Bristol-Myers says it has become the first pharmaceutical company to deploy Nvidia’s latest DGX Vera Rubin-based SuperPOD, expanding computing capacity more than tenfold for drug discovery work.
Bristol-Myers Squibb has expanded its AI computing footprint with a new Nvidia DGX SuperPOD built on the chipmaker’s Vera Rubin architecture, becoming the first pharmaceutical company to acquire the latest-generation system for drug discovery. The deployment consists of eight DGX Vera Rubin NVL72 systems packaged as a single DGX SuperPOD, which Bristol-Myers said delivers more than a tenfold increase in computing capacity versus the AI infrastructure it installed around 2023. The company said the system is the most powerful and energy-efficient Nvidia infrastructure operating in life sciences. The upgrade adds access to Nvidia’s BioNeMo platform and Agent Toolkit, which Bristol-Myers plans to use to train proprietary foundation models and build agentic AI workflows for complex research tasks. Bristol-Myers and Nvidia have been working together for roughly three years, and the drugmaker said its earlier DGX SuperPOD deployment shortened AI-enabled target identification from weeks to much faster timeframes while reducing computing costs by 55% for some workloads. The investment underscores how pharmaceutical companies are building dedicated AI infrastructure to tackle the slow and expensive process of drug discovery, where modeling molecular interactions, screening compounds and predicting toxicity require large-scale parallel computing. Traditional drug development typically takes more than a decade and costs billions per approved therapy.