Discovered Materials Uses AI Agents to Discover New Cooling Materials for Hotter Chips
Summarized by AI from reporting by Hacker News AI, published under our editorial policy.
Discovered Materials, a Y Combinator startup, uses AI agents to discover new materials for cooling semiconductor chips. This could help reduce the massive power and water consumption of data centers as chips like Nvidia's Rubin reach 2.3 kW of heat.
Key takeaways
- Discovered Materials uses AI agents to discover new materials for cooling semiconductor chips.
- Nvidia's upcoming Rubin chip is expected to generate 2.3 kilowatts of heat, up from 700 watts in the H100 chip.
- Better cooling materials could reduce the power and water consumption of data centers.
Discovered Materials, a Y Combinator-backed startup, launched an AI-powered platform to discover new materials that can better cool semiconductor chips. The company's AI agents analyze vast datasets to identify materials that can dissipate heat more efficiently, addressing a critical challenge in the semiconductor industry.
The Heat Problem in Semiconductors
GPUs and other semiconductor chips are getting hotter with each new generation. Nvidia's H100 chip, released in 2022, has a thermal design power (TDP) of 700 watts. The Blackwell chip, released in 2024, pushes out 1.2 kilowatts of heat, and the upcoming Rubin chip is expected to reach 2.3 kilowatts. This trend is driven by the increasing power and performance demands of modern chips, but it also poses significant challenges for data centers, which need to dissipate this heat to keep chips running efficiently.
How Discovered Materials' AI Agents Work
Discovered Materials' AI agents sift through vast amounts of data to identify materials with superior thermal properties. The agents use machine learning algorithms to predict how different materials will perform under various conditions, allowing the company to quickly narrow down promising candidates for further testing and development. This approach is much faster and more cost-effective than traditional materials discovery methods, which rely heavily on trial-and-error experimentation.
Why This Matters for Everyone
The heat generated by semiconductor chips is a major contributor to the power and water consumption of data centers. These facilities require massive amounts of energy to keep chips cool, and the trend toward hotter chips is making this problem worse. By developing more efficient cooling materials, Discovered Materials could help reduce the environmental impact of data centers and make them more sustainable. This, in turn, could lead to lower energy costs for cloud services and other data-intensive applications that we use every day.
What You Can Do Today
While Discovered Materials' technology is primarily targeted at the semiconductor industry, you can stay informed about their progress and the broader efforts to improve chip cooling. Visit their website at https://discoveredmaterials.com/ to learn more about their research and how they're working to address this critical challenge. You can also follow their updates on Hacker News and other tech forums to stay up-to-date on the latest developments.
Frequently asked
- Is Discovered Materials' technology available for commercial use?
- The source does not provide information on the commercial availability of Discovered Materials' technology.
- How does Discovered Materials' AI compare to traditional materials discovery methods?
- Discovered Materials' AI agents are much faster and more cost-effective than traditional trial-and-error methods.