Keep up with the innovative tech transforming business
Tech Brew keeps business leaders up-to-date on the latest innovations, automation advances, policy shifts, and more, so they can make informed decisions about tech.
Last Thursday, a buzzy new supercomputer startup debuted—with $105 million in Series A funding.
The startup’s ethos? Software has come a long way, but the hardware to support everything AI can do doesn’t exist yet. And Luminous Computing believes it has the solution: building the world’s most powerful supercomputer.
- Luminous’s investors include Bill Gates and Gigafund, the VC firm founded by PayPal and Founders Fund alumni.
Behind the hype: Luminous’s business plan centers on an alternative approach to chipmaking: photonic chips. It posits that the efficiency of light—which moves faster and generates less heat than electricity, which is typically what processors use to send signals—could lead to higher performance.
In July, we spoke with Nick Harris, cofounder and CEO of a Luminous competitor called Lightmatter, who told us that “the major challenge—it’s really about energy density. Essentially, computer chips today are way too hot.”
- He added that, using light, “we’re able to get around this fundamental technology challenge, this energy problem. And since light defines the speed limit in our universe, it’s the fastest thing out there.”
- Other competitors include Celestial AI and Intel.
Luminous also has rivals on the “world’s most powerful AI supercomputer” front. Meta is working on one that it claims will become the fastest of its kind in the world. The Meta system’s processing power is currently on par with the world’s fifth-fastest supercomputer.
Looking ahead, Luminous plans to use its newfound millions to double its engineering team, invest in the startup’s chips and software products, and scale up for the commercial market.
Big picture: Luminous wants to make it “much easier to build huge AI models,” Marcus Gomez, CEO and cofounder, said in the company’s release. As loyal readers of this newsletter may know, ever-faster and ever-larger AI models are an increasingly controversial debate in the field right now, for their potential to learn and spread patterns of bias.