On June 22nd, JP Morgan released a summary of Coherent's in-depth communication: The golden cycle of optical interconnection may have just begun I have shared with you the current pullback. If the market continues to give opportunities, I will prioritize investing in AI infrastructure targets that I have long been optimistic about, such as DRAM, MRVL, INTC, NOK, GLW, COHR, and AMKR. The reason is simple, I believe that the main thread of AI's current market trend has not been falsified. In the short term, the market may fluctuate due to interest rates, end of season rebalancing, or various macro noises, but what truly determines the stock price in the next one or two years is the industry trend. Among all the AI infrastructure tracks, I believe that optical interconnection will be one of the most important directions in the next two years. Today, I happened to read a Coherent (COHR) investor communication summary released by JP Morgan a few days ago, which revealed many noteworthy information. I will organize the core content and share it with everyone. Although this report did not provide any surprising target price, I believe the information revealed inside is very important. Many investors believe that the price of optical modules has risen significantly. But from the perspective of the industry cycle, it seems more like the eve of mass production. The first key point: the core data communication business The management said that the situation of insufficient supply in the optical module market has not changed, and there is a rather abnormal phenomenon: the time when 1.6T began to increase in volume is earlier than the time when 800G reached its peak shipment. This means that the pace of upgrades is significantly accelerating. Previously, it took several years to go from 100G to 400G, but now 800G is not fully mature, and 1.6T has already taken over ahead of schedule. More importantly, 1.6T sells more expensive and has higher gross profit, which means that the profitability of the entire optical communication industry chain is still continuously improving. The second key point: Will CPO, which everyone is most concerned about, replace traditional optical modules Coherent's answer is straightforward: no, at least not for the next few years. The management believes that standardized Scale Out networks will still prioritize the use of pluggable optical modules, as they are easier to maintain, lower in cost, and have a more mature ecosystem. CPO will grow in the future, but it will be more like an incremental market, not a replacement for existing stock. And the advantage of Coherent is not just selling individual products such as CW Laser or ELS. They cover almost all the core components required for the entire CPO optical engine - VCSEL, PIC, isolator, etc FAU、 Polarization maintaining fiber, detector, thermoelectric cooler, and microlens array. Many components can only be produced by a few companies worldwide. I paid special attention to the aspect of polarization maintaining fiber optic: the management said that the two companies that truly have the ability to supply this product globally are Coherent and Corning, and Coherent feels that their market share is higher. That is to say, regardless of which route wins, CPO or NPO, Coherent can benefit. The third key point: Indium Phosphide (InP), I think this part may be the most underestimated content in the entire summary The company once again confirms that the production capacity will double by the end of Q3 2026, and double again by 2027, with a total production capacity increase of four times in two years. Moreover, all of this expansion comes from the 6-inch production line, and more importantly, the yield rate of 6-inch has exceeded that of mature 3-inch processes. This actually indicates that the industry has entered a true stage of industrial expansion. Why is it important? Because whether it is 200G EML, 1.6T module, ELS or coherent optical module, they are essentially inseparable from a large number of InP devices. The biggest bottleneck in this industry in the past few years has not been insufficient demand, but the inability of supply to keep up. Coherent is now becoming one of the few players with the ability to truly scale up production. The fourth key point is OCS (Optical Circuit Switching), which I think is one of the most noteworthy new tracks in the coming years JP Morgan has previously raised its forecast for the OCS market size to $4 billion, but Coherent management believes that this number is still conservative. The reason is that the market originally thought OCS was mainly used in Scale Out networks, but now new demands are emerging: Scale Across (cross cluster interconnection), DCI (data center interconnection), and Scale Up (GPU cluster internal interconnection). Especially Scale Up, the management even believes that this may be the largest market in the future. If this judgment holds true, OCS may become one of the largest incremental markets in optical networks in the coming years. By the way, I have been tracking the trend of Scale Across recently. From Ciena, Nokia, Lumentum to Coherent, almost all optical communication companies have started to frequently mention this direction in their recent financial reports, indicating that the industry chain has formed a consensus on this matter. Fifth key point: Pump laser The management revealed that there are actually only two high-quality pump laser suppliers worldwide, and Coherent has a market share of up to 70% in some segmented markets. What is more noteworthy is that they do not plan to only sell components in the future, but rather upgrade to modules, subsystems, and even complete systems. The ASP of a single product has the opportunity to increase by more than 10 times. That's also why the market is increasingly favoring companies with complete vertical integration capabilities - the ones that will truly make money in the future may not necessarily be the ones selling individual parts, but the ones selling complete systems. Finally, let's talk about gross profit margin Coherent's long-term goal is to achieve a gross profit margin of over 42%. But based on the management's statement, I think 42% is more like a phased goal, not the endpoint. There are three reasons: the pricing power is increasing due to the tight supply and demand in the optical module industry; The cost reduction brought by the 6-inch InP production line; CPO、OCS、 Multi track systems, Thermadite and other high margin new products are gradually increasing in volume. Even the company has directly linked management bonuses to the increase in gross profit margin, indicating that the company's focus in the coming years has shifted from simply pursuing revenue growth to simultaneously increasing revenue and profitability. summary After the scale of AI clusters shifts from tens of thousands to hundreds of thousands or even millions of cards, the importance of optical communication will only continue to increase. The biggest advantage of Coherent is not a single popular product, but it covers almost the entire optical value chain - from InP materials to lasers, from optical modules to CPOs, from OCS to heat dissipation solutions. It is transforming from an optical device company into an optical infrastructure platform for the AI era. That's also why I've always felt that one of the biggest incremental opportunities in the AI industry chain in the next two years is likely to be in optical interconnection, and now it may just be starting.