Eli Lilly and Nvidia The two companies announced Tuesday that they will partner to build the pharmaceutical industry’s “most powerful” supercomputer and so-called AI factory to accelerate drug discovery and development across the field.
It’s the latest effort by Nvidia and the pharmaceutical industry to leverage AI to accelerate time to treatment for patients while reducing costs at every stage of drug discovery and development. Diogo Lau, Eli Lilly’s chief information and digital officer, said in an interview that the process from administering a drug to the first human to bringing it to market typically takes about 10 years on average.
Eli Lilly plans to complete construction of its supercomputer and AI factory in December. It is scheduled to go online in January. But it wouldn’t be until the end of the decade that new tools would significantly benefit the company’s business, and those of other pharmaceutical companies.
“The things we’re trying to discover with this kind of power that we have now will actually benefit us in 2030,” Lau said.
Industry efforts to use AI to get medicines to people faster are still in the early stages. Although there are no drugs on the market designed using AI, progress is evident in the number of AI-discovered drugs entering clinical trials, recent AI-focused investments, and partnerships between pharmaceutical companies.
Eli Lilly owns and operates supercomputers equipped with more than 1,000 Blackwell Ultra GPUs, a new family of chips from Nvidia, connected on an integrated high-speed network. The supercomputer powers the AI Factory, a specialized computing infrastructure that develops, trains, and deploys large-scale AI models for drug discovery and development.
Supercomputers are “really novel scientific instruments. They’re like giant microscopes for biologists,” said Thomas Fuchs, Eli Lilly’s chief AI officer. “This allows us to do things on a huge scale that were previously impossible.
Scientists will now be able to train AI models on millions of experiments to test potential drugs, “drastically expanding the scope and sophistication” of drug discovery, according to an Eli Lilly release.
The focus of the new tools is not just on discovering new drugs, but Lau said there is “huge opportunity” there.
“I hope that we will be able to discover new molecules that humans would never be able to discover on their own,” he said.
Several AI models will be available on Lilly TuneLab, an AI and machine learning platform that gives biotech companies access to drug discovery models trained by Eli Lilly based on years of proprietary research. That data is worth $1 billion.
Eli Lilly launched its platform in September as a way to expand access to drug discovery tools across the sector.
“It’s really powerful to be able to give these startups another springboard, which otherwise might take years to spend money to get there,” said Kimberly Powell, Nvidia’s vice president of healthcare, adding that the company was “pleased to be a part” of the effort.
In exchange for access to AI models, biotech companies are expected to provide some of their own research and data to help train the AI models, Lau noted. The TuneLab platform employs so-called federated learning. This means businesses can leverage Lilly’s AI models without either side having to share data directly.
Eli Lilly also plans to use supercomputers to speed up new drug development and bring treatments to people faster.
The company said its new scientific AI agents can support researchers, and advanced medical imaging will give scientists a clearer picture of how diseases progress and help develop new biomarkers (measurable signs of biological processes or conditions) for personalized care.
“We really want to fulfill the promise of precision medicine,” Powell said. “Without the AI infrastructure and foundation, we’re never going to get there, right? So we’re doing all the building that we need to do, and now we’re seeing this real rise. Lilly is a perfect example of that.”
Precision medicine is an approach that tailors disease prevention and treatment to an individual’s genetic, environmental, and lifestyle differences.
