It collaborated with Nvidia in the construction of one of the world’s most powerful AI supercomputers for drug discovery and development.
The DGX SuperPOD system will contain more than 1,000 Nvidia Blackwell Ultra GPUs and will function like an “AI factory,” accelerating every step of data analysis, model training, and molecular design.
The collaboration seeks to accelerate pharmaceutical innovation while infusing the latest computing technologies in every aspect of Lilly’s research and manufacturing pipeline.
For nearly 150 years, Eli Lilly has led a wide array of breakthrough therapies in global healthcare, ranging from diabetes, oncology, immunology, and neuroscience. The company well understands that with change coming to the pharmaceutical landscape, the future of drug discovery will increasingly depend upon computational power and AI.
Traditional drug discovery is a notoriously slow and expensive process, with each new therapy taking years of laboratory testing, data analysis, and clinical validation. The collaboration with Nvidia is designed to overcome those bottlenecks using high-performance AI computing to execute predictive modeling and generative drug design at an unprecedented scale.
At the heart of that collaboration is the Nvidia DGX SuperPOD platform, powered by its latest Blackwell Ultra GPUs. The architecture focuses on large-scale AI tasks such as complex biological data analysis, like protein folding and genomic sequencing, including in real-time molecular behavior. With more than 9,000 petaflops of AI performance, Lilly’s system is among the most capable computing infrastructures in life sciences to date.
Another critical building block in Lilly’s AI-enabled ecosystem is its proprietary TuneLab platform, powered by the technology of federated learning. Biotech partners and research organizations can work with Lilly’s models using TuneLab-with collaboration but without ever sharing sensitive data. Such a setting enhances scientific collaboration while keeping the data secure-a major step toward open and protected innovation in drug research. The company wants to extend this computing power well beyond the early phases of drug discovery, using the supercomputer to fine-tune clinical trials, analyze patient imaging for personalized medicine, and refine manufacturing through predictive analytics and digital twins-virtual models of physical production systems that find and fix inefficiencies.
Lilly’s move also complements its wider environmental and operational strategy. The supercomputer will be fully powered by renewable energy, while advanced water-based systems will cool it to keep energy consumption at a minimum. These actions are part of the company’s pledge for carbon neutrality no later than 2030. A deal of this nature represents a paradigm shift for both the pharmaceutical and technology sectors. Merging Lilly’s biological acumen with Nvidia’s AI and computing capabilities redefines how technology can accelerate scientific discovery. In a world where competitive advantage is defined by speed to market and speed of innovation, the Lilly-Nvidia alliance shows us how AI-powered infrastructure will shape the future of medicine by condensing years into months and bringing new hope to patients faster than ever before.
