Flapping Airplanes, a one-year-old AI research lab, has raised a $1 billion Series A round at a reported $6 billion valuation. Index Ventures and Lux Capital co-led the round, announced on October 21.[1][2][3] The valuation is roughly four times the $1.5 billion price investors placed on the company during its $180 million seed round in January.[1][2][3]
Existing backers Sequoia, GV and Kleiner Perkins joined the Series A. The company also listed Valor, Atreides, A*, Nvidia, Jump Trading Group and XTX Ventures as institutional backers.[1][2][3] Index's Mark Xu will join the company's board.[3] The lab has not released a commercial product, according to SiliconANGLE.[3]
Flapping Airplanes' three co-founders are Aidan Smith, Asher Spector, and Ben Spector. Smith is a Thiel Fellow who worked at Neuralink during college. Asher Spector holds a statistics doctorate from Stanford. Ben Spector founded Prod, a student-run startup accelerator, while at MIT and is pursuing a computer science doctorate at Stanford. He was the lead author of a paper introducing ThunderKittens, an open-source toolkit for fast GPU kernels.[1][3]
The team is building what its founders call "the world's most complete megakernel framework." A megakernel bundles an entire program into one GPU kernel, cutting the time lost between steps. According to Flapping Airplanes, the framework runs a full training job, serves inference and plays a video game concurrently in one kernel, allowing memory-bound inference and compute-bound training to happen on the same GPU without idle time.[1][3] The goal is to train large AI models with far less data. Asher Spector told TechCrunch, "A model that's a million times more data efficient is probably a million times easier to put into the economy."[1][3] The company said on X that its mission is to "unlock new markets in Al, including robotics, scientific discovery, and more, by pioneering radical new techniques to improve efficiency."[2] The founders say AI has only scratched the surface of what GPUs can do, and they think they can create intelligent systems "without ingesting half the internet."[1]
Index's Mark Xu and Shardul Shah said they expect AI demand to keep outstripping available compute and data, and they pointed to robotics, trading and life sciences as areas that need more efficient AI.[3] Index describes a team of about two dozen people working together in person in San Francisco.[3] The company has raised a total of $1.18 billion, SiliconANGLE reported.[3]
Sources
To report an error in this article, write to corrections@venturistnews.com. Corrections log.
Elliot Marsh covers startups, IPO filings and the money behind them for Venturist. Elliot writes from SEC filings, company announcements and published reporting.