Eli Lilly partners with NVIDIA to build the strongest supercomputer and AI factory in the pharmaceutical industry: accelerating drug development and discovering molecules that humans cannot find

Wallstreetcn
2025.10.28 20:37
portai
I'm PortAI, I can summarize articles.

Eli Lilly collaborates with NVIDIA to build a supercomputer and AI factory composed of over 1,000 Blackwell Ultra GPUs, expected to go live in January next year. This system will accelerate drug development and support large-scale AI model training. Eli Lilly's Chief Information and Digital Officer Diogo Rau stated, "We hope to discover new molecules that could never be found by humans alone."

Eli Lilly and NVIDIA announced a collaboration to build the "most powerful" supercomputer and AI factory in the pharmaceutical industry, aimed at accelerating drug development across the industry.

On Tuesday, the two companies announced that Eli Lilly expects to complete the construction of the supercomputer and AI factory by December and to launch operations in January next year. This system will consist of over 1,000 NVIDIA Blackwell Ultra GPU chips connected through a unified high-speed network.

The supercomputer will power the AI factory, which is specifically designed as a computational infrastructure for large-scale development, training, and deployment of AI models for drug research.

However, Eli Lilly's Chief Information and Digital Officer Diogo Rau stated that these new tools may not bring significant returns to Eli Lilly and other pharmaceutical companies in the short term. Rau said:

The benefits of the computational discoveries we are discussing now will not be realized until 2030.

AI in Pharmaceuticals Still in Early Stages

Efforts to leverage AI to accelerate drug market entry in the pharmaceutical industry are still in the early stages.

Currently, no drugs designed using AI have been brought to market, but progress is reflected in the increasing number of AI-discovered drugs entering clinical trials, as well as recent investments and partnerships focused on AI by pharmaceutical companies.

Eli Lilly's Chief AI Officer Thomas Fuchs stated:

This supercomputer is a truly novel scientific instrument, akin to a giant microscope for biologists.

Fuchs emphasized:

Scientists will be able to train AI models through millions of experiments to test potential drugs, significantly expanding the scope and complexity of drug discovery.

Rau pointed out that while finding new drugs is not the sole focus of these new tools, "this is where the biggest opportunity lies." He said:

We hope to discover new molecules that could never be found by humans alone.

Precision Medicine Goals Require AI Infrastructure

Eli Lilly also plans to use the supercomputer to shorten drug development cycles, helping treatments take effect more quickly.

The company stated that new scientific AI agents can support researchers, and advanced medical imaging can allow scientists to observe disease progression more clearly and help develop biomarkers for precision treatment.

Precision medicine is an approach to disease prevention and treatment tailored to individual differences in genes, environments, and lifestyles.

NVIDIA's Vice President of Healthcare Kimberly Powell said:

We hope to fulfill the promise of precision medicine; without AI infrastructure, we will never achieve this goal. So we are undertaking all necessary construction, and we will see the technology take off, and Eli Lilly is a prime example.

Open Platform for Sharing R&D Data

Multiple AI models will be available on Eli Lilly's Lilly TuneLab platform launched last September.

This is an AI and machine learning platform that allows biotechnology companies to access Eli Lilly's drug discovery models trained on years of proprietary research. This data is valued at $1 billion.

Eli Lilly launched this platform to expand access to drug discovery tools across the industry. Kimberly Powell stated:

It is very meaningful to be able to help these startups; otherwise, they might take years to consume funds to reach that stage.

She also added that the company is "very pleased to be involved" in this work. In exchange, the biotechnology company needs to contribute part of its own research and data to help train the AI model