Skip to content
primarysourced Photonics sector Coherent
COHR
~7 min read · 1,632 words ·updated 2026-04-29 · confidence 69%

InP EML supply duopoly — Coherent vs Lumentum

The electro-absorption modulated laser (EML) is the source-laser device for 100G/200G/400G/800G/1.6T fiber-optic datacenter transceivers. Merchant EML supply is a structural duopoly between Lumentum and Coherent Corp (the latter via the Finisar-2019 acquisition’s Sherman TX wafer-fab heritage). Together they hold ≥80% of merchant supply, with the residual ~20% split among captive vendors (NTT Electronics, Sumitomo Electric — primarily telecom) and emerging Chinese fabs (HG Genuine, Accelink — primarily domestic-China supply). The duopoly is the structural basis for both companies’ margin expansion through the AI cycle.

Why the duopoly is structurally stable

Three reinforcing moats:

  1. Capital intensity — InP wafer fabs require multi-year qualification cycles, multi-hundred-million-dollar fab investment, and vertical integration through epitaxy → fab → packaging → test. Greenfield InP capacity is not a 12-month build. The Coherent 6-inch (150mm) InP fab in Sherman TX took multiple years to bring online (announced March 2024; entered production in the September 2025 quarter as the world’s first InP 150mm production line — pre-Sept 2025 wafer was 3-inch; full target wafer-start rate by Q4 CY2026 ✓).
  2. Process know-how — 200G/lane EML modulation requires sub-micron MQW (multi-quantum-well) layer control, optical-coupling-loss optimization, and reliability engineering measured in 100,000+ hours of accelerated-life testing. Yield curves climb slowly with cumulative wafer starts; incumbents enjoy a learning-curve moat.
  3. Customer qualification — hyperscalers (and the module-assembly vendors serving them) qualify EML chips at the device level for each module form factor and each speed-grade generation. Re-qualifying a new EML supplier is an 18–24-month project; customers prefer dual-source allocations within the existing duopoly to a true tri-source.

The pricing-power consequence is captured in management framing on the Q1 FY2026 call. Coherent management said EML supply was the binding constraint and would improve sequentially through CY2026 (Q1 FY2026 transcript Nov 6, 2025 via Motley Fool) ◐.

Comparative position: Coherent vs Lumentum

MetricCoherent CorpLumentumNotes
InP fab footprintSherman TX (6-inch) + Allen TX + Järfälla SwedenSan Jose CA (4-inch) + Towcester UK (4-inch) + new Greensboro NCCoherent has the wafer-size advantage on legacy 6-inch ramp
6-inch (150mm) InP capabilityWorld’s first 6-inch InP fab announced March 2024; entered production September 2025 quarter (world-first 150mm production line); ramping volume CY2025–2026Greenfield 6-inch fab announced post-NVDA partnership 2026; Greensboro NC mid-2028 rampCoherent ahead by ~2 years on wafer-size advantage
200G/lane EML productionQualifying multiple products on 6-inch; in production on 200G EML, DFB-MZ, photodiodesVolume shipping; first-to-volume claim on some hyperscaler qualificationsBoth ramping through CY2026
Datacenter & Communications revenue (Q2 FY2026)$1,208M (+33.6% YoY)$665.5M total LITE Cloud & Networking dominant shareCoherent’s segment is broader (incl. systems, telecom, undersea)
NVDA strategic investment (Mar 2026)$2B in 7,788,161 common shares at $256.80 (immediate dilution)$2B in 2,876,415 Series A Convertible Preferred at $695.31 (HSR-gated)Symmetric capital; structural-instrument asymmetry
Vertical integrationComponent → Finisar heritage transceiver assembly (deeper)Component → Cloud Light transceiversCoherent’s module business is longer-running
FY2025 total revenue$5.81B (+23.4% YoY)$1.65B (+~24% YoY)COHR materially larger; LITE more pure-play to optics
FY2025 segment-mix concentrationNetworking 59% / Lasers 25% / Materials 16%Cloud & Networking ~75%+ (post-Cloud Light) / Industrial Tech ~25%-LITE is purer optics-AI play

Coherent’s broader portfolio (industrial lasers, materials/SiC) dilutes its pure-play exposure to AI photonics; Lumentum is the higher-beta name on the InP-EML thesis specifically. Coherent’s industrial / SiC ballast is a feature when AI capex digests and a drag when AI capex accelerates — see 05_financials segment revenue mix.

6-inch InP wafer-fab — Coherent’s structural advantage

Coherent announced the world’s first 6-inch (150mm) InP scalable wafer fabs in March 2024, located in Sherman TX and Järfälla Sweden (Coherent press release) ✓. Sherman TX entered production in the September 2025 quarter as the world’s first InP 150mm production line — the prior generation was 3-inch wafer. The fab transition from 3-inch (legacy industry standard) to 6-inch delivers:

  • 4× more chips per wafer than 3-inch; ~2.5× more than 4-inch (Lumentum’s legacy)
  • ~60% reduction in die cost (Coherent disclosure)
  • Improved cycle time per chip via standard front-end semiconductor-fab automation tooling

Coherent disclosed in November 2025 commentary that the 6-inch ramp was at ~80% of the target wafer-start rate with internal InP capacity targeted to double by Q4 CY2026 (Coherent Q1 FY2026 transcript) ◐. The Texas Semiconductor Innovation Fund granted $14,076,031 in February 2026 to accelerate scaled InP production at Sherman, with Coherent’s project totaling >$154M in capital investment (Semiconductor Today) ◐.

Lumentum’s response is the NVDA-funded San Jose new fab + Greensboro NC fab announced post-March 2026. Lumentum is moving to 6-inch but lags Coherent by approximately 18–24 months on volume readiness — a margin-relevant gap during the supply-constrained CY2026–CY2027 window.

Capacity-build timing

Capacity eventCoherentLumentum
Sherman TX 6-inch (150mm) InP fab (COHR)Announced Mar 2024; entered production September 2025 quarter (world-first InP 150mm production line); ramping CY2025–2026; double internal capacity Q4 CY2026n/a
Järfälla Sweden 6-inch (COHR)Announced Mar 2024; complementary capacityn/a
San Jose new fab (LITE)n/aNVDA-funded; announced Mar 2026; volume-ramp CY2027–2028
Greensboro NC 6-inch (LITE)n/aNew 6-inch fab announced Mar 2026; volume ramp mid-2028
Towcester UK expansion (LITE)n/aOngoing; legacy Oclaro fab
200G/lane EML productionRamping through CY2026Ramping through CY2026
1.6T module qualificationEML + SiPh ramping CY2026; VCSEL-based 1.6T 2H CY2026EML ramping CY2026
CPO commercial volume2027 first NVDA pluggable CPO; 2028 NVDA scale-upSame industry timeline

Capacity-build dates are management-framed; actual capacity-coming-online milestones are confidential and only inferable from quarter-on-quarter capex disclosures and InP-supply commentary.

Pricing discipline mechanics

In a structurally supply-constrained duopoly with fungible end customers, pricing power is asymmetric:

  1. Spot ASPs rise when customer demand growth (running ahead of capacity adds) outpaces supply growth — this is the current regime, with double-digit ASP increases on 200G EMLs in CY2026 (per industry trade-press analyst commentary — ⚠).
  2. Volume contracts (multi-year) lock in floor pricing — NVDA’s $2B equity investment plus multibillion-dollar purchase commitment to Coherent (and parallel structure at Lumentum) is a volume pre-purchase against capacity each merchant is dedicating to NVDA. The implied ASPs on the committed-volume tier are confidential but presumed to bracket below merchant-spot ASPs while above pure cost-plus.
  3. Mix shift to higher speeds drives ASP growth — 800G EML chip ASPs are higher than 400G; 1.6T EML ASPs higher still. Coherent management noted on the Q2 FY2026 call that 1.6T module gross margins would exceed 800G levels “particularly early in the lifecycle,” aided by 6-inch InP cost advantages (paraphrased from Futurum Group analysis) ◐.

The pricing discipline is fragile in two scenarios: (a) hyperscaler demand inflects negative on AI capex digestion, leaving capacity adds stranded; (b) NTT Electronics or HG Genuine credibly enters merchant datacom EML supply at scale, converting the duopoly to a tri-opoly with attendant margin compression. Neither has materialized as of April 2026.

NVDA’s dual-source structure preserves the duopoly

NVDA’s parallel $2B equity investments in both Coherent and Lumentum on March 2, 2026 (CNBC) are the most informative signal on duopoly persistence. If NVDA had concluded one supplier was superior, a unilateral investment would have been the move. Instead NVDA chose the symmetric structure:

  1. Bilateral capacity dedication — NVDA gets dedicated capacity at both incumbents.
  2. Dual-source supply continuity — NVDA preserves its negotiating position by keeping two viable suppliers.
  3. No obvious tilt of competitive dynamics between the two — both vendors raised capacity in parallel.

This is also a tell against the “NVDA in-sources optics” risk. Owning equity in both incumbent merchant EML suppliers is consistent with NVDA’s view that the merchant model serves NVDA better than vertical integration — at least through the 2027–2028 capacity cycle.

What would break the duopoly

  1. A credible third-source emerges with hyperscaler qual at volume — most plausible candidates: Sumitomo Electric (Japanese; primarily telecom; could pivot), HG Genuine (Chinese; export-control-constrained for Western shipments), or Marvell-internal silicon photonics + heterogeneous-integration approaches that monolithically integrate the source laser.
  2. Hyperscaler in-sources — least likely; the qualification + capex burden is high relative to merchant supply availability. AWS has the largest in-house optics team, but even AWS qualifies merchant EML for the bulk of its volume.
  3. Silicon-photonics + heterogeneous-integration approaches displace InP at the source-laser layer — currently silicon photonics relies on InP gain-chip flip-chip-bonded onto SiPh waveguide; the InP component is still merchant-sourced. Disruption would require monolithic InP-on-Si or quantum-dot-on-Si lasers reaching production — multi-year horizon.

Sources