Discovered Materials raises $9M seed to hunt cooler AI chips
What's the deal? Discovered MaterialsDealroom has a profile for this one. Try Dealroom → has raised a $9 million seed round led by Lightspeed India PartnersDealroom has a profile for this one. Try Dealroom → to use swarms of AI agents to find new materials for more efficient integrated circuits. Peak XV Partners and angel investors Paul GrahamDealroom has a profile for this one. Try Dealroom →, Gokul RajaramDealroom has a profile for this one. Try Dealroom →, and Thariq ShihiparDealroom has a profile for this one. Try Dealroom → also participated. The startup recently emerged from Y Combinator.
Why now? Chips running AI workloads run hot, driving up data centre power use and cooling costs. Founders Advaith SridharDealroom has a profile for this one. Try Dealroom → and Akash RamdasDealroom has a profile for this one. Try Dealroom → are turning AI on the problem AI helped create.
What's the endgame? The two built a software pipeline that uses Anthropic models to generate material leads, then runs simulations with physics models they trained to verify candidates. The plan: patent the use of promising materials in GPUs — or the process to make chips from them — and license to chipmakers.
Sridhar hopes to have new materials worth patenting within the year. "[Ramdas] was doing maybe 20 guesses a day during his PhD," he said. "We're able to do thousands of guesses a day now by having these agents run 24/7 on the cloud."
Product update: The company released examples of hundreds of new materials, plus a "Material Discovery Bench" to track how frontier models take on the challenge. It says it has already found several materials matching the properties used by major chipmakers, though it won't share details.
What could go wrong? The engineering trade-space is tough: a material that dissipates heat may be too hard to manufacture, or its electrical properties may suffer. "It's a bit of playing whack-a-mole with atomic structures," said Hemant Mohapatra, the Lightspeed partner who led the round. No AI-discovered material has been commercially deployed at scale yet.
The signal: The round sits in the 92nd percentile by size among 1,025 comparable Indian seed deals in the sector — a sign of investor appetite despite unproven commercialisation. Rivals like MatNexDealroom has a profile for this one. Try Dealroom →, SandboxAQ, and CuspAI are chasing similar goals, but Mohapatra argues the bottleneck isn't finding candidates. As he put it, "filtering them correctly and synthesizing them is the bottleneck."
Read more: TechCrunch
Image credit: Discovered Materials
