Nvidia to Use Vera Rubin Chips in AI Biotech Tieup with Eli Lilly
By Reuters | 12 Jan, 2026

The firms will jointly spend $1 billion to build a new research facility to generate data to train biotechnology AI models.

Nvidia and U.S. pharma giant Eli Lilly will spend $1 billion building a new joint research lab in the San Francisco Bay area over five years which will use Nvidia's newest generation Vera Rubin AI chips, the firms said on Monday. 

The announcement at the start of a weeklong JPMorgan Healthcare Conference in San Francisco comes just months after Zepbound maker Lilly said it is building a supercomputer using more than 1,000 of Nvidia's current generation Grace Blackwell AI chips. 

Lilly is among a raft of drugmakers that are increasingly relying on sophisticated AI models to both design and discover new treatments, with the goal of slashing the time it takes to get new offerings to market.

The pharma group and Nvidia, the world's most valuable listed firm, did not say whether Nvidia cash would flow to Lilly and be used to buy Nvidia chips, a circular arrangement that has raised questions about other Nvidia investments.

Nvidia's strategy in the biotechnology market is to supply open-source AI models and software that drugmakers can then use to build their own drug development platforms using Nvidia's hardware. 

The company on Monday released a raft of new models, including an updated one for ensuring that drugs designed with AI tools are practical to synthesize in real-world labs. 

In a press briefing, Nvidia's vice president of healthcare Kimberly Powell said both firms were dedicating "incremental resources" to a new facility whose location will be announced in March, where researchers from Nvidia and Lilly will work side by side to generate new data to train biotechnology AI models.

(Reporting by Stephen Nellis in San Francisco; Editing by Jan Harvey)