LillyPod: Pharma’s New AI Supercomputer Ushers in a New Era of Drug Discovery
Eli Lilly has launched LillyPod, the pharmaceutical industry’s most powerful supercomputer, powered by over 1,000 NVIDIA Blackwell GPUs. This new AI factory, wholly owned and operated by Lilly, aims to accelerate medical advancements and reshape the future of drug discovery, and development.
Unprecedented Computational Power
LillyPod delivers more than 9,000 petaflops of AI performance, assembled in just four months. It’s the world’s first NVIDIA DGX SuperPOD with DGX B300 systems [NVIDIA]. The system’s computational power, once requiring approximately 7 million Cray supercomputers, now fits within a single NVIDIA GPU [NVIDIA]. Lilly’s genomics team can harness 700 terabytes of data using over 290 terabytes of high-bandwidth GPU memory.
A New Scientific Instrument
Lilly leaders frame LillyPod not as traditional IT infrastructure, but as a novel research instrument. It allows scientists to explore billions of molecular hypotheses in parallel, a significant increase compared to the roughly 2,000 a productive wet-lab team can test per target per year [R&D World]. This leap in computational capability is intended to “leapfrog the whole sector” with AI [R&D World].
Applications Across the Pharmaceutical Value Chain
LillyPod will support large-scale training of protein diffusion models, small-molecule graph neural network models, and genomics foundation models. Applications span genomics, molecule design, single-cell biology, imaging, and manufacturing operations [R&D World]. Beyond drug discovery, Lilly envisions AI integration into areas like self-driving forklifts, industrial robotics, and digital twins of manufacturing lines.
NVIDIA Partnership and AI Factory Architecture
The development of LillyPod is part of an up-to-$1 billion, five-year partnership with NVIDIA [R&D World]. NVIDIA’s full-stack AI factory architecture, including accelerated computing, NVIDIA Spectrum-X Ethernet networking, and optimized AI software, provides a secure and scalable platform for the highly regulated healthcare and life sciences industries. NVIDIA Mission Control software manages the DGX SuperPOD, orchestrates workloads, monitors performance, and automates AI operations.
Lilly TuneLab: Expanding AI Access
Select models developed using LillyPod will be made available through Lilly TuneLab, an AI and machine learning platform. TuneLab provides biotech companies access to drug discovery models built on Lilly’s proprietary data, generated at a cost of over $1 billion. It uses a federated learning infrastructure built on NVIDIA FLARE, allowing companies to tap into powerful AI models while maintaining data privacy [NVIDIA].
Bridging the Gap Between Dry and Wet Labs
Historically, drug discovery has been limited by the physical constraints of laboratory experimentation. LillyPod aims to overcome this limitation by creating a computational “dry lab” at massive scale, enabling scientists to simulate and evaluate billions of molecular hypotheses before physical testing. This allows for a significant acceleration of the discovery process [R&D World].
Sustainability Commitment
Lilly aims for LillyPod’s infrastructure to run on 100% renewable electricity by 2030, utilizing efficient liquid cooling and minimizing energy impact [NVIDIA].
“This machine is exactly how AI should be used,” said Thomas Fuchs, Lilly’s chief AI officer. “It should be used for science. It should be used to lessen suffering and improve the human condition.”
Related reading