News
Mar 9, 2026
News
Enterprise
Artificial Intelligence
Americas
NewDecoded
3 min read

Image by Nvidia
Eli Lilly and Company officially launched LillyPod this week, establishing the most powerful AI supercomputer wholly owned and operated by a pharmaceutical firm. Based in Indianapolis, the system was assembled in just four months through a strategic collaboration with NVIDIA. This massive infrastructure aims to transform how medicines are discovered and delivered to patients worldwide by supercharging scientific research. Built as the first NVIDIA DGX SuperPOD featuring DGX B300 systems, LillyPod houses 1,016 Blackwell Ultra GPUs. It delivers over 9,000 petaflops of performance, allowing researchers to process 700 terabytes of genomic data. This computational scale represents a monumental jump from traditional processing to industrial-strength AI capabilities.
The facility marks a transition toward a dry lab research model where scientists simulate billions of molecular hypotheses before physical testing begins. Traditionally, researchers were limited to analyzing a few thousand ideas per target each year due to physical laboratory constraints. LillyPod breaks these barriers, enabling parallel evaluations at an unprecedented scale to find new medical breakthroughs. Beyond discovery, the AI factory supports the entire pharmaceutical value chain, including manufacturing and clinical workflows. Lilly is utilizing the NVIDIA Omniverse platform to create digital twins of production lines to optimize supply chains and reliability. These agentic workflows help the company design better clinical trials and automate complex medical writing tasks.
The project aligns with Lilly’s goal to run on 100% renewable electricity by 2030 using efficient liquid cooling systems. It also integrates with Lilly TuneLab, a platform providing biotech partners access to proprietary models. This ecosystem approach ensures that advanced AI tools benefit the broader healthcare community while maintaining data privacy.
Lilly’s decision to own and operate its own massive supercomputing cluster signifies a definitive shift toward the AI-native pharmaceutical era. By moving away from renting cloud compute toward proprietary infrastructure, the company is insulating itself from the volatile revenue cycles typical of the industry. This investment suggests that the future of medicine belongs to firms that can combine a century of physical lab data with massive, closed-loop computational simulation. For the broader industry, this sets a new benchmark where the competitive bottleneck shifts from simply finding molecular candidates to proving their real-world safety and synthesizability at speed.
Related Articles