Nvidia bets on new technology to scale AI infrastructure

Traders Union
All News
Financial News
Nvidia bets on new technology

Highlights

Nvidia has made $2 billion commitments to Lumentum, Coherent and Marvell, and a reported $500 million investment in Corning.
The company also participated in Ayar Labs 500 million Series E round, aimed at scaling co-packaged optics for AI systems.
Photonics could reduce the power and bandwidth pressure created by large GPU clusters.
The main test now is whether suppliers can scale manufacturing quickly enough for data-center demand.

A supply chain built around light

According to CNBC, the push began in March, when Nvidia announced separate $2 billion investments in Lumentum and Coherent. Both deals include multiyear, nonexclusive agreements, purchase commitments, and future access to advanced laser or optical networking capacity. The stated goal is to expand manufacturing and research for next-generation AI infrastructure.

Later that month, Nvidia said it had invested $2 billion in Marvell as part of a partnership tied to NVLink Fusion, custom AI infrastructure, and silicon photonics. In May, Nvidia and Corning announced a long-term partnership under which Corning will expand U.S. optical connectivity manufacturing capacity tenfold, increase domestic fiber production by more than 50%, build three new plants in North Carolina and Texas, and create more than 3,000 jobs. Dow Jones, via Morningstar, reported that Nvidia is investing $500 million in Corning as part of that partnership.

The market has continued to price Nvidia as the central company in the AI infrastructure boom. Its shares recently traded at $214.25 on Nasdaq, up 0.78%, with an intraday range of $209.75 to $215.43 and a market value of about $5.23 trillion.

The AI bottleneck moves beyond chips

The bet reflects a change in the economics of AI infrastructure. Large AI systems require thousands of GPUs working together, and those chips must constantly exchange data with memory, switches, servers, and other data centers. Copper remains widely used because it is familiar and reliable, but optical links can move data at high speed with lower energy demands over larger systems.

Nvidia has already brought photonics into its own networking roadmap. The company says its Spectrum-X and Quantum-X silicon photonics switches are designed to connect millions of GPUs across sites, with 1.6 terabits per second per port, 3.5 times energy savings, and 10 times better network resilience at scale compared with traditional methods.

Connectivity becomes a strategic constraint

The scale of Nvidia business explains the urgency. In the first quarter of fiscal 2027, Nvidia reported record revenue of $81.6 billion, up 85% from a year earlier, with data center revenue of $75.2 billion. Its shares recently traded at $214.25, giving the company a market value of about $5.23 trillion.

For Nvidia, photonics is not a side bet. It is a way to reduce the risk that AI systems run into a wall of power consumption, heat and data-transfer limits.

In an earlier report, we noted that Nvidia chip case puts Asia export controls under scrutiny.

This material may contain third-party opinions, none of the data and information on this webpage constitutes investment advice according to our Disclaimer. While we adhere to strict Editorial Integrity, this post may contain references to products from our partners.