World's first light-powered neural processing units (NPUs) could massively reduce energy consumption in AI data centers
- Lightmatter unveiled the Passage M1000 optical interposer and Passage L200 co-packaged optics this week, showcasing them at the Optical Fiber Conference in San Francisco.
- The interconnects address the increasing demand for chip-to-chip bandwidth driven by denser AI deployments and bandwidth scaling impediments.
- Passage M1000 targets XPUs and multi-die switches, using light to transfer data and supporting 256 fibers, while Passage L200 is a 3D CPO product for XPUs and switch silicon designs.
- Lightmatter claims the M1000 supports up to 114 Tbps of optical bandwidth, and the L200 promises 32 Tbps of bidirectional bandwidth, which is a 5 to 10x improvement.
- Lightmatter plans to ship the Passage M1000 in the summer of 2025, manufactured by GlobalFoundries, and the Passage L200 in 2026, offering improved AI performance scaling.
19 Articles
19 Articles
World's first light-powered neural processing units (NPUs) could massively reduce energy consumption in AI data centers
Q.ANT's new chip uses photon power in a bid to solve AI's big energy issue. It's also 50 times faster than silicon-based equivalents, the company says.

Lightmatter releases new photonics technology for AI chips
SAN FRANCISCO : Lightmatter, a startup valued at $4.4 billion, on Monday released two pieces of technology aimed at speeding up the connections between artificial intelligence chips.Instead of moving information between computer chips as electrical signals, Lightmatter's technology uses optical connection
Coverage Details
Bias Distribution
- 100% of the sources are Center
To view factuality data please Upgrade to Premium
Ownership
To view ownership data please Upgrade to Vantage