The 400G-per-lane era is here. At OFC 2026, Coherent showed InP components with more than 100 GHz of bandwidth - but how do you keep TDECQ low while absorbing a 5 dB sensitivity penalty? A deep dive into the technical and commercial tradeoffs of IM/DD for AI interconnects.

The OCP 2026 white paper, Lumentum's capacity lock-up race and Anthropic's 1-million-TPU deal mark the arrival of the optical direct-connect era. The full edition adds STT's golden-ratio formula, cube-position formula, a $22k OCS BOM, an 800G/1.6T BOM comparison and 13 stock view cards.

When the biggest buyer reaches upstream and index providers add optical names to their benchmarks, the industry's capital structure has shifted gears. NVIDIA's $2B stake in Coherent is not an investment; it is supply insurance.

Tower put a $1.3B 2027 SiPho contract on the table, and that is only the committed portion; management hinted 2028 will be "substantially larger." The question is no longer whether SiPho takes off, but whether 300mm capacity can keep up, which is the real reason Tower took full ownership of Fab 7 in Japan.

This week: Lumentum's revenue nearly doubles as 1.6T ramps, NVIDIA invests $2B in Coherent's InP capacity, China's NDRC blocks Meta's Manus deal, and TSMC's COUPE silicon photonics platform heads toward mass production.

A not-yet-effective FCC draft rule repriced the optical supply chain in a week: Innolight plunged, Coherent soared over 40%. But the real bottleneck isn't module assembly; it's upstream EML and InP. The policy arbitrage is real, the capacity shift is slower.

After 1.6T there is no single winner: every reach has its own lowest-power answer. Two companies on this panel gave the same number — the reverse gearbox feeding slow-and-wide optics costs 4 pJ/bit, more than the optics themselves. And Oracle's field data pulled the spotlight back to the 4,096 fiber contacts on a front panel.

Scale-across planning has shifted from "how many wavelengths do I need" to "how many fibers do I need", because regenerator sites can no longer supply the power to light that many fibers. Acacia mapped the whole optics landscape onto four scales in a seven-step chart, and the 400G/lane physics line is now setting the handover timeline for campus DCI.

At ECOC 2026, Coherent skipped the CPO sales pitch and brought a working machine: a 6.4T NPO with its own ELSFP at 4.7 pJ/bit — laser included. LightCounting puts the crossover at 2031. The CPO debate is over; what remains is a fight over the slope.

Coherent turned "AI scales on optics" into financial results: accelerating revenue, gross margin up in eight of nine quarters, and leverage down to 0.7x. The real variable is not demand but how fast 6-inch InP capacity ramps.

Semtech moved its full-year data center growth language from "at least 50%" to "no cap." More telling than the $291M in revenue: 1.6T, CopperEdge and HieFo all ramp together in the second half, and "capacity is king" is becoming the deciding factor of this optical cycle.

The optics problem at 400G/lane is essentially solved; what remains is electrical, packaging and connectors. Coherent pinpointed the bottlenecks: wire bonds collapse above 110 GHz, OSFP connectors dip at 90 GHz, and PAM6 in optics hits an SNR wall. IM/DD can stretch to 800G, but it needs electronics, not photonics.

With a three-year deal and a ~US$22.3M prepayment, Coherent has locked in the scarcest 6-inch InP wafers. This isn't ordinary procurement — it's supply insurance, marking the InP shortage's shift from waiting in line to grabbing capacity at the source, and forcing a choice between cheapest China capacity and Non-China supply insurance.

The FCC wants to move optical module orders away from China, but the Western vendors taking those orders still grow their laser dies on substrates China approves for export. 6-inch InP is up 250% with a supply gap above 70% — de-Sinicization works at the module layer, but has no answer at the materials layer yet. This week's three earnings calls and OCP's CPO move point the same way: the decisive move has shifted down to indium phosphide.

AI's deepest bottleneck isn't the GPU — it's a 6-inch InP wafer. Silicon photonics and TFLN still need it to emit light, supply falls more than 70% short, volume relief waits until 2028, and NVIDIA has stepped in to lock up capacity. STT maps the chain from phosphorus to laser and where Taiwan stands.

Four industry heavyweights on one stage reached an unromantic consensus: demand isn't a bubble, technology isn't heading for a revolution, and the real battlefield is manufacturing. For Taiwan's supply chain, that means the winners will be decided by yield, process, and automation.

In one week, TSMC moved CPO from slideware onto a production schedule, and Lumentum seized pricing power at the light source. Two narrow gates, TSMC COUPE capacity and 200G EML supply, now decide who gets to play in the 2026 1.6T cycle.

AXT didn't just grow this quarter — it shifted gears. Record InP revenue of $30.7M lifted gross margin to 45% and flipped the company to profit. Supply is now the only ceiling: backlog tops $100M and this year's capacity target is raised to more than triple.