Photonics start-up Luminous Computing makes $105 million Series A • The Register
Luminous Computing, a startup using photonics to drive artificial intelligence, has raised venture capital, securing $105 million in Series A funding from a range of investors including Microsoft co-founder Bill Gates.
The Bay Area upstart, founded in 2018, announced the funding round, which included companies including Gigafund, 8090 Partners, Third Kind Venture Capital, Alumni Ventures Group and Strawberry Creek Ventures. He adds to the $1 million in pre-seed money the company received in 2018 and the $9 million in seed funding was withdrawn a year later.
Luminous officials said the new funds will be used to double the size of the company’s engineering team and build its custom chips and software as it moves towards commercial-scale production. . It is also continuing to recruit photonics designers, very large scale digital and analog integration (VSLI) engineers, packaging and system integration engineers and machine learning experts.
The company is among a growing number of vendors looking to photonics to solve thorny compute, latency and memory issues in AI and machine learning technology that are fast becoming the stones cornerstones of modern applications and industries, from autonomous vehicles to pharmaceuticals. .
The goal is to use light rather than electricity flowing over wires to send signals, which proponents say will enable the development of more sophisticated and less expensive AI systems.
“AI has gone superhuman,” said Marcus Gomez, CEO and co-founder of Luminous. in a report. “We can interact with computers in natural language and ask them to write a piece of code or even an essay, and the result will be better than most humans could provide. The frustrating thing is that we we have the software to solve monumental, game-changing problems that humans can’t even begin to solve, we just don’t have the hardware to run those algorithms.
According to Luminous, the problem is that current computers use electrical signals, which reduces performance as the signals travel. Electrical signals consume more power and carry less information over longer distances, and businesses now rely on software to compensate for bottlenecks in communications caused by hardware.
Additionally, today’s “AI supercomputers” (high-end processor plus accelerator) cannot keep up with the computing power needed to train models, and the software techniques used in the models are increasingly complex. The systems themselves aren’t scalable enough: Even machines with more than 1,000 processors often run at less than 20% efficiency, with the rest of the time sitting idle due to communication limits, according to the company.
The plans for Luminous come from years of research by the founders and other leaders of the company. Gomez was a research scientist at dating app company Tinder in 2018 and worked in research roles at Google (in artificial intelligence) and the Mayo Clinic, a software engineer at Bloomberg, and a network biology researcher at Harvard Medical School. . He dropped out of Stanford’s MS program to start Luminous.
Mitch Nahmias, co-founder and CTO of Luminous, earned his BS, MA, and PhD in electrical and electronics engineering at Princeton. While at school, Nahmias studied the relationship between a laser and a biological neuron, which helped create the field of neuromorphic photonics.
Matt Change, the company’s vice president of photonics, also earned a doctorate in electrical engineering from Princeton and spent two years at Apple designing hardware to reduce interference between coexisting wireless radios on the Apple Watch. , leaving in 2019. Prior to that, he was CTO at Rebeless, a startup creating photonic integrated circuit technology for broadband signal processing.
Luminous does not reveal details of its plans, although a 2019 report of its $9 million seed said the goal is to create a single photonic chip that can replace 3,000 TPUs, which are small ASICs created by Google and designed for low-power machine learning inference.
Luminous has more than 90 employees and has working prototypes of its chips in its labs, with plans to ship development kits to organizations within two years.
Although there is a lot of work going on in the area of silicon-photonic computing, companies are likely to tackle barriers such as cost, scalability, power consumption, and reliability. ®