Discovered Materials is leveraging AI to address the heat issues in chips running AI workloads, a problem that contributes significantly to data center energy consumption. The startup, which recently closed a $9 million seed round from Lightspeed India Partners, aims to find new materials that can improve the efficiency of integrated circuits. Founders Advaith Sridhar and Akash Ramdas are using a software pipeline that incorporates Anthropic models to generate material leads and foundational physics models to simulate and verify candidate materials. Source: techcrunch
The company has released examples of hundreds of new materials and its 'Material Discovery Bench' to track how frontier models tackle this challenge. Discovered Materials is focusing on the thermal properties of semiconductor materials, believing this is the key to success. However, the startup acknowledges the engineering trade-offs involved in manufacturing materials that may reduce heat generation or improve dissipation. “It’s a bit of playing whack-a-mole with atomic structures,” said Hemant Mohapatra, a Lightspeed partner who led the investment. Source: techcrunch
The startup has already identified several materials with properties similar to those used by major chipmakers but cannot disclose more details. Mohapatra expects the field of predicting novel substances will become commoditized as models improve, but the real challenge lies in filtering and synthesizing these materials. Discovered Materials aims to patent its materials for use in GPUs or chip manufacturing processes and hopes to have new materials ready for licensing within a year. Source: techcrunch
