EmbeddingGemma 2: an open, lightweight multimodal embedding model
deepmind.google·6d ago
AI Engineering
Hacker News·Aug 12·5 min read
Launch HN: Discovered Materials (YC P26) – AI agents to discover new materials
TL;DR
Discovered Materials, a startup from Y Combinator, developed AI agents to identify novel crystalline materials with specific properties for semiconductor applications. Their models propose materials that meet stringent criteria for thermal conductivity, dielectric constant, and mechanical properties, along with viable synthesis recipes.
✦ Why It Matters
Explore AI-driven material discovery to enhance your semiconductor research and development efforts today.
Key Takeaways
- Explore AI tools for discovering new materials in your projects.
- Implement a structured grading system for synthesis recipes.
- Focus on dynamic stability and thermal properties in material selection.
- Utilize machine learning models to predict material behaviors.
How It Works
The AI agents leverage machine learning interatomic potentials to predict material properties like thermal conductivity and mechanical stability. They generate random configurations to assess phonon modes and use the Boltzmann transport equation to estimate thermal conductivity.
This allows for a comprehensive evaluation of potential materials before synthesis.