year. It’s not only in the industrial business; in fact, in our data center and communications businesses it’s also quite substantial. I’m very satisfied with the progress to date. We will continue to work to achieve our targets, and we are highly focused on that. I’m quite satisfied with where we are right now. I think we’re still in the early stage.

Thomas O’Malley:

Then as a follow-up, Jim, you mentioned in your opening remarks that some bottlenecks in the OCS business are being resolved.

Thomas O’Malley:

What exactly do you mean, and how much impact will it have on production output?

Jim Anderson:

There are some internal components—or some components produced internally at Coherent—that constrain the pace at which we can expand our capacity. We have been able to significantly increase production volume of internal components. This truly unlocks the acceleration in our capacity expansion. In the past one or two months, we’ve seen very good improvement in production speed, and we expect this momentum to continue. We’ve seen the OCS production ramp be much faster than a few months ago, which is really very good.

Operator:

Thank you. We will now take the next question from Blaine Curtis of Jefferies. Please go ahead.

Blaine Curtis:

Hi everyone. Thank you for taking my question. I actually want to ask about scale-across, which is becoming a hot topic. You mentioned it in the communications business. Maybe you can talk about what the current state of things is there. When you look ahead to FY2027, how would you describe the growth trajectory of scale-across?

---

Jim Anderson:

Thank you, Blaine. We are seeing tremendous growth in the scale-across portion of the business. This is within our communications business segment, and I mentioned it in my prepared remarks. Scale-across—or DCI—is also within that communications segment, along with traditional telecom business. We are seeing the fastest growth in that scale-across portion. You know, in the most recent quarter we saw 16% sequential growth and 60% year-over-year growth. Here, like in data centers, demand is extraordinary. Visibility is extraordinary. We have entered into long-term agreements (LTAs) with customers in this segment. We are seeing broad growth across nearly every product in our segment, and the customer base is also very broad.

---

Jim Anderson:

Let me walk you through the products in this segment. You know it will include components such as pump lasers. It will include modules such as ZR+ transceivers, including the growing 100G, 400G, and 800G ZR+. It will also include line cards and amplifiers, and full systems. Given the demand we see in front of us and the visibility we have into this area, we expect this space to be a very strong growth area for the future. We believe one new system that will continue to accelerate our growth here is Multi-Rail.

---

Jim Anderson:

Our Multi-Rail technology was highlighted at OFC. It helps deliver huge capacity increases within the same power and physical space as the previously addressed solutions. This is a huge benefit for customers. We have many highly differentiated component technologies integrated into the system, which puts us in a very favorable position. We are selling full systems, and we expect revenues to begin in the first half of the 2027 calendar year. You know, this is just another growth driver layered on top. Growth in this area is very strong, and given the strong growth outlook we see, we expect this to continue.

---

Blaine Curtis:

Thank you, Jim. I just want to follow up on Thomas O’Malley’s question about gross margin. I’d like to better understand those tailwinds. You mentioned 6-inch as the biggest driver. I assume the 6-inch products you’re shipping—along with the unit volumes—are still relatively small. Are there any startup costs that will gradually fade, and is that where the savings come from? Does 1.6T also bring gross margin improvement?

---

Jim Anderson:

When I mentioned 6-inch in the previous quarter, I said it is one of the enabling factors. In fact, there were many other contributing factors to gross margin expansion in the prior quarter. It’s similar in our guidance for this quarter. 6-inch is a contributing factor, but there are other factors as well. Pricing and other cost structure improvements are also drivers. Yes, I would say we are still in a fairly early stage in ramping 6-inch capacity. If you consider 6-inch, last quarter we shipped the first batch of transceivers containing components from our 6-inch production line—that was the initial production. It will grow significantly over the next few quarters. I think most of the benefits of 6-inch are still ahead of us.

---

Jim Anderson:

You know, if you consider the total amount of capacity doubling and the fact that all that doubled capacity is 6-inch, then by the end of this calendar year, in the next quarter, half of our capacity will be 6-inch. I think the benefits from 6-inch are more ahead of us. Regarding 1.6T, we do believe it is favorable for gross margin. We expect, as we have seen in prior speed transitions, that at the beginning of the new data-rate lifecycle, gross margins are typically better than at the previous data-rate. We expect 1.6T to be favorable for gross margin in the transceiver business.

---

Blaine Curtis:

Thanks, Jim.

---

Operator:

Thank you. We will now take the next question from George Notter of Wolfe Research. Please go ahead.

George Notter:

Hi everyone. Thank you. I was just wondering if you could tell us more about the new long-term agreements you are signing. Obviously, we learned a lot from the NVIDIA deal that you mentioned, and you also said there are some other deals that you’ve already reached. What can you tell us about these deals—like, you know, how big they are? What kind of term are we talking about? Do these deals help fund your capital expansion? For example, you know, from a financial perspective, it would be interesting to hear more details. Thanks.

Jim Anderson:

Yes. Thank you, George. Yes, we signed several additional long-term agreements in the prior quarter. I’ll say there are some other discussions still in progress. We expect to complete some additional long-term agreements soon in the current quarter. Those, you know, long-term agreements typically have three parts. You asked about capital expenditure commitments. Yes—typically there is upfront investment from customers to help pay for capital expenditures, and that can take a number of different forms. Usually there is some upfront investment, which in a sense represents customer interest and binding, and we view that very positively. Of course, there is also supply commitments from our side.

Jim Anderson:

The third element is, almost always, some minimum-level demand commitment from the customer to ensure that the capacity is utilized. That is the three parts of long-term agreements. Almost every long-term agreement includes these three parts. Yes, I would say we made good progress on additional long-term agreements last quarter, and we expect more long-term agreements to come. Yes, it’s quite a sizable scale.

George Notter:

Yes. Regarding the types of customers here, what can you say? Are they cloud service providers? Are they systems manufacturers? What else can you say? Thanks.

Jim Anderson:

Both, right? We expect long-term agreements from hyperscale cloud service providers and other systems customers. I expect both.

George Notter:

Thank you.

Jim Anderson:

Sure.

Operator:

Thank you. We will now take the next question from Vivek Arya of Bank of America Securities. Please go ahead.

Vivek Arya:

Hi. I’m Michael Mani, on behalf of Vivek Arya. Thank you for taking our questions. I’d like to go deeper into some of the long-term agreements related to CPO—including NVIDIA, but maybe also other deals you’re looking at over the next few years. What combination is in these agreements between lasers, ELS modules (which you highlighted at OFC), and the various other components you can sell into CPO solutions, such as Fiber Array Units? How does this differ across customers? And depending on the deal, what are the trade-offs? Thanks.

Jim Anderson:

This could vary by customer. You know, it’s important to remember that we bring a very broad CPO technology portfolio to customers. I think that’s a real advantage for us. At OFC, we laid out all the different types of technologies we can bring into CPO solutions. Lasers—high-power continuous wave lasers—are clearly an important component, but not the only one. We can also provide 200G and, going forward, 400G VCSEL. In some applications, VCSEL can be a better laser technology, such as in near-packaged optics. Beyond that, if you look at external laser sources, we can provide that module.

Jim Anderson:

In those, almost all of the key optical components are also developed internally. Not just the lasers, but also isolators and thermoelectric coolers. All components within that are viewed as a big advantage by customers because we don’t rely on others for these technologies. And then the actual Fiber Array Unit—this is the component that connects switching chips or XPUs to the panel or external laser source modules. We can also provide entire components because we have lens arrays and polarization-maintaining fiber. We have all the components needed for CPO solutions. I would say that most customers are using it—if not all of these combinations, then a significant portion of them.

Vivek Arya:

Great. Thanks. For my follow-up, I just want to ask about the two incremental opportunities you highlighted for 2027—Multi-Rail and thermal management products. You said Multi-Rail revenue timing is first half, and I think thermal management products are second half. From now until then, from a customer perspective, what milestones are there? When will we get a better sense of how large this growth really is? What does the competitive landscape look like for the two areas? And if you can clarify, in what areas do you believe you’re particularly differentiated? Thanks.

Jim Anderson:

Yes. Michael, let me start with Multi-Rail, which is the nearer-term one. I would say the milestones are the typical engineering milestones that we go through with customers. You know it includes qualification, pilot runs—very normal engineering milestones—we’re pushing that forward. Again, we expect revenue to start in the first half of 2027. I think as we get closer to that revenue growth, we can provide more, and better, understanding of the pace and cadence of that revenue growth. We view it as a substantive new product line with a significant revenue opportunity. What I mean is that we estimate the market size for Multi-Rail at at least $2 billion over the next few years, and it could be larger.

Jim Anderson:

Our technologies are very differentiated. For Multi-Rail, the truly critical point is the underlying technology. I won’t go into a lot of technical detail, because those are already covered at OFC. But within Multi-Rail there are many key components that are unique to us, or that we have unique differentiation in, which puts us in a very favorable position. We are very satisfied with our competitive positioning for Multi-Rail. Then the second part of your question—absolutely, thank you for asking about thermal solutions. We’re very excited about this. This is about taking industrial technologies and some material technologies that we apply to industrial markets and reusing them in data centers. One example is our Thermadite technology. Thermadite is a proprietary material only Coherent provides.

Jim Anderson:

If you look at Thermadite applied to, say, cooling for switching chips or XPU or ASIC chips, compared with the currently common copper-based thermal solutions, Thermadite—or other types of materials we can provide—can deliver 2 times, and sometimes even 5 times better thermal conductivity than copper solutions. This is a huge improvement for customers because it means if we use one of those thermal solutions with 2 to 5 times better thermal performance, it allows, for example, XPU or GPU to run at higher frequencies or higher utilization, because it can be cooled more effectively.

Jim Anderson:

It’s almost like getting more tokens from the same CPU or GPU. That’s a huge win for our customers. We’re really excited about it. Customer engagement there is very strong. Again, it’s going through normal engineering milestones, but we expect revenue to arrive in the second half of next year. By the way, another great technology I want to mention is our thermoelectric generator. We collect waste heat—from CPUs or GPUs—then convert that waste heat back into electricity and pump it back to the data center. That is a huge efficiency improvement for data center power efficiency. Yes, we are excited about these new thermal solutions.

Vivek Arya:

Thank you.

Operator:

Thank you. We will now take the next question from Papa Sylla of Citigroup. Please go ahead.

Papa Sylla:

Thank you. Thanks for taking my question. Congratulations on the strong performance. Maybe, Jim, my first question is about overall pricing at the transceiver level. Obviously, you are both a seller of transceivers and a buyer of lasers and electrical components. At least yesterday or over the past few days, we heard some news about laser price increases, especially for EML (electro-absorption modulated lasers). I’m curious whether you are seeing this on one side, and if so, whether you can pass those costs through at the transceiver level. Given the supply-demand imbalance, do you have enough levers at the transceiver level to raise pricing?

Jim Anderson:

Okay. Let me start with pricing, and then I’ll come back to the cost question. On pricing—yes, I think pricing is very healthy and the pricing dynamics are very healthy. Because of supply-demand dynamics, I think pricing has been very good, right? Every time we transition to a new data rate, one thing always happens: the average selling price of the new data rate increases. The pricing for 1.6T is higher than 800G, and so on. I think the pricing dynamics are very healthy. On costs—remember, most of the components used in our transceivers are sourced internally. This insulates us from external component price increases to a certain extent, providing some buffering.

Jim Anderson:

Of course, we do use some externally sourced components. We do so for strategic reasons. Yes, we believe we have successfully passed through those external component price increases, or offset them through our own internal production. The combination of pricing and costs results in higher gross margin. I think Sherri mentioned in her prepared remarks that especially in data center and communications, we saw that the majority of gross margin improvement came from this part of our business.

Papa Sylla:

Understood. That’s very helpful. For my follow-up question, it seems obvious that you’re seeing very strong 1.6T demand, at least during early deployment. I’m curious if you can talk a bit about the product mix you’re seeing among EML, silicon photonics (SiPho), and even VCSEL. As a follow-up, overall, does selling more silicon photonics transceivers versus EML—or vice versa—have an impact on gross margin?

Jim Anderson:

Regarding the second part of your question, we actually haven’t seen any significant gross margin difference between EML and silicon photonics transceivers. The gross margins for the two transceiver types are on the same level. We are expanding production for both at the same time. For 1.6T, we are expanding 1.6T capacity for both EML-based and silicon-photonics-based solutions. Remember that even silicon-photonics transceivers still need continuous wave (CW) lasers based on indium phosphide, right? In either case, they need indium phosphide capacity again—which links back to one of the reasons we are driving rapid expansion of indium phosphide capacity. For us, the combination between EML and silicon photonics is determined by customer applications.

Jim Anderson:

We work with customers to determine which of these two technologies is better suited for their applications. Depending on the type of application, there are pros and cons. We do expect that VCSEL will be adopted later as well. Our 200G VCSEL development progress is going very well. In addition to 200G VCSEL used for transceivers, we also expect 200G VCSEL to be adopted in some CPO (co-packaged optics) applications or NPO applications. The initial 1.6T capacity expansion is based on a combination of EML-based and silicon-photonics-based 1.6T.

Papa Sylla:

Understood. Thank you, Jim.

Operator:

Thank you. We will now take the next question from Ruben Roy of Stifel. Please go ahead.

Ruben Roy:

Yes, thank you. Jim, discussions about CPO seem to have clearly accelerated since the beginning of the year, after OFC—then even in the past few weeks, some of your peers and you yourselves have been talking about it. The first question is just a clarification about the second half scale-out and the 2027 scale-up capacity ramp. Are these capacity ramps specifically related to Nvidia, or are other customers also contributing to your initial scale-out CPO revenue? The second part of the question is: when you think about CPO and new opportunities, such as Multi-Rail and the components that make up Multi-Rail, as I understand it, some of those have a different—higher—margin structure than other indium phosphide or silicon-photonics components.

Ruben Roy:

How do you think about allocating capacity among these newer growth areas over the next 12 to 18 months?

Jim Anderson:

Thank you, Ruben. On CPO, obviously they could be our main customer for CPO. We do expect other customers to follow as well. We are talking with multiple different customers. The customer base is fairly broad, and we expect to deliver CPO solutions across multiple customers. Nvidia is certainly one of our main customers. On the second question about Multi-Rail, the gross margin structure for that portion of the business is indeed higher. You’re absolutely right: within Multi-Rail solutions, some particular components have fairly high gross margins, and those components depend on indium phosphide capacity as well.

Jim Anderson:

Overall, how we think about capacity allocation is to allocate indium phosphide capacity to the business that can generate the highest profit amount. Whatever can generate the largest profit amount for the company, we allocate capacity there.

Operator:

Thank you. We will now take the next question from Sean O’Loughlin of TD Cowen. Please go ahead.

Sean O’Loughlin:

Hi, Jim, Sherry. Thank you for letting me join the call. As always, congratulations on a solid set of results. One thing I think addresses, to a large extent, the question that both Blaine and Tom raised earlier in the call is: investors are trying to better understand the timing difference when you expand 6-inch indium phosphide capacity—from the initial SKU shipped, initial transceivers, to achieving revenue (as you’ve mentioned)—versus having the full qualification process at certain customers for mass production. I’m going to ask this in a way that I know may not be the perfect framing, but if I think about doubling indium phosphide capacity next quarter, why doesn’t that translate into doubling revenue?

Sean O’Loughlin:

I think that’s the confusion I’ve had talking with a lot of people. If you could comment on that, it would be helpful.

Jim Anderson:

Yes. Please remember that there is a time lag between the indium phosphide devices and when we actually ship transceivers, right? When indium phosphide devices—whether EML or CW lasers—come out of the production facility, typically we see transceiver shipments based on those devices only in the next quarter, i.e., 2 to 3 months later, right? For example, those transceivers shipped in our March quarter used indium phosphide devices manufactured in the early part of our September quarter or the early part of our December quarter. From device manufacturing to when we see them show up in transceiver shipments, there is generally a lag of a few months.

Sean O’Loughlin:

Jim, could you comment on the customer side? Should we assume that once transceivers are shipped, we have completed the qualification process? Is that the way we should be thinking about it—because—

Jim Anderson:

Oh, yes.

Sean O’Loughlin:

—everything is still ongoing?

Jim Anderson:

Yes. For qualification, there’s nothing particularly special between 6-inch and 3-inch devices. In some cases you may need qualification, but that should already be completed before mass production shipments, right? When we talk about mass production shipments, qualification is already complete at that time.

Sean O’Loughlin:

Got it. That’s helpful. And maybe related to the CW EML question—I know I didn’t miss it, I know you’ll say it’s somewhat irrelevant, and you’ll follow customer choices. If you could comment on the 400G silicon photonics technology you showcased at OFC, and also some industry doubts about the feasibility of 3.2T silicon photonics and CW lasers, that would be helpful. Thanks.

Jim Anderson:

Thanks, Sean. As you mentioned, at OFC we demonstrated a 400G silicon photonics technology that can support 3.2T. We demonstrated that capability. It can be used for transceivers or for CPO. We are just demonstrating the ability to do that. The exact form could be CPO or transceivers, or both. But we believe that based on this demonstration, we have a path to 3.2T—or 400G silicon photonics per lane. Of course, we expect to have solutions based on 400G differential EML—which we already have—and solutions based on 400G silicon photonics.

Jim Anderson:

By the way, we are developing 200G VCSEL and also developing 400G VCSEL. These will take longer, but we are working on these as well. We believe we have