AI bandwidth demand still favors Coherent and Lumentum, but tight customer financing can delay orders and lasting valuations need durable profits. A model provider losing share need not reduce total bandwidth use. Coherent could beat and raise guidance; Lumentum may only meet estimates this quarter, improving next quarter. If margins partly reflect scarce indium phosphide, added supply will test whether upgrades, volume and cost gains sustain earnings. These earnings calls remain forecasts. … Show more

AXT cannot be the dominant indium phosphide substrate supplier when its customers' supply depends on the Chinese government's export permitting. Lumentum has locked up InP supply for several years with Sumitomo, a larger competitor that does not face the same export restrictions. Doubling down on manufacturing in China is an odd choice given continuous permitting roadblocks. Substrate pricing is unlikely to go parabolic, despite Serenity's claims of 10x+ price increases. x.com  x.com 

AXT's indium phosphide substrate chokepoint gives it pricing potential, though its Coherent agreement caps annual wafer-price increases. Channel checks suggest increases of 50-60% in Q2 2026, 10% in Q3 and over 10% in Q4. With supply agreements to Lumentum and Coherent, AXT probably becomes No. 1 ahead of Sumitomo Electric. A SanDisk-style hike now looks less likely because of the annual cap, whose percentage is redacted. x.com 

Astera Labs looks relatively expensive because similarly fast-growing AI infrastructure names trade at much lower multiples. Broadcom and Nvidia are at 14x, Credo at 15x and Lumentum at 18x; comparable peers look 50-80% cheaper. Astera is a great company and not necessarily expensive in absolute terms. Its roughly 2x multiple over Credo may persist, but the setup looks harder; with no valuation support, it could fall 50% if weakness continues. x.com  x.com  x.com  x.com 

Lumentum and Coherent sit at the center of a multiplicative optical bottleneck, where parts, capacity, yield and customer sign-off each limit output. Next-gen AI clusters need optical interconnect, and McKinsey sees 800G transceiver output 40-60% short of demand through 2027 as the market doubles to $26B in 2026. Nvidia invested $2B each in both to lock in laser supply, and the FCC is drafting curbs on Chinese transceiver imports; Chinese suppliers hold ~60% of optical-component revenue. x.com  … Show more

AI bottleneck semiconductor stocks look to have gotten ahead of themselves, with roughly a hundred near-vertical charts where downside now seems to outweigh upside. They are semis and highly cyclical, and many smaller caps may end up trading like meme stocks. A substantial shakeout would offer a better entry. AI, robotics and energy still have years of growth ahead, but chasing these names now is not advisable. x.com  x.com  x.com  x.com  x.com 

Lumentum and Coherent look like early winners as a second round of the tech and AI trade begins. In the first round every tech stock seemed to win regardless of what it did; now the field is narrowing, so investors need to be discerning. Nvidia CEO Jensen Huang said the silicon photonics capacity needed is substantially higher than the world has today, explaining Nvidia's $4B investment in the two. x.com  x.com 

Lumentum, Coherent and Applied Optoelectronics remain buys amid AI optics demand: Lumentum's CEO said its AI optical parts are completely sold out through early 2029. x.com 

AXT stays central to the InP substrate thesis because Coherent's 6-inch ramp relieves only its own device capacity and pushes the constraint upstream to substrates. Coherent does not plan to sell InP lasers externally soon, so the merchant shortage persists, which is also the context for Lumentum's long-term agreement with AXT. AXT's InP is made in Beijing, and China's InP export licensing remains despite the topic being raised in the Trump-Xi talks, so permission is the key variable. x.com  … Show more

Interconnect, not compute, is now the binding constraint on AI clusters. Copper, pluggable optics and CPO will coexist for years, so the strongest names span both paths or own the tightest bottleneck. Top five: Credo, the fastest-growing full stack; Applied Optoelectronics, the 800G/1.6T ramp; Marvell, broadest but richly priced; Lumentum, the laser chokepoint; Tower, the foundry bet. Coherent, Amphenol, Fabrinet, GlobalFoundries, TSMC and Broadcom also gain. Astera Labs is pricier and narrower. … Show more