NVIDIA and Eli Lilly have announced an AI co-innovation lab in the San Francisco Bay Area. Jensen Huang (NVIDIA founder and CEO) and Dave Ricks (Lilly chair and CEO) detailed the $1 billion initiative, which aims to shift drug development from "artisanal discovery" to a systematic engineering process. The lab will focus on modeling biological complexities to solve medicine's toughest challenges, such as diseases of the aging brain. The facility will utilize a "scientist-in-the-loop" framework, where autonomous systems (Agentic AI) manage wet labs synchronized with computational "dry" labs. This creates a continuous cycle where AI-driven experiments generate physical data that immediately refines new computer models. To support this, Lilly is deploying an NVIDIA DGX SuperPOD—a high-powered AI "factory"—to train foundation models (Foundation Model) specifically for biology. NVIDIA also announced updates to BioNeMo, its platform for digital biology. These include tools for predicting RNA structures and "nvMolKit," which uses graphics processors to speed up cheminformatics—the application of computer techniques to chemistry problems. Jensen Huang also honored leaders like Max Jaderberg (president of Isomorphic) for advancing protein structure prediction. This collaboration advances the field of Physical AI, applying digital intelligence to the physical world of molecules. By simulating millions of molecules digitally (in silico) before physical testing, researchers aim to "bend the arc of history" in pharmaceutical research and bring life-saving treatments to market much faster.

NVIDIA and Eli Lilly have announced an AI co-innovation lab in the San Francisco Bay Area. Jensen Huang (NVIDIA founder and CEO) and Dave Ricks (Lilly chair and CEO) detailed the $1 billion initiative, which aims to shift drug development from "artisanal discovery" to a systematic engineering process. The lab will focus on modeling biological complexities to solve medicine's toughest challenges, such as diseases of the aging brain. The facility will utilize a "scientist-in-the-loop" framework, where autonomous systems (Agentic AI) manage wet labs synchronized with computational "dry" labs. This creates a continuous cycle where AI-driven experiments generate physical data that immediately refines new computer models. To support this, Lilly is deploying an NVIDIA DGX SuperPOD—a high-powered AI "factory"—to train foundation models (Foundation Model) specifically for biology. NVIDIA also announced updates to BioNeMo, its platform for digital biology. These include tools for predicting RNA structures and "nvMolKit," which uses graphics processors to speed up cheminformatics—the application of computer techniques to chemistry problems. Jensen Huang also honored leaders like Max Jaderberg (president of Isomorphic) for advancing protein structure prediction. This collaboration advances the field of Physical AI, applying digital intelligence to the physical world of molecules. By simulating millions of molecules digitally (in silico) before physical testing, researchers aim to "bend the arc of history" in pharmaceutical research and bring life-saving treatments to market much faster.