Bezos-backed CuspAI launches AI Materials Foundry with Nvidia to accelerate chip material discovery

CuspAI, a British startup using AI to simulate novel materials, has raised $450 million at a $2.6 billion valuation and launched the AI Materials Foundry, a coalition of more than 45 organizations including Nvidia and Meta, to accelerate discovery of chipmaking materials and other advanced materials. The round was led by Kleiner Perkins and NEA, with significant participation from Jeff Bezos' Bezos Expeditions.

What happened

CuspAI announced the $450 million Series B funding round led by Kleiner Perkins and NEA, with Bezos Expeditions, Glade Brook Capital Partners, Lux Capital, AMD Ventures, and Britain's Sovereign AI Venture Fund also participating. The startup also unveiled the AI Materials Foundry, a network of data, labs, compute, and scientific expertise running on CuspAI's AI platform. Nvidia will supply accelerated computing infrastructure, and Meta's Fundamental AI Research Team joined the initiative. CuspAI is expanding offices and labs in Singapore, the UK, the Netherlands, Germany, Japan, and the US.

Why AI builders should care

This signals a shift of AI from software to physically grounded discovery. The coalition pools computing and scientific resources to build software that helps researchers develop new materials faster and at lower cost. For AI builders, this represents a new domain where AI models are used to simulate material performance and narrow experimental search spaces, potentially accelerating breakthroughs in semiconductors, energy, and manufacturing.

Practical implications

The AI Materials Foundry is a distributed R&D model for materials science. CuspAI's platform simulates material performance to reduce the number of laboratory tests needed. The coalition includes Nvidia for compute, Meta for AI research, and industrial partners like Hyundai. The global lab expansion indicates a multi-site approach to materials discovery. For builders, this model could be replicated for other physical science domains.

Caveats

Details on the exact AI methods, specific materials targets, and deployment benchmarks are not specified in the provided materials. The coverage centers on announcements and partnerships; technical claims about capabilities and timelines are not independently confirmed. Pricing, deployment specifics, and long-term outcomes remain uncertain.