Nvidia Commits at Least $6.5 Billion to Photonics as AI Data Centers Strain Power
Updated
Updated · TradingView · May 29
Nvidia Commits at Least $6.5 Billion to Photonics as AI Data Centers Strain Power
8 articles · Updated · TradingView · May 29
At least $6.5 billion has been committed by Nvidia since March to photonics suppliers including Lumentum, Coherent, Marvell and Corning, while it also joined an Ayar Labs funding round.
Photonics moves data with light rather than electrical signals, a shift Nvidia is backing to connect growing numbers of AI chips and servers without the rising power draw and cost of copper wiring.
Jensen Huang has said current silicon photonics capacity is still insufficient for future AI demand, underscoring why Nvidia is investing early across the supply chain.
The bet could open a new growth leg for Nvidia and lift optical-component and networking suppliers, but large-scale AI data-center deployment remains the key test.
With manufacturing yields so low, is Nvidia's $6.5B photonics bet a solution or just a gamble against AI's energy crisis?
Will making AI more efficient with photonics ultimately lead to a greater drain on global power grids?
Nvidia’s $4 Billion Photonics Investment: Securing the Future of AI Infrastructure
Overview
In 2026, Nvidia announced a $4 billion investment in photonics technology to address the growing bottlenecks in data transfer between AI chips. As traditional electrical signals struggle to keep up with the massive data volumes and speeds required by advanced AI models, Nvidia is turning to photonics, which uses light instead of electricity for chip connections. This shift enables faster, more energy-efficient, and higher-bandwidth data pathways, essential for scaling AI supercomputers and data centers. By pioneering these foundational technologies, Nvidia aims to significantly enhance its AI infrastructure and maintain its leadership in the industry.