Luminous is developing a custom AI supercomputer stack centered on proprietary silicon photonics to eliminate data‑movement bottlenecks and improve performance and programmability. The company is building nearly every part of the stack from chips to software, aiming for order‑of‑magnitude gains in performance and simplified programming models. Luminous says its work will enable much larger, more efficient AI models and support commercial‑scale AI applications. The company plans to double its engineering team and is actively recruiting photonics designers, digital and analog VLSI engineers, packaging and system integration engineers, and machine learning experts. Funding announced in the articles will support custom chip and software development and gearing up for commercial production. Luminous frames its mission around delivering useful, usable, and safe AI by providing the hardware needed to run tomorrow’s AI applications. Luminous Computing is developing photonics chips designed to handle AI workloads by using light to move and multiply dense arrays of data at high speed. The company says its approach will remove major bottlenecks that traditional processors struggle with, and its architecture is based on CTO Mitchell Nahmias’ Princeton research. The one-year-old, seven-person startup was founded by Michael Gao, CEO Marcus Gomez and Nahmias; Gomez and Gao have prior software and data-science experience. Luminous reports it already has working silicon and aims to ship development kits within the next few years. The company claims a single chip could replace the computing power of roughly 3,000 Google TPUs. It raised capital to grow the team and specifically hire people with semiconductor industry experience to accelerate engineering and development.

Total raised

$114M

Funding rounds

2

Latest round

Series A

Latest activity

Mar 2022