Coherent Corp. (COHR) Earnings Call Transcript & Summary

My name is still Paul Silverstein. So I want to -- I'm Head of Investor Relations at Coherent, it's my pleasure to welcome all of you to Coherent's OFC investor and an analyst meeting. I want to thank all of you who are physically with us here in San Diego at the conference for joining us this morning at this event, which were once again highlighting our optical communications business, especially with respect to extraordinary and exciting market inflection presented by AI. Also want to welcome those of you who are joining our live webcast who may be tuning into this webcast and some future date via the archive link that will be in the Investor Relations section of our website. If you've not already done so, I also would encourage each of you to visit or revisit our previous investor communications market webinar that we hosted on September 19, 2023, which also is archived in the Investor Relations section of our website. We'll begin this morning's event with presentations by four executives from our leadership team. We'll have plenty of time following the presentations for Q&A from the audience. [Operator Instructions] Presenting this morning will be the following Coherent executives in the order in which they're presenting. Dr. Sanjai Parthasarathi, our Chief Marketing Officer; Dr. Julie Sheridan, our Chief Technology Officer; Dr. Lee Shu, Executive Vice President, Datacom Transceivers; and Dr. Beck Mason, Executive Vice President, Telecommunications. Following the conclusion of their presentations, they will be joined for the Q&A session by Dr. Chuck Mattera, our CEO and Chair; And Dr. Sunny Sun, President of our Networking segment. The presentations and Q&A session will be available for replay in the Investor Relations section of Coherent's website for 12 months. Please feel free to reach out to me following the event. I'd welcome any feedback that you may have. Feedback's welcome, positive feedback [indiscernible]. Before I turn it over to Sanjay, I need to address the customer disclosures, my apologies. I'd like to remind everyone on this call that this presentation contains forward-looking statements relating to future events and expectations, including our expectations of future financial, operational results and growth and the health, growth and opportunities in the markets we serve, that such statements are based on certain assumptions and contingencies and involve risks and uncertainties, which could cause actual results to differ materially from those expressed in the forward-looking statements in the presentation or in previous disclosures. Risk factors that could cause actual results to differ materially from those set forth in this presentation, are set forth in Slide 3 of the investor presentation, the company's annual report on Form 10-K for the fiscal year ended June 30, 2023, and as may be identified in filings of the company. Coherent disclaims any obligation to update any forward-looking statements, which, unless otherwise indicated in this presentation, speak only as of March 26, 2024. It is now my pleasure to turn the microphone over to Sanjai. Thank you.

Thanks, Paul. Good morning. We are so glad you're all here. We are eager to share our excitement about our communications market opportunity. Our positioning and our competitive differentiation in this market through both our current and future innovative technologies and products. I will start with the communications market update, followed by a review of new products and the exciting demos at our booth as well as our OFC speaking engagements. Let me first begin with our total available addressable market across all of our markets, our total addressable opportunity across all of our markets are for diversified markets are industrial, communications, electronics and instrumentation. We just updated our market sizing. We now estimate our current TAM across all of our markets to be almost $70 billion. And over the next 5 years, we expect this TAM to increase by almost 100% to $135 billion. We further segment our markets by defining verticals within each market. Our communications market, which is a focus of this event, includes our Datacom and Telecom verticals. Communications accounts for approximately 45% of our revenue, with the split within communications, approximately 70% Datacom and 30% Telecom. We provided deep insight in this market -- into this market on our September 19, 2023, Investor Market Webinar, which is archived on our website. If you have not already done so, I would encourage you to review the presentations. Our industrial market includes precision manufacturing, semiconductor capital equipment, display capital equipment and aerospace and defense. Industrial currently accounts for approximately 37% of our total revenue. Precision manufacturing, semi-cap and display capital equipment collectively account for the bulk of our industrial revenue, with each making a meaningful contribution. We provided in-depth insight into this market on our December 14, 2023, Investor Market Webinar, which is also archived on our website. We segment our instrumentation market, which accounted for 9% of our second quarter fiscal '24 revenue into life sciences and scientific instrumentation verticals. We plan to address this market in detail on our next investor market webinar, which is currently scheduled for May 14, 2024. Our electronics market, which accounted for 8% of our second quarter fiscal '24 revenue consists of our consumer electronics and automotive electronics verticals. Communications is our largest market, measured by SAM -- by TAM, by SAM and by revenue. Communications accounted for 46% of our revenue last quarter. We are a well-recognized innovation leader in the communications market and have led the market for many years in terms of market share. We enjoy leading positions in both datacom and telecom, and we leverage many of our key materials and compound semiconductor platforms across both these verticals. For example, Gallium Arsenide as a platform enables both terrestrial and submarine pumps and telecom as well as VCSELs [indiscernible] for datacom. Another example is our Indium Phosphide platform which enables EMLs, DFB Mach-Zehnder, Continuous Wave lasers for datacom, AI applications as well as enables our photonic integrated platforms for Coherent transceivers in telecom. Owing in part to the fact that about 2.6 billion people, 1/3 of the global population are still not connected to the Internet, the demand for bandwidth remains insatiable. The matavese powered by Immersion technology and IoT devices, enabled by AI generating data are two examples of applications that drive the demand for bandwidth. The applications of tomorrow are coming at us very quickly, and the bandwidth demand far outstrips capacity. We segment our communications market into datacom and telecom verticals. All the connectivity within the data center or data center intraconnects, we classify under datacom. All parts of the communications network outside of the data center, including submarine, long haul, metro and access networks, and data center interconnects, we classify under telecom. We estimate that the datacom and telecom markets combined represent for us a served available market opportunity of $17 billion in calendar year '24, which we estimate is enjoying a 17 -- going to enjoy a 17% compound annual growth rate over the next 5 years, increasing to $37 billion in calendar year '29. The beyond market growth, we continue to expand our communication SAM by entering new market opportunities, including Coherent transceivers, disaggregated systems and our recently announced optical circuit switch. Each of these three market opportunities offer the potential for attractive incremental growth along with attractive attending margins. We think we have extremely strong solutions to address and gain meaningful traction in each. I, along with Beck and Lee, will address each of these in greater detail later on this morning. With the extraordinary growth of AI clusters, we continue to see Internet content providers or ICPs ratchet up their capital investments in building out their optical network infrastructure. Their optical infrastructure investments not only just include transceivers for connecting servers and switches within the data center and along with OCS for data center interconnect or DCI, but also for metro, long haul, submarine and even satellite networks. Coherent offer solutions that addresses all of these market segments. We've been the market leader in datacom transceivers for 3 decades. We are vertically integrated and offer all the leading-edge technology platforms, including Gallium Arsenide and Indium Phosphide lasers as well as integrated platforms such as silicon photonics. We have climbed the experience curve. Julie and Lee will talk more about our technology and market leadership position in the industry. Turning to Datacom. AI has rapidly emerged as an explosive driver of Datacom Transceivers. Last week, we all witnessed a revolutionary announcement by the market leader in AI systems, which further reinforces our market outlook. Well, the basic architecture remains the same, namely, GPUs internally connected by electrical connections and externally connected by optical connections. The bandwidth of the new external optical connections is significantly increasing. The previous state-of-the-art system with four 800 gigabits per second optical transceiver ports is now being replaced by a system with 36 1.6 terabits per second optical transceiver ports, which represents a bandwidth growth of 18x per system. The arrival of generated AI in 2022 -- late 2022, drove an inflection point in the datacom market. We are one of the first -- we were one of the first to cease the impact of AI on the overall datacom market, and we've been constantly refining our view of that impact ever since. Our comprehensive market analysis utilizes inputs from our customer value chain, our interpretation of customer systems architectures to arrive with GPU to transceiver ratios as well as external market research. This chart shows the latest view with a focus on higher-speed transceivers, which drive the majority of the AI opportunity. While our commercial launch of the 1.60 transceivers is not far off, we expect our 800G transceiver revenue to grow for years to come, with overlapping growth in both 1.6 and 800G, consistent with previous transceiver adoption cycles in the datacom market. Moving to telecom. I want to return to our robust effort to expand our SAM. Transceivers with Coherent modulation technology are the largest part of that expansion. Industry analysts project the current transceiver market will grow at a 20% 5-year CAGR from $3.8 billion in calendar '23 to over $8.7 billion in calendar '28. We are entering this market with products that offer a compelling value proposition. We are leveraging our vertical integration with our industry-leading Indium Phosphide platform as well as our new DSP platform to offer uniquely differentiated products. In turn, customers are leveraging our innovative products to create new markets and applications. The number and the quality of our customer design wins for our Coherent transceivers, especially for our new 100G ZR QSFPDD transceiver, which leverages our internally developed Steelerton DSP, underlie our confidence in our ability to drive market penetration and accompanying strong incremental revenue growth. Beck will talk more about those innovations and outstanding customer reception for the same. A relatively new trend in telecom networks is disaggregation. Disaggregation is gaining increasing momentum. Disaggregation promises to help operators reduce cost and broaden their supply chain through standardization and multi-vendor interoperability. As a general proposition, service providers and data center operators want to avoid vendor lock-in and to be able to build telecom networks using products from a mix of vendors. The same way of app scalers built data center networks. With the broadest and the most vertically integrated portfolio in the industry, we are well positioned to address this market. Offering the potential to further meaningfully expand our datacom SAM yesterday, we announced our brand-new optical circuit switching, or OCS platform for data center networks. We estimate the OCS market, which is currently a multi-hundred million dollar market, to grow to over $1 billion over the next few years. Our differentiated offering, combined with our expertise and decades of experience with our liquid crystal WSS platform and our vertical integration strongly positions us for this opportunity. Let me spend the next few minutes reviewing what promises to be our best OFC ever. We are in both 3412 right by the main entrance. We kicked it off yesterday at what has always been OFC's headline event, the executive forum. As in the years past, extremely well attended. Chuck was at the CEO panel where he spoke about how we have executed on our decade-long strategy to cement our position as the largest optical component supplier with the broadest and the most vertically integrated portfolio in the industry. Yesterday evening, we received three awards from Lightwave Innovation, including two of our differentiated 800G components and transceivers. Our differentiation and innovation is critical to our current transceiver market position and revenue growth. Beck will cover our innovation that led to these awards later this morning. We announced 11 new product platforms, many of which we are demonstrating at our booth. Please do come to our booth and see these demos of our differentiated offerings, including our unique 100G ZR Coherent transceiver, featuring our in-house DSP for outside plant applications, our 800G DCO transceiver and our 800G DR4-LPO transceiver, featuring the first of its kind DFP Mach-Zehnder laser, which can address link up to 10 kilometers. Julie will discuss more about that innovation in greater detail. We innovate at all levels of the value chain from materials to optics, to components to systems. Our products at OFC include optical building blocks such as Grisms, Micro Lens Arrays all of these new products, BiDi Adapters and so on. We are excited about our 6 inch Indium Phosphide platform and our breakthrough on our path to the 200G VCSEL, which promises to be a game changer for AI connectivity. At 12:30 p.m. today, I will be moderating a session on component platforms for AI networks, which will feature our own Vipul Bhatt, VP Datacom Marketing, who's here. Joining us on this panel would be Cedric Lam, Google's Director of Infrastructure; and Craig Thompson, NVIDIA's Vice President of the Networking Group, among others. I'm sure it's going to be a lively session, especially after last week's announcements by NVIDIA and all the excitement it represents for the optical community. We continue to lead the industry in innovation and our extremely talented team our extremely talented technical team will be presenting at various technical sessions throughout the week. Thank you for your attention. And with that, I will turn it over to Julie.

Thanks, Sanjay. Hi, everybody. I'm Julie Sheridan Eng. I am the Chief Technology Officer of Coherent. And today, I'm going to talk about our technologies inside of our datacom products. which, as Sanjay mentioned, are contributing materially to the data center build-out of artificial intelligence or AI. So let me start first with a bit of background. AI requires training on data sets, which can contain billions or even trillions of parameters. This requires significant compute capacity, which can be provided by tens of thousands of processors simultaneously. To address these new requirements, data centers are adding new equipment specific to machine learning and AI. And in this AI portion of the network, a significant amount of information must be transferred between processing units such as GPUs and TPUs. And to enable this, the GPUs and TPUs are interconnected in a dense grid with electrical connections within the rack and fiber optic connections through transceivers for beyond the rack. In this way, the fiber optic transceivers are playing a critical role in the build-out of the AI network inside the data center. The speed of transceivers is critical to the performance of the network. Today, transceivers are shipping with 100G per lane components. Transceivers can use either 4 lanes or 8 lanes of 100G to create 400G or 800G transceivers, and we and others have also demonstrated 200G per lane components, which can support 800G with 4 lanes, 1.6T with 8 lanes and even up to 3.2T with 16 lanes of 200G. Network upgrades and architectural changes to address artificial intelligence and machine learning are driving us and the industry to introduce higher speed transceivers at a faster pace than ever before. This requires tremendous innovation in high-speed components and especially semiconductor lasers. Coherent is one of the only transceiver manufacturers that is materially participating in the AI build-out that also designs and manufactures lasers in-house. Our multi-decade strategic investments give us a unique level of vertical integration and differentiated capability. So why do we develop and manufacture lasers in-house? The primary reasons are: first, to provide value and extract returns at all level of the manufacturing chain. Second, this allows us to control our own supply chain for this critical component. Third, we have the ability to innovate and lead with technology-driven products. And finally, we're able to offer our customers differentiated capacity for their supply chain risk management, which is very valuable and makes us more valuable to our customers. So we developed several types of laser devices, why don't we use a single laser for all reaches? And the reason for that is we need to optimize for power consumption and cost. So we start with the lowest cost, lowest power-consuming device, and then we only transition to a higher cost higher power consuming device when it's required by the application, which is generally as the data rate increases, fiber length increases or the fiber loss distance increases. The lowest cost, most energy-efficient laser for datacom applications is the Gallium Arsenide vertical cavity surface emitting laser, or VCSEL. Many of the links in the AI network are shorter than 50 meters, and the Gallium Arsenide VCSEL is a great fit for that application. Beyond 50 meters, the signal from a VCSEL over multimode fiber begins to degrade so we need to move to other technologies, including Indium Phosphide and silicon photonics to support 500-meter links to 10-kilometer lengths, which can connect switches together or switches to routers or routers to routers. Why do we use Indium Phosphide for some applications in silicon photonics for others? Generally speaking, Indium Phosphide has higher optical output power and better modulator efficiency. So it makes it a best fit for the highest speed and most challenging applications. Silicon photonics can offer advantages by allowing the integration of compact passive optical components. So in links where silicon photonics can meet the specifications, it can be smaller sized and lower cost. But it's important to note that it's not an either/or with silicon photonics and Indium Phosphide, it's an and. And this is because silicon does not [ links ]. So every silicon photonics-based transceiver will contain an Indium Phosphide high-power laser to provide the light. Because we have all the technologies in-house, we're able to objectively compare and select the best technology for the application, which is a differentiator for Coherent. We're one of the pioneers in VCSELs and one of only two high-speed datacom VCSEL manufacturers. We're also a leader in VCSEL-based transceivers, which account for a significant portion of the AI market opportunity. Coherent has multiple 6-inch Gallium Arsenide VCSEL fabs in the U.S. and in Europe. We're the only datacom VCSEL manufacturer in the world with a 6-inch vertically integrated platform, which we believe provides significant cost and volume advantages. We're one of the highest volume manufacturers of VCSELs in the world. Last August, we announced that we've shipped over 200 billion VCSEL emitters for sensing and datacom. Our 100G per lane VCSELs are in production to support 400G and 800G transceivers, and are experiencing a significant high-volume production ramp primarily for AI applications. Beyond 100G per lane, we are nearing the physical limitations of directly modulated lasers and a 200G per lane VCSEL is a significant challenge. We were the first in the industry to announce that we think a 200G per lane VCSEL was possible last March at OFC 2023. Here at OFC 2024, we have a technical paper showing our results from a new VCSEL platform called the oxide-free lithographic VCSEL. We developed this new platform because we believe it has strong potential to achieve the bandwidth necessary to support 200G per lane. And if we're successful in our commercialization effort, we expect 200G per lane VCSEL-based transceivers to gain strong customer interest and adoption for short-reach AI links. Coherent is also a leader in indium phosphide lasers and transceivers. We have multiple indium phosphide fabs in the U.S. and in Europe. Our indium phosphide technology platform has been field proven at scale with more than 200 million datacom lasers deployed in the field over 2 decades. Our indium phosphide lasers have been deployed and qualified by virtually every network OEM and hyperscaler in the world. And we leverage that deep expertise to develop lasers to support longer reach 800G and 1.6T transceivers. For reaches 500 meters to 10 kilometers, electro-absorption modulated lasers are often used. We manufacture 100G per lane EMLs to support 400G and 800G transceivers, such as our EML-based 800G DR8 transceiver, which we've demonstrated publicly. We've also publicly demonstrated our 200G per lane EML, for which we received the Lightwave Innovation Review Award. As we and the industry transitioned to 200G per lane and 1.6T transceivers, we introduced our newest laser technology called the DFB-MZ, and this stands for Distributed Feedback Laser with Mach-Zehnder. This is an Indium Phosphide laser monolithically integrated with an Indium Phosphide Mach-Zehnder modulator, which is a very advanced photonic integrated circuit. We believe this new laser technology will be valuable for 1.6T, 10-kilometer reach and also linear pluggable optics, or LPO. We demonstrated an 800G FR4 OSFP based on the DFB-MZ at the European Conference on Optical Communications, or ECOC in 2023. And here at OFC 2024, Sanjay said, we're demonstrating an 800G DR4, which is 4x200G, LPO OSP based on the DFB-MZ. We won the ECOC Industry Most Innovative Product Award for the DFB-MZ, and this is an example of the competitive advantage provided by our internal component capability. We're also very excited to announce that we are bringing up 6-inch Indium Phosphide lasers in both our Jarfalla, Sweden, fabrication facility and in our high-volume state-of-the-art fabrication facility in Sherman, Texas. We expect to be the first to 6-inch indium phosphide, which gives us lower cost and also added capacity to support the growing needs for Indium Phosphide for AI. I brought a little demo for you on which I'll hold up here. So this is a 3-inch Indium Phosphide wafer of lasers. And here is 6-inch. So what you see is you get a lot more lasers per wafer, which will give us a lot of -- we believe that investment will position us as the vendor of choice in Indium Phosphide for Datacom and AI. We also have a strong position in silicon photonics-based transceivers. Our position in silicon photonics is based on over a decade of in-house silicon photonics investments and technology development starting in 2010. By 2014, we demonstrated the industry's first 50G NRZ end-to-end link based on internally designed silicon photonics. For many generations of products, we determine that Gallium Arsenide and Indium Phosphide offered meaningful advantages compared to silicon photonics-based solutions. Over time, as transceivers became multichannel, and as parallel single-mode fiber applications emerged, and as we entered the coherent optics transceiver market for telecom, we started shipping silicon photonics-based products in production. We have an in-house silicon photonics design team, and we rely on Tier 1 foundries as is common in the silicon industry. Like our integrated circuit design teams, we differentiate ourselves based on design, and we have many patents on our silicon photonics designs. We have sampled 800G transceivers based on silicon photonics to Tier 1 customers. And in addition, we showed our 200G per lane silicon photonics eyes at ECOC last year. We have developed and are manufacturing in-house Indium Phosphide, high-power lasers for silicon photonics. We're making these lasers commercially available in the market, and we'll leverage them for our own transceiver design, such as our silicon photonics-based 800G DR8 transceiver that we have demonstrated publicly. Transceivers are transitioning to higher speeds at a faster pace than ever before. We and others are shipping 800G transceivers based on 100G per lane lasers in production, and we expect to ship our first 1.6T transceivers based on 200G per lane lasers this year. We're one of the few vendors in this space that has the technology and the manufacturing scale to keep pace. But what happens after 200G and 1.6T? We believe that most of our customers will transition to 3.2T pluggable transceivers by adding additional lanes rather than by increasing the data rate per lane. So we believe that 200G per lane lasers and optics we're developing today will support the development of 3.2T pluggable transceivers, which should serve the industry for the next 5-plus years. But what's beyond that 200G per lane and 3.2T transceivers? As usual, the industry will come together through standards and multi-source agreements to establish a few options and arrive at a few competing solutions. It's likely that either the number of lanes will increase, the line rate will increase to 400G or more complex signal modulation will be used. And each of these pads is possible and each of them is challenging in their own way. There's been discussion for the past decade about coherent optics inside the data center. For clarity here, I'm referring to coherent optics with the small c rather than Coherent, the company. Well, coherent optics provides more capacity for channel. It requires more complex optics consumes more energy and is higher cost than direct detect or non-coherent solutions. So just as we use a waterfall of laser solutions, starting with the lowest cost, lowest power-consuming solution and transitioning to a higher cost, higher power consuming solution only when required, we think the same will be true of coherent optics. We see a path to 3.2T, 10-kilometer Direct Detect pluggable transceivers, which we think will serve the industry for the next 5-plus years. If at some point in the future, coherent optics is lower cost, lower power consumption and smaller size than the other options, or if coherent optics is the only way to achieve a certain distance at a certain data rate, then we and others in the industry will implement it. But if this happens, we will have an excellent position given we are one of the few major datacom suppliers that also makes coherent optics transceivers. Our 100G ZRQSFP28 DCO that Sanjay mentioned, that coherent receiver is an excellent example of how we are using our internal capabilities to decrease the cost, size and power consumption of coherent optics solutions. So while we don't see coherent optics in the data center as a near-term trend for datacom, if and when this does happen, we will be in a strong position to support it. There's also significant industry discussion about linear pluggable optics, or LPO, linear receive optics, or LRO, as well as near package optics, or NPO, and co-packaged optics, or CPO. It's important to note that these are different implementations can be thought of as packaging and architectural partitioning changes compared to traditional retimed optics. And all the underlying component technologies remain largely the same. So for this reason, our internally designed components can support LPO, LRO, NPO and CPO, and when these implementations emerge, we'll be ready to support them. Based on customer interest, we are developing LPO products. We have demonstrated an 800G DR8 LPO based on 100G per lane at ECOC last year, and here at OFC, we are showing our 800G DR4 LPO based on 200G per lane. We've actually been shipping VCSEL-based NPO for over a decade. In terms of CPO, we see the market evolving into optical chiplets, which can contain the modulator, detector, optics, optical packaging and integrated circuit, which is much like an LPO pluggable transceiver just in a much smaller package. For a very short reach CPO, we believe Gallium Arsenide VCSELs will continue to have a place in the ecosystem, owing to their low-cost, low-power consumption and ease of integration into parallel modules supporting many lanes. At last year's OFC, we demonstrated a VCSEL-based CPO that we developed with IBM through a contract with ARPA-E. Under that program, we delivered an 800G VCSEL-based CPO chiplet in a very small package of 13 millimeters x 13 millimeters x 4 millimeters. And we are currently extending that work to 3.2T. For longer reach co-packaging, Indium Phosphide high-power lasers will be used likely with silicon photonics. This implementation is generally shown as silicon photonics co-package with a switch or processor chip with a remote indium phosphide high-power laser connected to the co-package chip through optical fiber. We at Coherent are working on high-power Indium Phosphide lasers, silicon photonics and compact optical packaging. These can be combined either into pluggable transceivers or optical chiplets for co-packaging. So when the co-packaging market emerges, we will be ready. While MTO and CPO are hot topics, we believe we will see the industry innovate around pluggable optics paradigm as long as possible, as we've already seen with the LPO and LRO. We believe that pluggable transceivers will support the industry needs through the rest of this decade. While we do see co-packaging as a trend for the future, it's important to understand the value pluggable transceivers bring to our customers. First, there's a rich ecosystem of transceiver manufacturers that support these standardized interfaces, which creates competition and economies of scale. Second, pluggable transceivers can be added as needed, enabling the so-called pay-as-you-grow model. Third, pluggable transceivers offer flexibility, enabling our customers to use lower cost, lower power consuming links for shorter lengths and higher reach -- longer reach transceivers for longer lengths without committing to a configuration upfront. And finally, pluggable transceivers also offer an easy replacement as needed, offering flexibility to our customers. As a leader in transceiver for over 2 decades, we expect to continue to innovate to remain a technology leader and to deliver volume solutions to our customers to enable data center and AI networks that meet their needs, and we're excited to be part of the hardware solution that is helping to bring forth the power of artificial intelligence. Now let me turn it over to Lee.

Thank you, Julie. Good morning. My name is Lee Xu, and I want to thank you for the honor of presenting our Datacom business today. I'm excited to share our vision and strategy with you. Let's go to next. So our datacom business is a vital growth engine for Coherent. We expect substantial growth for 800G in fiscal 2024 and beyond. With our firm development commitment for 1.6T technologies, we see the potential to more than double our revenue -- telecom revenue in the next few years. I'm here eager to showcase our unique strength and competitive advantages. Today, my focus will be on the exceptional AI-driven growth over the next several years. It's really a singular opportunity for us. Leveraging the collective strengths of our company, we're poised to meet this challenge. Julie and Sanjay have already addressed various aspects of AI market, including applications and technologies. I will delve into our accelerating revenue growth in 800G, a key driver of our growth in the near term. And then our online -- our product and technological advancement aimed at enabling us to exceed the future AI market growth. We enjoy a very long history of market leadership as evidenced in this slide, which presents our Datacom revenue over the past 18 years, and also offers a glimpse into the future. Our journey from a leading the shift from 10 to 100G and then to 200 and 400G transceivers underscores our industry leadership. In fiscal 2023, our 200G and 400G sales constituted over half of our datacom revenue. really positioning us at the industry's forefront. Our consistent growth, especially since the acquisition of Finisar in September 2019 and has solidified our leadership in the transceiver market. With the emergence of AI, we expect an even more rapid expansion starting 2024. All right. So our portfolio really encompasses every key variance of 800G products that the market demands from short to long reach. We are aggressively ramping our 800G transceiver shipment and revenue. Having only commercially launched our 800G product transceiver approximately -- a little bit over a year ago, our 800G revenue sequentially grows -- increased by over 100% to over $100 million in our most recent December fiscal quarter. Following over 200% sequential growth in our first fiscal quarter, that's the September quarter last year. Looking ahead, we expect our 800G revenue to increase by roughly 3x to $450 million in the second half of this fiscal year, fiscal 2024, from approximately $150 million in the first half. We expect our 800G transceiver revenue to continue to grow long after the along after we commenced the shipment of 1.6T transceivers in the second half of calendar year 2024. Slide 46. Helping to fuel this expected long-term 800G revenue growth, this slide show our high-level product family road map. We have launched our mainstream 800G products based on cutting-edge 100G per lane optical and electrical technologies. As this slide indicates, we will be enhancing our 800G product portfolio with 800G transceivers based on 200G per lane lasers and other optical components, which also should reduce the cost of our bill of materials. As also indicated on the slide, we will have 800G products that specifically address the linear pluggable optics. Julie mentioned that early on. We expect to commercially launch both of these new 800G products later this calendar year. As Julie also reviewed, we do not see meaningful, if any, risk posed by LPO or CPO for that matter should either of these technologies gain eventual adoption from customers. As previously noted, we expect our 800G transceiver revenue to ramp for years to come, even as we commence shipping and ramping our 1.6T generation. Our ambitious goals do not stop here. We are hard at working 1.6T products, which we expect to start shipping by end of this calendar year or early next year. We also have already commenced exploring 3.2T and co-packaged technologies, as Julie indicated earlier. Slide 47 showcases the cornerstone technologies for our success. Julie addressed a lot of that. I want to say that our differentiated transceiver design and laser technologies are set to further enhance our industry leadership position in the 1.6T generation. Our diverse customer base, each with unique architectural designs and needs, demands a very versatile approaches and ranging from silicon photonics to EMLs to DFB-MZ, is essentially DFB Mach-Zehnder and even VCSELs, right? And that catering to a variety of application scenarios such as connectivity length and system architectures for this next 1.6T and 3.2T generation. Our internal design team is at the industry forefront, proving that silicon photonics can achieve 200G per lane. Our fab in Sweden produces EML lasers, which are pivotal to our 400 and 800G single mode transceiver shipment right now. With a head start in Sweden for the DFB Mach-Zehnder laser technology, known for its superior distance coverage and linearity, and the potential for even higher bandwidth, we're pushing the boundaries. And our 100G per lane VCSELs already on the market in mass and with 200G per lane versions in development. All right. So as set forth on Slide 48, we're very excited to introduce here at OFC 2024, a new product. We call it Optical Circle Switch. Julie -- Both Julie and Sanjay talk about this. We call it OCS, just to brief it, right, and enhancing our AI-driven market opportunity, this product offers the potential for very meaningful incremental revenue increase for our datacom vertical market. Based on industry analyst's estimates, OCS is currently already a multi-$100 million market that could grow to over $1 billion in a few years. We plan to gain a very healthy share of this market opportunity. In addition to confirming competitive performance advantage, our vertical integration of liquid crystal design and fabrication plus other advanced electronics and microoptics, all these should enable a very profitable product for us. Interest in this OCS for data center and AI applications really has increased dramatically in the past year after a technical paper published, showing a 30% cost saving and 41% power saving using OCS relative to electrical switch, right? In addition to the cost and the power savings, unlike traditional switches, the OCS is data rate agnostic, right? And which offers data center and AI customers the ability to upgrade to new data rates without changing hardware. And that is a significant advantage compared to electrical switching. While our OCS is a new platform, we think that we enjoy what we think a -- has a significant competitive advantage. These include the long experience we have with this key underlining material technologies, which we think offers performance advantage over other OCS platforms. More specifically, it's based on digital liquid crystal technology that we developed for our -- one of our wavelengths selective switch products. We have shipped over 160,000 of these WSS systems, including for ultra-high reliability undersea applications. And we have adopted this technology into our OCS product right now so the advantage is -- of this liquid crystal for particular this application are intrinsically high reliability and low-voltage electronics used. Altogether, they can result in high reliability and lower cost. In this OFC, we're showing 300 port by 300 port live demo of the OCS in our booth. And we're very excited to be able to extend our long history of innovation in liquid crystal technology to this exciting product space. And together with transceivers, we can support the AI in the data center, right? So in the very fiercely competitive datacom transceiver market, we've maintained our leadership for over 16 years, commanding a 20% plus market share with excellent financial performance. Based on our projected 800G revenue growth for third and fourth quarter of this fiscal year, we expect our share of the AI-driven 800G market to exceed 20%. Our success is rooted in our scale, the breadth and depth of our product portfolio and our broad customer base. Those provide us with significant leverage and economy of scale. Our progress in transceiver level design made us a leader in the 800G generation and will further enhance our leadership position in 1.6T, right? Our vertical integration in design and manufacturing really give us advantage in cost and ramp-up speed. Our vertical integration on lasers allow us early and stable access to advanced devices such as the current 100G per lane VCSELs. So our diversified manufacturing characterized by high-quality systems, advanced automation and efficient new product introduction underpins our rapid scaling capability. our production expansion in China and more expansion outside of China, in Malaysia, Vietnam and Philippines are steps toward a global resilient manufacturing footprint, right? The fusion of Finisar, the 3 decades of leadership was to [ fix ] assets and mindsets, especially in the era of AI underscores our technological and manufacturing excellence. This is a summary page. In summary, we see ongoing significant impact of AI on our Datacom business in terms of both magnitude and duration. We expect our commitment to innovation, particularly in 1.6T generation and laser advancement to fuel our continued and even greater success, that's to drive our AI revenue growth above the AI market growth, right? And as previously noted, we see the potential to more than double our datacom transceiver revenue in the next few years and for our launch of the optical circuit switch to generate significant incremental telecom revenue, we're excited for the future. Thank you for your attention. I'll now turn to Beck.

Have to clean up after the previous speakers so. Good morning, everyone. So my name is Beck Mason, and I'm the EVP of the telecommunications business at Coherent. So Coherent has been in the telecommunications business for more than 30 years. And over that time, has grown that business both through internal innovation and through the acquisition and integration of companies with critical enabling technologies. So our technology roots date back more than 30 years, and include foundational IP developed at [indiscernible] Labs, Corning, Nortel Optoelectronics, Alcatel Optronics, Pirelli, IBM and others. Our acquisitions include Marlow, which brought industry-leading solid-state thermoelectric cooler technology, PhoTop with its broad array of vertically integrated passive optical solutions, Faraday Rotators from Integrated Photonics, the Telecom pump laser, amplifier and signal monitoring business from Oclaro, wavelength selective switch platforms from both Koradna and Finisar and Ghana, and finally, Indium Phosphide lasers and optical transceivers and transmission components from Finisar. Telecom fiber optic networks cover a wide range of applications, from access networks at the edge that connect to homes and businesses through to metro and long-haul core networks, which interconnect the data centers and central offices across the country, and finally, with subsea networks that connect them around the world. They also play a critical role in 4G and 5G wireless networks where they support front haul and backhaul of the traffic. Coherent produces the products that transmit and receive the information over those networks, enabling hundreds of separate optical transmission channels on individual wavelengths of light to be carried over a single fiber with aggregate capacities into the several tens of terabits per second. We produce the optical line system components that are used to add, drop, route and switch these individual channels over the network. Our pump lasers and fiber amplifier products regenerate the signals and extend their reach over thousands of kilometers in long-haul applications, and our monitoring solutions enable the performance of these channels and the health of the fiber infrastructure to be managed effectively. Global Internet traffic continues to grow at a healthy pace, driven by video, wireless, web services, machine-to-machine communication and more recently, AI applications. Today, there are more than 5 billion Internet users and almost 30 billion network devices, while Internet traffic is growing at 24% per year. The telecom market experienced a slowdown in the second half of calendar 2023, and we are in the middle of that now. And we do not expect full market recovery until the second half of 2024. Despite the slowdown, we expect our SAM to grow at a healthy 16% CAGR over the next 5 years coming out of our fiscal '24. This is driven by the fact that we are well positioned in 2 very high-growth segments. The first is coherent transceivers, which is the largest segment in the telecom space, is projected to grow at a 20% compound annual growth rate over the next 5 years, reaching a market size of $8.7 billion by 2028, as Sanjay showed in his presentation. The second is the disaggregated systems market. We have differentiated leading-edge technology that supports these solutions. This market is just over $3 billion today annually and projected to exceed $4.5 billion by 2027. These 2 areas are going to be the largest growth drivers in our telecom business over the next 5 years. Coherent has 1 of the industry's broadest portfolios of technologies and products that power telecom fiber optic networks. This portfolio breadth should allow us to fully benefit from telecom market recovery. We serve all segments from the edge, where subscribers are connected through to the metro, and core networks that interconnect them to the data centers that make up the cloud. Our products benefit our customers by simplifying their user experience, enabling them to reduce both capital and operational expense while paving the way for the evolution of future networks and increased traffic that we expect them to carry. In all applications, customers value high performance, high reliability, smaller size, lower power consumption and lower costs. Because of the realities of network operations and economics, we see a different emphasis on each of these attributes depending on which section of the network is being served. For example, in access networks, cost, participation and physical size play a significant role in core networks, spanning hundreds of thousands of kilometers, performance is key. In submarine and space networks, high reliability becomes a factor of primary importance. Optical line systems comprise the base layer of fiber optic networks. We are a leader in the space, and our products cover all aspects of amplification, routing, switching and monitoring in the network. Our amplifier technology is based on our industry-leading pump laser solutions and delivers best-in-class capability for receiver pre-amplification, transmitted booster applications and low-noise line amplification. We cover both C and L-band network requirements with both Erbium and Raman capability. For routing and combining DWDM channels, we have fixed multiplexer and demultiplexer solutions, along with reconfigurable optical add-drop multiplexes based on our industry-leading LCOS WSS technology. In submarine applications, our ultra low power, high reliability liquid crystal-based WSS solutions are deployed in situations where reliability and low power dissipation are critical. To help support these capabilities, we have the industry's most advanced solutions for monitoring and control. Our optical channel monitors can track hundreds of individual wavelength channels within the network to ensure their performance. Our most advanced optical channel monitor technology is based on coherent mixing, and is unique in the marketplace due to its ultra-high resolution capability, enabling it to deliver features beyond the capabilities of traditional OCMs. Finally, in order to monitor and ensure the integrity of the fiber infrastructure, we have developed optical time domain reflectometry tools that can scan the fiber real time, identify locations of brakes or excessive loss without disrupting live traffic. Increasingly, we combine many of these features together into integrated subsystems to support full node functionality for our customers. So we continue to advance our technology in a number of critical areas. We expect these advances to confirm meaningful competitive advantages, which in turn should lead to market share gains that will drive our telecom revenue growth over and above market recovery. In the wavelength selective switch space, we are scaling to opt solutions that enable both C and L-band channels to be managed within a single device. This more than doubles the bandwidth capability of our products versus our competitor solutions. On the amplifier side, we are leveraging our higher pump power capability to deliver more efficient amplifiers and to broaden their spectrum coverage so that more channels can be supported on a single fiber, moving from 4.8 terahertz solutions today to 5 and 6 terahertz solutions in the future in both the C and the L-band. This is increasing the total information carrying capacity of the fiber, and it lowers the cost of the network. We have new optical channel monitoring technology that enables a single device to cover both C and L-bands together, cutting the cost in half for monitoring these broader networks. And finally, on the pump laser front, where we continue to lead the industry, the highest power per pump emitter, we are now transitioning to designs where multiple pump lasers are integrated into a single package, yielding much more compact and efficient amplifier designs. Pump lasers are used to power the fiber amplifiers that boost the optical signals in virtually all telecom networks worldwide. We have been in the pump laser business for 30 years and have shipped more than 7 million devices to the field. Our output has grown more than 40% since 2018, and we are arguably the largest supplier in the industry. Our 21 different laser platforms serve everything from ultra small single-channel booster applications for silicon photonic-based optical transceivers to very high power, greater than 1 watt single emitter and dual emitter solutions that support high-performance amplifiers using C-band and L-band systems that can carry more than 100 wavelength channels. We have significantly enhanced our pump laser product portfolio, and we believe we have an industry-leading portfolio, we expect to deliver steady revenue growth at strong gross margins. We continue to innovate with new and differentiated solutions. One of our most interesting products in the transport space is the pluggable optical line system, which addresses the demand from network operators in connection with the emergence of IP over DWDM. Our POS solution enables a full bidirectional amplifier and an 8-channel colorless combiner to be plugged into a transceiver report on a switch or router. With the introduction of IP over DWDM approaches, network operators are using coherent transceivers plugged directly into their switch and router ports. This has created demand for a compact efficient solution to combine and amplify these signals. Our polls product delivers the solution. It enables the functionality of a compact transport OLS chassis to be delivered in a pluggable module. The solution can serve the dual function of multiplexing and demultiplexing the optical wavelengths to and from the various pluggable transceivers in the router, then amplifying them for transmission over the network. We see the opportunity to generate significant incremental revenue from expansion of our solutions in the coherent optical transceiver market. Coherent optical transceivers represent the largest and fastest-growing portion of our telecom market SAM at $3.8 billion today with a CAGR of 20%. We are expanding our solutions in this market with meaningfully differentiated products that we believe position us well for success. Notwithstanding the strong existing suppliers and solutions in this key product segment, our early design wins, both in number and quality for our new products reinforce our view. We especially note our new 100-gig ZR coherent transceiver that incorporates our Steelerton coherent DSP. This transceiver enjoys the advantage of extremely low power consumption and a QSFP28 form factor, which is the most compact form factor in the industry. This, in turn, allows for significantly greater faceplate density and bandwidth capacity relative to the prevailing competing CFP2 form factor. Traditionally, telecom networks relied on complex and expensive line cards to generate and detect the optical wavelengths to carry data over the network. Technology developments have enabled this capability to be complemented now in small, low-power pluggable optical transceivers and increasingly, these are becoming the primary solution for data center interconnects, metro and edge applications. In Edge, Access Networks, or compact 10G tunable SFP+ transceivers and our 100-gig QSFP28 coherent transceivers provide scalable, low power and cost-effective solutions that can be deployed in central offices and in remote terminal applications. For data center interconnect and metro applications, our high-performance coherent transceivers enable high-capacity networks that can operate over distances of a couple of thousand kilometers or carry data rates as high as 800 gigabits per second on a single wavelength. And they deliver aggregate capacities up to 64 terabits on a single fiber if you use both extended C and extended L-bands. So our strength overall is in the depth of the expertise we have in the critical enabling technologies for these transceivers from the indium phosphide and silicon photonic devices that generate, modulate and detect the optical signals, to the high-speed RF analog and digital ICs that are used to drive them, we support these fundamental building blocks with advanced packaging capabilities for integrating the photonics and electronic circuits for coupling the light from the photonic components into the optical fiber and for controlling the environment of the critical components. Our family of telecom digital coherent optical transceivers include solutions for QSFP28 form factor at 100 gigabits per second, for metro and Edge applications 400-gig QSFPDD solutions for ZR and ZR+, data center interconnect applications and high-performance 800-gig ZR and ZR+ solutions in QSFPDD and OSFP form factors. Every generation increases capacity and reach over the prior generation. We are among the first to market with 800 gig coherent in a small form factor transceiver, and we will be demonstrating this product here at the OFC show. This platform can support distances of over 1,000 kilometers at 800G and greater than 2,000 kilometers when operated in its 400-gig mode. In addition to scaling up on capacity and reach with our transceiver products, we are also driving down the size and power dissipation of the solutions. Over 3 generations of solutions for 100G coherent transmission, we have reduced the power dissipation by 70% and the size by roughly a factor of 4. This has been accomplished with our innovative low-power steelerton digital signal processor chip and our compact tunable laser and silicon photonic COSA components. An important building block in a coherent transceiver is the digital signal processor. This device converts digital data from a switch or router into the complex analog modulation signals that are used to transmit data over a fiber optic network. It also converts the received signal at the other end of the link back into digital data, and compensates for any signal impairments that occur over the fiber optic link. It's an important element in driving the overall performance of Coherent link, and it represents a significant portion of the cost of a coherent transceiver. We partner with leading vendors in the industry to leverage their DSP solutions for some of our products, and we design and develop our own custom DSP solutions with our internal design team. This enables us to cover a wide range of solutions. We focus our designs on advanced process nodes, which deliver the lowest power, smallest die size and ultimately the lowest cost solution. This combination of external partnership and internal capability lets us focus our engineering efforts where we can best differentiate with unique approaches that extend our market and enhance our ability to deliver disruptive solutions. We're building on our launch of our steelerton first coherent DSP with a number of additional coherent DSPs in development. Our internal DSP solutions not only give us a significant performance advantage, they also substantially reduce the cost of our products delivering a 10- to 15-point improvement in the gross margin of our transceiver portfolio. Our first generation DSP stellate was designed to dramatically lower the cost and power for 100G transceiver solutions, delivering close to a 60% reduction in power over prior generation devices and enabling 100G solution in a QSFP28 form factor, something that had never been done before, and still has not been done by anyone else yet. Our second-generation DSP Silverton is still in development. This device will be almost 20x the size and complexity of steelerton, and will use advanced 3-nanometer CMOS technology. This will deliver superior performance and power for 400G and 800G applications. Moving to 3 nanometers will give a significant power and cost advantage over legacy 7- and 5-nanometer devices, which are used today. This allows us to enable greater capability in the device within the pluggable form factor. This week at OFC, we will be demonstrating the industrial temperature version of our new 100-gig ZR QSFP28 DCO transceiver. It's the only coherent transceiver available today in the popular QSFP28 form factor. And now it's enabled by our compact tunable laser, advanced integrated silicon photonics coherent transmitter receiver and our steelerton DSP. It delivers breakthrough performance in a compact package, making it very popular among service providers and cable MSOs. This product is cable of operating over 125C temperature range from minus 40C to plus 85C, which enables it to be deployed in remote terminal applications at curbside or on the pole, has a capacity of up to 100 gigabits per second and a maximum reach of 300 kilometers. Increasingly, network operators are moving to IP over DWM strategies, particularly in the data center interconnect and metro DWM space. This approach enables coherent optical transceivers to be plugged directly into network routers and switches, bypassing the need for a separate trans monitor or mux monitor solution. IP over DWDM can enable substantial cost savings for network operators as it eliminates an entire layer of hardware and facilitates disaggregating the supply chain for coherent transceiver optics. This, in turn, significantly expands our market opportunities for coherent transceiver products. Coherent is a world leader in compound semiconductor wafer fabrication technology. For our telecom products, we have indium phosphide and gallium arsenide wafer fabs in Zurich, Switzerland, and Stockholm, Sweden, respectively, that provide the photonic chips used in our telecom and datacom products. In addition to Zurich and Jarfalla, we have state-of-the-art production fabs for gallium arsenide and indium phosphide photonic devices in Fremont, California, and Sherman, Texas. This week, we announced the availability of 6 inch indium phosphide fabrication in our Jarfalla, Sweden, and Sherman, Texas, wafer fabs. We are the first in the industry to transition to this larger wafer size for photonic chips, and we will show an example of a 6-inch device in our booth at OFC. Moving to 6-inch enables us to access newer generation automated equipment in our fabs, increasing yield and reducing processing costs. Our 6-inch wafers deliver 4x the die per wafer of current generation 3-inch indium phosphide wafers and can deliver a 60% reduction in die cost. At coherent, our strength is not just in our R&D and our technology but also in our manufacturing capability. We have vertically integrated manufacturing that spans all the way from the base device and component level to the module and subsystem level. Our production focus is on having geographic diversity of supply to ensure business continuity in the face of potential supply disruptions and to provide flexibility in country of origin to meet tariff and security-based sourcing concerns for our customers. Our internal photonic packaging production facilities rely on highly automated robotic assembly to ensure consistent quality and performance of our products and to enable high-volume production with reduced labor and overhead costs. Our telecommunications pump lasers are built in Shenzhen, China, and the Philippines, our photonic components are manufactured in our automated facilities in Wuxi, China. Our transceiver products are built in Wuxi, China, and in our factory in Epo, Malaysia, and we manufacture passive optics along with amplifiers, line cards and other subsystem products in Fuzhou, China. We also have manufacturing locations in Vietnam, Thailand and the Philippines. We have an experienced manufacturing technology development team that is expert in assembly and test automation. Their capability enables us to automate most of the manufacturing stages of our production. The use of internally developed automation helps us minimize our capital expense and customize our solutions to best fit the requirements of our products. It also ensures more consistent product quality, less waste and better manufacturing efficiency and cost. We pride ourselves on having 1 of the best quality reputations in the industry. We have a strong focus on continuous improvement in product quality, and we drive a zero-defect culture approach. Finally, Coherent is 1 of the largest suppliers in the telecommunications component market, and we leverage that scale to keep ahead in product design, manufacturing technology and product quality so we can bring superior value and quality to our customers. The breadth of our technology portfolio in both transport and transmission applications, allows us to bring to market vertically integrated solutions that are higher in performance, more cost-effective and better optimized to work together in telecom networks. Thank you.

Thanks, Beck. I want to thank Sanjay, Julie, Lee and Back. I thought those were great presentations. They, along with Chuck and Sunny, will now field your questions. We'll try -- I know it's late, we're going to try to answer all of your questions. We'll hold -- we were scheduling at I see you out there. We will -- we'll try to answer all of your questions. We're going to keep it up beyond 12. Please, once again, for those of you listening in via the webcast who would like to ask a question, please feel free to send your questions to me at [email protected] and I'll ask on your behalf. And with that, we'll open up to questions.

A lot of great opportunities, particularly in data center. I think you guys have been talking about ramping capacity for data center I'd like kind of an update on those plans. But just more broadly, talking about, call it, VCSELs versus EMLs, what capacity need do you guys most have out there right now to address this? Yes. And just talk about just kind of where you guys are kind of behind the curve? Or are you getting -- your competitor today talked about talked about capacity reservation fees. Are you going to have something similar? Would love to know just kind of the state of the ramps in the various opportunities you guys have in terms of capacity.

Lee, why don't you take it?

Yes. Well, thanks for the question. First of all, we're ramping up both on the VCSEL-based transceivers and the EML-based transceivers and some of the silicon photonics based to transceivers. So all of that are in ramping -- and we did say in the shell saying that, hey, particularly on 800G, where roughly $15 million shipment in Q1, $100 million in Q2, and Q3 and Q4 combined is going to be $450 million. So as you can see, that's quite the ramp, right? And we see this demand is quite broad-based and we hope that we can continue for many years to come.

Chris, are you good?

Yes.

Go ahead, Carl.

Yes Karl Ackerman from BNP Pariba. How do you see the adoption of optical circuit switches across hyperscalers who already deployed third-party switches and the routers I guess would optical switches be used in greenfield data center deployments at a certain insertion points, which is 1.6. How do I think about the adoption of that? SP-8 Okay. Sunny, you want to.

Yes. We have been working on that key customers for making key components for optical circuit switch for quite a few years, it's 7 or 8 years ago. So actually, we have supplied large volumes of the OCS optical circuit switch components to the customers who make it their own optical circuit switch. So Neste deployment has been implemented in certain AI machine learning customers. So right now, just more about broadly covered and now is that adopt for more customers with the AI requirements. We could see this invitations with increased capacity, especially for AI network requirements for switching noting requirements, this will be broadly covered by more customers. We have seen more engagements with additional hype scale and AI network operators Okay.

And that could add, if it's okay, sunny, is that, yes, it's a significant network change. And so I think you'll see it mainly in greenfield. But as Sunny said, because of the papers that were published showing the dramatic cost savings and the dramatic power consumption savings a lot of interest, yes.

So Ananda Baruah, Loop Capital. Thanks for doing this. question is you had mentioned share as part of the presentation. The question is, is innovation cycles continue to compress Gen AI-related innovation cycles? Do you think that has any impact on share implications in transceivers for folks who can innovate who have the ability to innovate more quickly? And what might the share implications be? And I'd love to then also just to the extent you have an opinion on that, how -- what kind of conviction do you have in that opinion?

So that's a very critical question for us. We've been thinking about that. So we do see this for AI, particularly the product introduction cycle does compress relative to the traditional hyperscaler data center generation not change. And that actually plays into our strength in terms of our intimacy with top kind of leaders on the AI cluster design and manufacturing. And we have a chance to use our cutting-edge technologies, right? Just to give you an example, right now, everybody knows there is 100G per channel VCSEL shortage, right, since we have our own internal VCSELs and the transceiver team and the VCSEL team working together, we can probably get more upward than our competitors. So combining all that and also with our strong commitment with the next 2 generations for product development, we think that will end up gaining more share in this AI race relative to previous hyperscaler general market.

And a follow-up to that is, if you look out 3 to 5 years, do you think that the share composition of, let's call them, sweet spot, sweet spot bandwidth speed looks meaningfully different in 3 to 5 years than it does today or it does -- or it has historically?

You want to take on that?

I think the 1 thing I can say is, as you alluded to, and, it's hard, this stuff is hard -- and is very hard, and it's ramping very fast, and the generations are coming faster and faster just as we're approaching like device physical limitation. So it's very hard that. And so because of that, for instance, now at 800G, 1.6T, you see very few really strong competitors compared to like 10G or even 100G. And I think that should benefit us, and I don't see it slowing down, Yes.

Just talking a little bit about the GTC announcements on the Interact interconnect side, now that you can load up the -- upto 10-K GPU cluster on a 2-layer network. So like what do you guys see as the content impact for the transceivers for you guys? And maybe do the ASPs increase with the higher bandwidth ports just make up for that? And also on your projections to that have an impact?

For anybody asking a question, just for the benefit of those out on the web, if you could introduce yourself before the question, that would be appreciated.

Yes, I'll try that. So the generational change that was announced last week, you're talking about from going from 4 optical ports of 800G capacity to 36 optical ports of 1.6 terabits per second capacity. So 18-fold increase in terms of the opportunity for optical transceivers for a fully loaded system. So it is a big -- and in line with what we have been projecting. And you saw there was a big inflection in 2023 of the market, and that inflection is all about AI.

Does that answer your question? Thank you. Questions on the line? Anybody else?

Yes. Dave Kang, B. Riley. Just a follow-up question on 800-gig ramp. So if you do the math, you said you're going to do about $450 million in second half. If you just go sideways, that's about $900 million next year. Just wondering how much of a further growth from the current half that we're in?

In terms of 800 shipment, we grow tens of -- more than tenfold this fiscal year versus last year. We're certainly not going to grow that much next year, but we still want to see at least our current projections still see a healthy growth. And so -- and once we reach that, that's -- yes, it's going to be more than half -- 800G alone is going to be more than half of our revenue, close to our half of our revenue.

This is [ Papa Sila ] from Citi. I have maybe a couple of questions on the DSP front. On the telco, I know you started doing some work there. I guess what's the target on being internal versus using merchant DSPs? And is -- are therefore you're making on the telco transferable to datacom? Or is that even the goal? And I have a follow-up.

Okay. Let me try the first part of that. So obviously, DSPs are a very large investment. When you go and develop a 3-nanometer DSP, just to put it in context, the mask set is about $25 million. So it gives you an idea of the scope and scale of those developments. So in our market, if we think about it, telecom moves a little slower in the transceivers than in datacom. And so we have 100 gig, 400 gig and 800 gig as the primary operating modes today. Even at 800 gig, the DSPs support multiple different rates. So for example, our Silverton DSP will support 400 gig for a long reach. It will support legacy 4 gig ZR modes, and it will support 800-gig high performance. So I think our goal is primarily over time to migrate the bulk of our portfolio to internal designs. But we're having partnerships with leading external suppliers helps us is, in some cases, we can't invest in every area to be first to market or some markets are smaller and more niche or less proven. And so, in that case, those external partners are very helpful. On your second part of your question, from a technology perspective, I would say that the front-end high-speed ADC DAC technology and the back-end SerDes technology on a Coherent DSP are very similar to what you use on a PAM4 DSP. The internal functionality, though, is very, very different. And so there is some synergy with that and some less so. But I think right now, we're not really commenting on any plans to vertically integrate around Datacom DSP technology.

Yes. No, that's very helpful. And a follow-up on the pricing just general comment on pricing on 1.6T on the transfer side. I guess, there, we heard a merchant -- I guess, the leading merchant vendor kind of will probably raise prices almost at 2x, but obviously, you cannot go all the way to 2x because of, I guess, TCO. But I guess on the transceiver side, how much leverage do you have between the DSP vendor kind of almost being 2x and then kind of having a 2x limit on how much you are able to raise pricing on the overall device level? I don't know if the question makes sense.

That's certainly a critical question for our profitability. I can't comment on detailed pricing. What I can comment is that traditionally, transceiver is indeed a very combative market, right? And we make money on several occasions. When is we're ramping up faster initially than other people when we can command a very healthy margin. The second is that we have very unique product or unique technologies that we can command a very healthy margin. For this 1.6T, just as Julie earlier on mentioned, it's pushing limit on multiple fronts, right? Electrical, optical, your own -- your own transceiver design and the thermal and optics. And we believe that with our technology breadth and depth, we'll be able to gain a healthy margin on that.

Karan Juvekar from Morgan Stanley here. So clearly, you're seeing a large revenue ramp on the second half for 800 gig going from 150 to 450. I guess how do you think this will differ on the 1.6 side? Do you think you see a similar sort of revenue step function as you go into 1.6T? And then given it's on 200 gig per lane, do you think that impacts your production? Or do you think you'll be able to sort of ramp your production on 1.6T similar to the way you were able to on 800 gig?

Yes. We're confident that we're going to kind of perform better in terms of market share and gaining market share and profitability of 1.6T than 800G. But the exact time for commencing the shipment of 1.6T is later part of this year or early next year. That's still -- we're just as our competitor and our customers, we're mentally kind of fiercely working on it. So...

And just if I could add let some amount of the -- for instance, if you're making an EML at 100G or 200G, it's still an EML running through your fab. So a lot of the capacity that we have is fungible kind of to other data rates that maybe singular exception is test equipment, right? Test equipment is data rate specific, but everything else is pretty much data rate unspecific. Yes.

Lee, we received a question from the web. I think I know the answer, but the question is when you referenced 800G being half of our revenue after the strong 25% growth, what were you referring to? And I trust you were referring to, as we said back in the shareholder letter that 800G this year, we think will be half of -- or AI will be half of our total revenue for the year. But was there something else you're referring to on the 800G growth in and of itself?

We're referring to just as what we referenced in our investor letters. It's essentially with 800G further growth. And with the AI, we also ship 400G for AI, and all that combined is half of our revenue for fiscal 2024.

Anybody else before we wind up? Please use the mic.

[ Jeff Kochi ] with Raymond James. In your updated datacom transceiver opportunity by laser type. You just have EML at like 68% of the opportunity of the $15 billion you updated that last quarter, and that was a material increase. So it seems like you're suggesting EMLs are going to be the major beneficiary. I just want to know how that squares with like the linear drive technologies and what your thoughts are there? Because I believe EMLs are not compatible with.

Good question. When we presented that in -- as part of our earnings call package in February, the assumption was that 200G VCSELs will not make it into 1.6. That was the assumption. That's why the EML piece went very large. But we know that this week, 2 vendors, including 1 of them being us, have talked about a path to 200G VCSELs. So this market keeps changing. The architecture is evolving, the laser types are changing. So we will be refreshing that. And I do anticipate that the VCSEL portion will grow larger than what it is currently and the EMLs will shrink a little bit because if -- as Julie often says, if you're able to make a transceiver of the XL -- that will be the lowest cost, the lowest power consumption. You will do it that way. Everything else comes later.

Just repeat the question. The thought being on linear drive.

Yes. For the linear drive, we do think there are customers that like that interface. It's generally where it's a bookended application where the customer owns both ends of the application because it is more of an engineered link. You get certain power savings and cost savings, but you pay price to get that power savings and cost savings by having to really engineer the link. And so our estimate is more like subset of the volume will be linear. And for that, we are looking at the application of VCSELs for that, but also for us, for the long way, we're looking at our DBM for that application. That's a very linear device. But I think it will be a subset of the total pluggable market. Yes. Yes.

Anybody else? Going, going. All right. Before we wrap up, I just -- I want to thank again the team, Julie, Sanjay, Beck and Lee and Sunny and Chuck. And on a personal note, I want to thank on behalf of myself and the team, Chuck for his exceptional leadership. You really -- I've been here 9 months, and I've just been blown away. I really mean that. We look forward to continuing our discussion with many of you. For those of you who I know, it's always great to see. For those of you I don't, please feel free to come up and introduce yourself. And for all of you, those of you who I have known, those of you who we meet, I'd be more than happy to try to be more help in your investment process, I truly mean that. And I wish you a great rest of the conference. Thank you, everybody.