Bristol Myers Squibb expands drug discovery compute with NVIDIA Vera Rubin

Bristol Myers Squibb is deploying its second NVIDIA DGX SuperPOD, built on eight DGX Vera Rubin NVL72 systems, to expand AI access across its research organization. The company says the rack-scale systems, each combining NVIDIA Vera CPUs and Rubin GPUs, will deliver up to 10x the performance per megawatt of the infrastructure being replaced.

The new system will be combined with BMS’s existing DGX SuperPOD into a unified environment with a single data plane accessible from every BMS site globally. Researchers will use the platform for predictions, model training and agentic workflows across the drug discovery pipeline, including biological AI tools from NVIDIA BioNeMo Agent Toolkit and infrastructure management through NVIDIA Mission Control.

BMS has operated a DGX SuperPOD for about three years and is already using AI for target identification, CELMoD compound library expansion and lead optimization through a “Predict First” methodology. Executives described the goal as moving from abstract AI potential to measurable impact, with human scientific judgment augmented by more quantitative insights and shared learning across programs.