Bristol Myers Squibb said it will expand its relationship with Nvidia by building what it calls a “most powerful” AI supercomputer system for pharma, deepening its AI infrastructure footprint. BMS plans to deploy an NVIDIA DGX SuperPOD enabled with AIDGX Vera Rubin NVL72 systems, supported by a three-year collaboration already aimed at R&D computing. BMS positioned the investment as infrastructure for hybrid intelligence, where AI co-scientists and human researchers work in tandem to increase the probability that pipeline decisions are correct. The company cited using the platform across multiple disease programs and operational workflows. The move places BMS among pharma companies increasingly tying discovery and development pipelines to large-scale compute relationships, including references to prior AI collaborations with Anthropic and Microsoft for parts of its R&D and diagnostic efforts.