Bristol Myers Squibb (BMS), a Princeton, New Jersey-based biopharmaceutical company, announced it will expand its collaboration with NVIDIA by deploying NVIDIA’s DGX SuperPOD with DGX Vera Rubin NVL72 systems to support artificial intelligence-driven drug discovery and development.

According to the company, the expanded computing infrastructure will provide greater computational performance and energy efficiency, allowing researchers to run larger AI workloads while supporting proprietary foundation models trained on BMS’s scientific data. The company said the infrastructure will underpin AI-assisted research workflows designed to accelerate target identification, molecule design, and clinical development.

BMS said the investment builds on nearly three years of collaboration with NVIDIA, which began with deployment of DGX SuperPOD infrastructure to support research and development. AI is already being used to automate target identification and validation and to support its “Predict First” approach, in which AI-generated predictions help guide experimental design before laboratory work begins.

Robert Plenge, executive vice president and chief research officer at BMS, said in a statement that the expanded infrastructure is intended to help researchers learn from experimental and clinical data more efficiently, allowing scientists to spend more time evaluating scientific hypotheses and making research decisions.

The announcement reflects the pharmaceutical industry’s broader investment in AI infrastructure. As previously reported by Pharma Manufacturing, NVIDIA has also partnered with Eli Lilly to improve drug discovery and manufacturing operations, using virtual models to identify production bottlenecks, optimize processes, and expand manufacturing capacity for high-demand medicines including GLP-1 therapies.