OPA Silicon Photonics Chip Market

The OPA silicon photonics chip market closed 2025 at $2,860M, up from $2,440M the prior year. SiLC Technologies held 15.7% with $448M in recognized revenue, Analog Photonics took 14.2% at $406M, and Voyant Photonics claimed 12% at $343M. Intel Corporation sat at 11.08% with $317M, though our desk counts roughly half of that as co-packaged optics shipments into hyperscale customers rather than pure OPA beam-steering chips. North America accounted for 43% of the global installed base by our reckoning. The market is young. Most of the revenue today comes from prototype LiDAR modules sold into automotive tier-ones for integration testing, not volume production vehicles.

Automotive solid-state LiDAR and free-space optical comms drove the bulk of 2025 OPA chip demand. Industry sources suggest tier-one automotive suppliers locked multi-year LiDAR contracts with SiLC Technologies and Voyant Photonics during 2025, each deal tied to specific vehicle platforms launching in the 2027 or 2028 timeframe. Those design wins carry seven-to-nine-year lock-in periods because re-qualifying a LiDAR sensor for functional safety takes eighteen months minimum. The 1550nm wavelength camp (SiLC, Elenion) argues for eye safety at higher power budgets; the 905nm camp (Voyant, Analog Photonics) counters with lower-cost silicon germanium photodetectors and established CMOS foundry compatibility. By our count, the wavelength question is splitting the market into two non-interoperable camps, and the one that wins the 2027–2029 vehicle-platform cycle will compound share for a decade.

SiLC Technologies's 15.7% lead in 2025 came from partnerships with automotive suppliers and Chinese EV makers we're tracking under NDA. Analog Photonics has been gaining share, reportedly tied to partnership activity with Mobileye around OPA-based surround-view LiDAR for Level 3 highway automation. Market participants report pricing pressure in late 2025 as vendors competed to defend their tier-one automotive sockets. Intel's position is harder to read: the company's silicon photonics group ships both data-center co-packaged optics and OPA prototypes under the same P&L line, so the 11.08% share likely overstates pure OPA exposure by 40–50%. We think the real OPA-only share for Intel is closer to 6%, which drops them to fifth behind Elenion. The top three (SiLC, Analog, Voyant) held a combined 42% at year-end 2025, and we don't expect that to move more than 500 basis points through 2027 because the design-win queue is already set.

Three scenarios break the 17% CAGR view. First, if NHTSA or Euro NCAP downgrade the safety credit for solid-state LiDAR in upcoming ADAS testing protocols, then tier-ones pull forward cost-reduction roadmaps and OPA chip ASPs compress faster than our model assumes. Second, if any of the top three players hit a yield wall at 200mm wafer scale, then automotive customers revert to MEMS-mirror LiDAR, which is out of scope but still shipping at higher volume today. Third, reports suggest China's SMIC is developing silicon-photonics process capability targeting sub-$50 OPA chips for domestic EV makers; if that process ramps in 2026, the entire pricing curve shifts and Western vendors lose share to Jianguang, Accelink, and other Shenzhen-based entrants our desk hasn't modeled yet.