Bristol Myers Squibb deploys Nvidia's most powerful AI system to accelerate drug discovery

Bristol Myers Squibb Builds Life Sciences' Largest AI Infrastructure

Bristol Myers Squibb announced it is deploying a DGX SuperPOD built on eight DGX Vera Rubin NVL72 systems from Nvidia AI, establishing the most powerful single-owned AI computing system in the life sciences industry 1

1

2

2

. The pharmaceutical giant becomes the first company in its sector to adopt the Vera Rubin architecture, which Nvidia unveiled earlier this year as the successor to its current generation of AI systems . Financial terms were not disclosed, but the investment builds on a smaller SuperPOD system the drugmaker purchased approximately three years ago, which is now two or three generations behind Vera Rubin 1

1

.

Strategic Shift: Owning Compute Over Renting

The decision to purchase rather than rent AI infrastructure reflects a calculated approach by Bristol Myers Squibb and other pharmaceutical companies including Eli Lilly and Roche, which made similar investments in October 2025 and March 2026 respectively 3

3

. When the BMS team analyzed the economics, they found that even when reserving rented capacity, they would access chips three to four generations old at significantly higher costs than purchasing outright 3

3

. The ownership model also ensures that proprietary data and intellectual property remain on compute infrastructure the company controls 3

3

. Greg Meyers, chief digital and technology officer at Bristol Myers Squibb, emphasized that the new system delivers 10 times more compute capacity per watt spent, addressing rising electricity costs 1

1

. The supercomputer is expected to be online by January 2027 and will be hosted by data center company Equinix under a co-location agreement 3

3

.

Accelerating Drug Discovery With Measurable Impact

Bristol Myers Squibb is already demonstrating concrete results from AI-driven drug research. The company currently uses AI tools to cut the time to make medicines for clinical trials by 20% to 30%, with potential to reach 50% in coming years, according to Robert Plenge, chief research officer 1

1

4

4

. Plenge noted that the new capabilities would allow the company to cycle through many more potential drug candidates early in development: "Maybe before we could do 10 and now we can do dozens" 1

1

. One experimental sickle cell disease treatment currently in early clinical development would likely not have been discovered without AI-enabled research 1

1

. The company has used AI to expand its library of CELMoD compounds—molecules engineered to selectively degrade cancer-causing proteins with applications in blood cancer treatment and beyond 2

2

.

Unified Platform Powers Research Across Therapeutic Areas

The expanded AI infrastructure will support research across oncology, cardiovascular disease, immunology, hematology, and neuroscience 5

5

. Researchers will gain access to NVIDIA BioNeMo Agent Toolkit for biological AI, enabling them to run predictions, train foundation models, and power agentic workflows across the full drug discovery pipeline 2

2

5

5

. Payal Sheth, senior vice president of therapeutic discovery sciences, explained that the company employs a "Predict First" methodology, using AI predictions to prioritize synthesis of molecules with multi-parameter optimization and weed out candidates that wouldn't meet required properties 2

2

. This approach ensures laboratory experiments focus on molecules with the highest probability of success in clinical trials 2

2

. The company now uses AI in all of its small-molecule and most of its large-molecule programs 1

1

4

4

.

Democratizing Access to Computing Power

Erin Davis, vice president of research business insights and technology at BMS, calls the new deployment the "SuperDuperPOD," reflecting both its scale and ambition 2

2

. The team is combining the existing DGX SuperPOD and the new DGX Vera Rubin NVL72-powered system into a unified environment accessible from every BMS site globally 2

2

. "Instead of equipping a small group of researchers with access to the supercomputer, we're opening it up to literally every scientist," Davis said. "No one has to wait, and no one is told they have a limit" 2

2

. Barriers from past acquisitions and the need for deep computational expertise are being replaced with AI-native tooling managed through NVIDIA Mission Control, allowing researchers to initiate complex predictions in plain English 2

2

. The deployment extends a strategic partnership with Anthropic announced in May, which gives more than 30,000 employees access to institutional knowledge through Anthropic's AI models 5

5

. Meyers emphasized the company's commitment: "BMS has made a deliberate bet on AI, and we are beginning to see it pay off in our pipeline and operations" 3

3

.