Muhammad Zuhair
Data-Center Researcher at SemiAnalysis
Interconnect and memory-bandwidth bottlenecks could limit AI scaling, which makes silicon photonics and better networking significantly more important and should make Lumentum and Coherent much more prominent. Google's From AGI to ASI paper warns that as models grow, time spent moving data between memory and compute, or between nodes, can become the dominant factor, causing diminishing returns unless novel hardware or interconnect technologies are developed. x.com
Lumentum and Coherent stand to gain because the market moves on early co-packaged optics signals, such as NVIDIA's $2 billion investments in both, even if real benefits wait until Rubin Ultra or Feynman. The supply chain positions itself months or years before integration happens. Per @FundaAI, relayed by @firstadopter, Jensen Huang said at Computex that CPO will ramp on schedule, and scale-out deployments for CoreWeave, Lambda, Oracle and a major CSP are on track for mass production. x.com
Tighter Lumentum laser allocations would strengthen Coherent's pricing power if Marvell's rumored acquisition leads it to prioritize Nvidia and its own modules. Other buyers would face tighter supply, and hyperscalers would push harder to second-source lasers from Coherent. x.com
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
Co-packaged optics standards could make finished transceiver modules smaller and more interchangeable as chip designers build optical engines inside their packages. That is another reason to avoid Lumentum, Coherent and Applied Optoelectronics and prefer Soitec. Lumentum and Coherent now earn a lot from finished modules, leaving them exposed if the architecture changes. x.com
Optics and photonics should be among AI's biggest winners as faster GPU networks need optical connections that suppliers cannot expand quickly enough. Lumentum says some components are sold out through early 2029; Sivers is investing $30M to target more than 100M lasers a year, with operations expected in Q4 2027. Potential US restrictions on Chinese transceivers could reshape supply in 2027. Applied Optoelectronics and Sivers are the favorites. x.com x.com
Networking, power and compute look more durable than cyclical memory in the AI buildout. Lumentum's revenue compounded year after year across the 2010s and 2020s, unlike Micron's; AWS revenue also compounded. Lumentum, Nebius, Bloom Energy and Credo should heavily outperform SanDisk and Micron from here, though AI itself remains heavily doubted. x.com x.com
Lumentum and Coherent face a threat from microLED optics, which could replace laser optics with 20x less power and no temperature sensitivity by late 2027. The use case is short-reach, high-volume links under 30 meters. The copper versus optics debate is a distraction, since every AI cluster needs both and Credo now does both while Applied Optoelectronics handles long-haul; the bigger question is what microLED does to laser makers. x.com x.com
Scale-up moving to optics adds demand for optical suppliers even if new AI architectures need fewer switches. Bigger data centers and more GPUs add to that need; Jensen Huang expects future optics requirements to be several orders of magnitude above what NVIDIA ships today. This challenges the optics bear case for Lumentum, Coherent, Tower Semiconductor and Credo. x.com
Optical interconnect is where the next physical choke point in AI shows up, now that HBM memory has become crowded. Lumentum's lasers are sold out and Coherent's backlog runs to 2028, so Marvell raising its interconnect growth outlook to 70% from 50% is just more demand that supply can't answer. Per @PhotonCap, all three companies secured $2 billion strategic commitments from Nvidia. x.com x.com
Applied Optoelectronics, Lumentum and Coherent face share losses at US hyperscalers to Innolight. Channel checks favor Innolight gaining share in Q2 2026, at the expense of those three. Per @poezhao0605, a $7B Hong Kong IPO, the largest in years and upsized from $5B on strong demand, has cleared its listing hearing. x.com
Networking is the next AI hardware trade after memory, whose share of rack value rose from 15% to 35%. Optics is moving inside the server, as Nvidia's Feynman architecture builds in co-packaged optics and silicon photonics. Scale-across networking could grow 30% a year just linking existing data centers, as inference moves closer to users and training spreads across sites. Supply is tight: Lumentum's CEO said it can meet only 30% of expected 2027 demand and sees optics tighter than memory. x.com
