Bristol Myers Squibb (NYSE: BMY) announced it will deploy an Nvidia DGX SuperPOD with DGX Vera Rubin NVL72 systems, expanding its existing compute infrastructure as part of a collaboration with Nvidia that began nearly three years ago.

According to the company, the Vera Rubin architecture delivers up to ten times greater performance per megawatt compared to its predecessor, which BMS said will allow it to run larger AI workloads without a proportional increase in energy consumption.

The infrastructure is intended to support AI-driven research across oncology, hematology, cardiovascular, immunology, and neuroscience. BMS said it will use the cluster to develop foundation models trained on its proprietary data and to power agentic workflows, including automated target identification and validation and an approach it calls "Predict First," where AI-generated predictions inform experimental design before lab work begins.

"BMS has made a deliberate bet on AI, and we are beginning to see it pay off in our pipeline and operations," said Greg Meyers, Chief Digital and Technology Officer at Bristol Myers Squibb. "Expanding our compute capabilities with Nvidia gives our researchers and teams across the business the scale they need to keep BMS at the leading edge of what AI can do for drug discovery and development."

Robert Plenge, Executive Vice President and Chief Research Officer at BMS, said the infrastructure is meant to improve decision-making quality rather than speed alone. "The goal isn't speed for its own sake; it's raising the probability that each program we advance is the right one," he said.

The agreement also includes access to BioNeMo, Nvidia's platform for biological AI, and the BioNeMo Agent Toolkit. Rory Kelleher, Senior Director of Business Development, Healthcare and Life Sciences at Nvidia, said the tools will allow BMS to convert its scientific data into proprietary intelligence to support molecule design and hypothesis evaluation.

Financial terms of the agreement were not disclosed in the press release.