Old Huang is about to turn optical modules into hard currency, the next battleground in the AI computing power arms race.

NVIDIA has committed since March of this year to invest at least $6.5 billion in photonics and optical networking-related companies, investing $2 billion each in Lumentum, Coherent, and Marvell, investing $500 million in Corning, and participating in Ayar Labs' $500 million Series E funding round to develop optical connection systems. Photonic technology is expected to improve bandwidth and reduce energy consumption, becoming a core infrastructure for AI data centers; as model sizes expand, the demand for interconnection between GPUs has surged, and network interconnects have become a new bottleneck. Jensen Huang stated at GTC that he is accelerating the application of silicon photonics in network platforms and GPU interconnects, with related capacity demand far exceeding global supply capabilities. The high-speed optical interconnect demand in AI server racks has benefited the industry chain and the stock prices of invested companies.