Nvidia announced Monday that it is investing $2 billion each in Lumentum and Coherent , which develop photonics technology for data centers, such as optical transceivers, circuit switches and lasers, used to transfer data at high speeds over long distances .
See also: New Meta and Nvidia agreement for millions of AI chips
Their technology could improve energy efficiency, data transfer speeds and bandwidth in future AI data centers, following Nvidia's successful move in 2020 with the acquisition of networking hardware company Mellanox to strengthen NVLink and increase the amount of data moving between its GPUs.
For Lumentum, the non-exclusive multi-year agreement includes a “multi-billion dollar purchase commitment and future capacity access rights for advanced laser components,” as well as support for expanded research and development and manufacturing. The agreement with Coherent is described similarly, with a “multi-billion dollar purchase commitment and access and capacity rights for advanced laser and optical networking products.”
See also: Nvidia CEO denies he is "unhappy" with OpenAI – Big investment coming
The development of agentic AI, such as Anthropic’s Claude Cowork and Microsoft’s Copilot Tasks , is increasing the bandwidth needs of AI data centers to support the rapid execution of multiple tasks. Photonics could offer a solution — optical fibers can support significantly higher bandwidth and lower latency than copper cables, while also consuming less power.
Nvidia isn't the only organization paying attention to photonics. Last month, DARPA issued a call for research proposals to improve photonic computing, especially in relation to AI applications.
See also: Trump says Nvidia can sell more powerful AI chips to China
Nvidia rival AMD also acquired silicon photonics startup Enosemi last year, saying it would “accelerate” AMD’s innovation in optics for its AI systems.
