Bristol Myers Squibb Expands AI Research with NVIDIA Vera Rubin SuperPOD

Bristol Myers Squibb said in a press release that it will deploy the latest NVIDIA DGX Vera Rubin NVL72 systems to expand its computing infrastructure for drug discovery and development. The installation will form what the company describes as the most powerful and energy efficient NVIDIA setup in the life sciences sector, offering up to ten times higher performance per megawatt than previous systems.

The new cluster will support BMS’s proprietary AI models and agentic workflows under a unified computational backbone. These systems will enable researchers to manage more complex workloads without a proportional rise in energy use. The initiative extends a collaboration that began nearly three years ago when BMS first adopted NVIDIA’s SuperPOD technology.

In a company blog post, NVIDIA said the expanded deployment consists of eight DGX Vera Rubin NVL72 systems that will provide access to a unified AI platform and the BioNeMo Agent Toolkit for biological research. The infrastructure will be shared across BMS sites globally, facilitating predictions, model training, and automation in screening and molecular design.

BMS executives said the new capabilities will allow scientists to accelerate discovery processes and prioritize the synthesis of molecules more effectively, with AI informing both small and large molecule programs across disease areas including oncology, immunology, and neuroscience.