# Discovered Materials raises $9 million to find cooler semiconductor materials with AI

> Source: <https://mlq.ai/news/discovered-materials-raises-9-million-to-find-cooler-semiconductor-materials-with-ai/>
> Published: 2026-08-10 14:34:59.131208+00:00

# Discovered Materials raises $9 million to find cooler semiconductor materials with AI

- Discovered Materials raised $9 million in a seed round led by Lightspeed India Partners, with participation from Y Combinator, Peak XV Partners and angel investors including Paul Graham.
[[1]](https://techcrunch.com/2026/08/10/discovered-materials-is-playing-ai-whack-a-mole-to-hunt-cooler-chips/) - The startup uses Anthropic models in a custom agent harness, followed by physics-based simulations, to search for thermally conductive dielectric materials for 3D chip packaging.
[[1]](https://techcrunch.com/2026/08/10/discovered-materials-is-playing-ai-whack-a-mole-to-hunt-cooler-chips/)[[2]](https://discoveredmaterials.com/) - Its own benchmark found more than 500 computational candidates but only one with a plausible laboratory synthesis route, underscoring the gap between material discovery and manufacturing.
[[2]](https://discoveredmaterials.com/)

Discovered Materials has raised $9 million to use AI agents to search for materials that could help semiconductor companies manage the heat generated by advanced AI chips. Lightspeed India Partners led the seed round, which included Y Combinator, Peak XV Partners and angel investors Paul Graham, Gokul Rajaram and Thariq Shihipar. [[1]](https://techcrunch.com/2026/08/10/discovered-materials-is-playing-ai-whack-a-mole-to-hunt-cooler-chips/)[[2]](https://discoveredmaterials.com/)

The San Francisco startup was founded by Advaith Sridhar and Akash Ramdas. Ramdas holds a Stanford doctorate in materials science and has researched semiconductor materials, while Sridhar previously worked on AI agents at Persona and Luma Labs. [[1]](https://techcrunch.com/2026/08/10/discovered-materials-is-playing-ai-whack-a-mole-to-hunt-cooler-chips/)[[2]](https://discoveredmaterials.com/)

## From model outputs to lab work

Discovered Materials says its software uses Anthropic models to generate candidate materials, then applies physics models trained by the company to test whether those candidates meet targets for thermal conductivity, dielectric performance, mechanical strength and stability. The company is focused on materials for semiconductor packaging, including thermally conductive dielectrics that could help make 3D-stacked chips practical. [[1]](https://techcrunch.com/2026/08/10/discovered-materials-is-playing-ai-whack-a-mole-to-hunt-cooler-chips/)[[2]](https://discoveredmaterials.com/)

The company’s Material Discovery Bench reports that seven tested models generated more than 500 previously unknown materials that met computational targets. Only one candidate came with a synthesis recipe judged plausible enough to attempt, however. The startup says it is working to make that material in the lab. [[2]](https://discoveredmaterials.com/)

Discovered Materials plans to patent materials or manufacturing processes for use in GPUs and license the resulting intellectual property to chipmakers. Sridhar told TechCrunch that the company hopes to have materials worth patenting within the next year. It has not disclosed a named chipmaker customer, a production prototype or a commercial deployment. [[1]](https://techcrunch.com/2026/08/10/discovered-materials-is-playing-ai-whack-a-mole-to-hunt-cooler-chips/)

## Further sources

The stories that matter, in one email. Free — unsubscribe anytime.