Over the past three months, NVIDIA has invested at least $6.5 billion in multiple photonic technology development companies, accelerating its strategic push to overcome a major bottleneck in practical AI application.

Photonic technology, which uses light signals to transmit data, is an emerging field with significantly higher efficiency than the current mainstream electrical signal transmission. Electrical data transfer consumes more energy, a growing impediment to the widespread adoption of artificial intelligence.

Since early March, NVIDIA has completed investments totaling $2 billion in three photonic technology firms: Lumentum, Coherent, and Marvell. The chip giant also announced a $500 million investment in Corning for joint development of high-end optical interconnect solutions and participated in a $500 million Series E funding round for Ayar Labs, a startup in the optical space.

Alvin Nguyen, a senior analyst at Forrester, stated in an interview, "With photonic technology, NVIDIA can scale AI infrastructure while avoiding the high energy consumption issues that come with continuing to use copper cables and electrical signals."

"Investing in photonic technology companies ensures continuous iteration in this field. If NVIDIA sticks with traditional copper and electrical solutions, it will eventually hit a ceiling in computing power expansion and performance enhancement. Focusing on photonics effectively mitigates that risk."

Within AI infrastructure, photonic technology can replace traditional copper-based electrical signals, using light to transmit data between graphics processing units (GPUs), memory, network chips, servers, and data centers.

Brian Colello, a senior equity analyst at Morningstar, noted that while copper cables currently dominate as the mainstream interconnect solution due to low cost and high reliability, photonic technology is poised to play an increasingly vital role in AI infrastructure over the long term.

He explained, "NVIDIA's planning for its next-generation rack-scale AI solutions will require more optical interconnect technology to handle the exponential growth in bandwidth driven by surging demand for new models and computing power."

Currently, NVIDIA has already integrated some photonic technology into its networking products. The company has introduced related supporting tools that can help AI computing centers connect millions of GPUs across sites while significantly reducing energy consumption and operational costs.