[JPM 2026] NVIDIA Forms AI Drug Development 'Triangular Alliance' with Lilly and Thermo Fisher

AI Drug Design to Experimentation to Retraining
Accelerating Drug Development with a Closed-Loop System
Reaching the "Transformer Moment"

NVIDIA has announced its intention to create a "closed-loop artificial intelligence (AI) system" to dramatically accelerate drug development, partnering with pharmaceutical company Eli Lilly and research and laboratory equipment company Thermo Fisher Scientific.

On January 12 (local time), at the 44th JP Morgan Healthcare Conference held at the Westin St. Francis Hotel in San Francisco, NVIDIA announced that it would jointly invest up to 1 billion dollars (approximately 1.472 trillion won) over five years with Eli Lilly to establish a joint AI innovation research center in the San Francisco Bay Area.

Kimberly Powell, Vice President of Healthcare and Life Sciences at NVIDIA, took the stage at the 44th JP Morgan Healthcare Conference. NVIDIA

This research center will leverage NVIDIA's next-generation AI chip "Vera Rubin" and DGX-based infrastructure to conduct large-scale computations from drug candidate discovery to optimization, aiming to shorten development timelines. DGX refers to a high-performance computing (HPC) system developed by NVIDIA, optimized for deep learning and AI workloads.

Kimberly Powell, Vice President of Healthcare and Life Sciences at NVIDIA, who took the stage that day, described the essence of this collaboration as "continuous learning." Powell emphasized, "Drug development is reaching a 'transformer moment'-a decisive turning point in technological paradigms. 'BioNemo,' a generative AI for drug development, converts experimental data into AI-usable data and creates a 'continuous learning cycle' where every experiment informs the next." In other words, the plan is to redesign the drug development process as a "closed-loop," where AI designs drug candidates, experiments validate them and generate data, and those results are used to further refine the process.

Powell also stated, "Researchers from NVIDIA and Eli Lilly will work side by side in the same space to jointly generate new data needed for AI model training." The specific location of the joint research center is scheduled to be announced in March this year.

NVIDIA also revealed for the first time its collaboration with Thermo Fisher Scientific on this day. By combining Thermo Fisher's expertise in research equipment and experimental manufacturing with NVIDIA's AI software, the strategy is to reduce repetitive manual tasks from experimental design to equipment operation and result interpretation, thereby increasing the speed, accuracy, and reproducibility of data generation.

If this AI-centric closed-loop structure for drug development becomes established, NVIDIA could move beyond being a mere "semiconductor supplier" to playing the role of a drug development platform. For Eli Lilly, the expectation is that reducing the time required for candidate discovery and optimization will enhance the competitiveness of its pipeline.