South San Francisco is poised to become an even hotter address for biotech and chip companies. Nvidia and Eli Lilly said Monday they are teaming up on a co-innovation artificial intelligence lab in the Bay Area, pledging up to $1 billion over five years to get it off the ground. The plan is to pair Nvidia engineers with Lilly scientists in a tightly linked setup where high-powered compute and wet labs constantly feed each other, with the goal of cranking out experimental data and training models that can speed up drug discovery and development. Work is expected to start in South San Francisco early this year, with the partners describing the project as a “scientist-in-the-loop” system instead of a traditional, one-and-done research facility.
The commitment, unveiled during the J.P. Morgan Healthcare Conference, will go toward talent, infrastructure and compute, according to NVIDIA. The companies expect to name a specific South San Francisco site in March, and the initiative builds on Lilly's October reveal of a Nvidia-powered “AI factory” supercomputer designed to turbocharge its drug pipeline.
How the lab will work
The new lab will sit on top of Nvidia’s BioNeMo platform and lean on the company’s next-generation Vera Rubin architecture to train models in biology and chemistry, according to NVIDIA. That stack is meant to create a tight loop where in silico experiments constantly inform real-world chemistry and biology, and vice versa. “AI is transforming every industry, and its most profound impact will be in life sciences,” Nvidia CEO Jensen Huang said, casting the co-innovation lab as a place where scientists can rapidly test ideas in silicon before they head back to the bench.
Why South San Francisco?
South San Francisco is already packed with life-sciences heavyweights, and Lilly is no stranger to the neighborhood. The drugmaker’s Gateway Labs network already operates at 681 Gateway Boulevard and 201 Haskins Way, according to Lilly Gateway Labs. Both buildings are owned by Alexandria Real Estate Equities, a setup that could streamline permitting and lab build-outs for the partners, Bisnow reports. Dropping the new AI hub into this cluster would plug Nvidia and Lilly directly into an existing pool of lab talent and early-stage startups already working on fresh drug programs.
Builds on Lilly's AI factory
The South City lab is essentially the physical-world counterpart to Lilly’s previously announced “AI factory.” In October the company laid out plans for an industry-scale Nvidia DGX SuperPOD supercomputer, powered by more than 1,000 GPUs, to train biomedical foundation models and offer select tools through its TuneLab platform, according to Lilly. Company statements say models developed on that infrastructure could be shared with partner biotechs to help design and validate new molecules faster, effectively extending the factory’s reach beyond Lilly’s own walls.
Thermo Fisher tie-in
Nvidia is also lining up the hardware and automation side of this AI push. Today, the company announced a separate collaboration with Thermo Fisher Scientific to build autonomous lab infrastructure and multi-agent workflows that can generate consistent, high-quality experimental data for model training, according to NVIDIA. Nvidia argues that automation and robotics are key to scaling the continuous lab-in-the-loop cycle that it and Lilly are betting on, where experiments run around the clock and feed straight into AI systems.
Timeline and next steps
The companies say they will hammer out site selection and staffing details over the next few months, and they still expect to reveal the exact South San Francisco location in March, according to Reuters. Lilly says models and tools from its AI factory will roll out through the TuneLab platform, giving collaborators a path into the new infrastructure. In the meantime, local life-sciences startups and real-estate players in South San Francisco will be watching closely to see which block lands the billion-dollar AI drug lab.
