By Jane Lanhee Lee
OAKLAND, Calif. (Reuters) – Lightmatter, a Boston-based startup that uses light for AI computing, said on Wednesday it raised $154 million and tripled its valuation as customers in need of faster computing and more energy-efficient for AI work buy its systems.
“We use light to link computer chips together and we also use light to perform calculations for deep learning,” said Lightmatter co-founder and CEO Nick Harris. “The reason we’re getting these customer- and datacenter-scale deployments with our interconnect is because the generative AI boom is driving high-end chips like crazy.”
Light has been used for decades in the telecommunications industry to move data around the world with fiber optic cables. Today, major technology companies are looking to use light-based technology in their data centers to reduce energy costs. The challenge has been to reduce the size of the devices used to create or control the light.
As chips become more advanced, they consume more power. “That can be up to half the total cost of running a data center just to power the chips,” Harris said, adding that light transmits information much more energy efficiently than light. electrical signals flowing on the wires.
Harris said his chip called Passage, which connects other chips by translating electrons into photons and back, will be mass-produced and used in data centers next year. Much of the funding will help roll out this product.
Its more futuristic product Envise is a system that uses light, instead of electrons, to do computation. Harris said Envise will be able to run large AI models but will not train them. Customers will test the system this year, he said.
GV (Google Ventures), the venture capital investment arm of Alphabet Inc, and Fidelity Management & Research Company were among the investors participating in this latest round, the company said.
(Reporting by Jane Lanhee Lee; Editing by Lisa Shumaker)