Eli Lilly taps Nvidia for AI drug discovery supercomputer
Eli Lilly, Nvidia to build AI supercomputer for drug discovery
Eli Lilly will partner with Nvidia to build a Blackwell GPU-powered supercomputer to accelerate drug discovery. It is slated to go live by January.
Eli Lilly has partnered with Nvidia to build what they say will be the most powerful supercomputer run by a pharmaceutical company, aiming to speed the search for new medicines and shorten lengthy research timelines.
The Indianapolis-based drugmaker—known for the weight-loss drug Zepbound and treatments for diabetes and cancer—plans to use the Nvidia-powered system to discover novel molecules and improve other parts of its business, including clinical-trial design, manufacturing and sales operations. The companies announced the deal Tuesday but did not disclose financial terms. Lilly said some of the hardware arrived at its Indianapolis data center over the weekend and expects the system to be up and running by January.
Why it matters
Drug developers see artificial intelligence as a way to better pinpoint disease biology and identify promising drug mechanisms, potentially improving the industry’s historically low success rates. Major pharma companies like Johnson & Johnson and Roche have been expanding their AI capabilities, while startups are building businesses focused on AI-driven drug development. So far, AI tools have shown more near-term impact in tasks like trial design and administrative work than in producing new therapies that make it to market.
Key takeaways
- Lilly is targeting faster discovery of novel molecules and better clinical-trial design.
- The effort complements broader industry moves to embed AI across R&D and operations.
- Results in patient outcomes will take years, given multi-phase clinical testing.
Leadership perspective
Thomas Fuchs, Lilly’s chief AI officer, said the company is investing to bring AI’s potential to bear on difficult diseases, arguing that this is the kind of application that can materially improve health outcomes.
Nvidia’s healthcare head Kimberly Powell added that the technology won’t deliver overnight breakthroughs but could accelerate progress toward personalized medicine over the next decade.
Under the hood
Lilly’s new system will be powered by more than 1,000 of Nvidia’s Blackwell B300 graphics processing units, among the company’s most advanced AI chips. The supercomputer will operate on 100% renewable electricity within existing Lilly facilities and use chilled-water cooling.
- >1,000 Nvidia Blackwell B300 GPUs
- Runs on 100% renewable electricity
- Chilled-water cooling inside Lilly data centers
- Target go-live: by January
Strategy and data control
Fuchs said Lilly chose to build and operate the system in-house rather than rely on a third party to reduce exposure to geopolitical supply risks and to maintain tighter control over infrastructure that will incorporate decades of proprietary data.
Broader ecosystem
The partnership also extends Nvidia’s growing presence in medical research; earlier this year, the chipmaker announced collaborations with the Mayo Clinic, Illumina and other life-sciences groups. Nvidia CEO Jensen Huang has long worked with academic and industry labs to apply the company’s processors to complex scientific computing, though these efforts are not a major revenue driver.
Timeline and patient impact
Despite the investment, Lilly cautioned that benefits for patients will take time. Diogo Rau, the company’s chief information and digital officer, said any molecules discovered with the help of AI would still need to pass through multi-year clinical trials, meaning the first results could arrive later this decade or into the 2030s.
Near term
Operational gains in trial design, manufacturing, and sales analytics.
Long term
Potential acceleration toward personalized and targeted therapies.
The bottom line
If successful, the effort could make drug discovery more efficient and help bring targeted therapies to market faster, while cementing Nvidia’s role as a key technology supplier to the life-sciences sector.
