Nvidia Bets $6.5 Billion on Photonics to Break AI’s Energy Barrier

Energy

Share on :

Facebook
X
LinkedIn
Pinterest
WhatsApp
Email

Prime Highlights

  • Nvidia invested $6.5 billion into photonics firms over three months.
  • Large-scale photonics adoption across AI infrastructure expected from 2028 onwards.

Key Facts

  • Nvidia is the world’s leading GPU chipmaker powering global AI infrastructure.
  • Photonics uses light to transmit data, replacing energy-intensive copper electrical systems.

Background

NVIDIA has committed at least $6.5 billion into photonics technology companies over the past three months, as the chip giant moves aggressively to address one of AI infrastructure’s biggest bottlenecks. Photonics uses light instead of electricity to transmit data and is considered a more energy-efficient alternative to copper-based electrical transfer, which increasingly limits broader AI deployment.

Since early March, Nvidia has announced $2 billion investments into Lumentum, Coherent, and Marvell, all of which are actively developing photonics technology. The company also committed $500 million into Corning for advanced optical connectivity solutions and participated in optics startup Ayer Labs’ $500 million Series E funding round.

Alvin Nguyen, senior analyst at Forrester, said photonics allows Nvidia to scale AI infrastructure without the energy costs associated with electrical and copper systems. He added that these investments ensure continued photonics advancement and prevent Nvidia from hitting a scalability and performance wall.

NVIDIA CEO Jensen Huang, speaking at GTC in March, said the company is scaling its silicon photonics technology and adding it to GPU-to-GPU interconnect systems. He noted that the required photonics capacity far exceeds what the world currently produces, prompting Nvidia to work closely with its supply chain to build capacity ahead of demand.

Stocks of invested companies have surged sharply, with Lumentum up 134%, Marvell up 122%, Corning up 111%, and Coherent up 96% since the start of the year.

According to experts, large-scale manufacturing is still challenging, considering widespread adoption is expected only from 2028 onwards.

Read Also : PingPong and Visa Launch B2B Card Payment Tool for Global Supplier Transactions

Related Articles: