<cite index="61-5">OLIX Computing, a London-based developer of Optical Tensor Processing Units (OTPUs), raised approximately $312 million (€270.5 million) in Series B funding at a $3.3 billion valuation in early August 2026, led by Fundomo with participation from Arm, Hudson River Trading, Reed Hastings, and existing backers</cite>. The company's thesis addresses a critical AI infrastructure constraint: photonic (light-based) on-chip interconnects can reduce the energy and speed bottlenecks caused by copper wire connections in large-scale AI computing systems.
<cite index="61-2">The deal signals strong investor interest in chips, power use, cooling, and other AI bottlenecks that make computing expensive</cite>. OLIX's OTPU architecture uses light instead of electrical signals to move data within and between processors, enabling higher bandwidth density at lower power draw compared to copper-constrained traditional processors. The company is positioning itself alongside other optical interconnect startups (Lumilens, Ayar Labs, Lightmatter) targeting the same physical limits: hyperscalers scaling to 100,000+ GPU clusters face copper signal degradation at ~1.5m distances.
For production teams: OLIX's Series B and Lumilens' contemporaneous emergence reflect investor conviction that the optical interconnect stack—both scale-up (within-rack, co-packaged) and scale-out (cross-rack fiber)—is now table-stakes infrastructure for AI factories. The energy efficiency gains (photonics vs. copper) become material at 300+ megawatt deployments. Architects planning multi-year capacity expansions should evaluate OTPU suppliers alongside network silk (Broadcom, Marvell) and co-packaged optics, understanding that switching costs accumulate early in the design phase; optical integration decisions lock in topology and cooling design.