Lilly's LillyPod: A Quantum Leap for Pharmaceutical Research

Eli Lilly has taken a monumental step forward in its quest to save and improve lives through groundbreaking medical advancements. This week, the company unveiled what is being hailed as the world’s most powerful AI factory wholly owned by a pharmaceutical firm: LillyPod. The facility leverages NVIDIA DGX SuperPOD technology with 1,016 Blackwell Ultra GPUs, delivering over 9,000 petaflops of computational power in an impressive feat that took just four months to assemble.

From Cray Supercomputers to a Single GPU

The sheer scale and efficiency of LillyPod are nothing short of astounding. What once required the equivalent computing power of 7 million Cray supercomputers now fits into this single, state-of-the-art NVIDIA DGX system. This monumental shift in computational capacity is not just about speed; it’s a game-changer for how pharmaceutical research can be conducted.

“LillyPod represents more than just technological advancement,” said Diogo Rau, executive vice president and chief information and digital officer at Lilly. “It symbolizes our commitment to pushing the boundaries of what is possible in medical science.”

A Symbol of Purposeful Innovation

The launch of LillyPod marks a significant milestone for Eli Lilly as it aims to harness AI-driven research to accelerate drug discovery, genomics analysis, and other critical areas. The facility’s capabilities are set to support large-scale training of protein diffusion models, small-molecule graph neural network models, and genomics foundation models.

“Computation is at the heart of biology,” emphasized Thomas Fuchs, senior vice president and chief AI officer at Lilly. “Our ability to compute at scale has become a necessity for our work in drug development.”

The Future of Pharmaceutical Research

LillyPod’s infrastructure includes over 1,000 NVIDIA DGX B300 systems, enabling the genomics team to process vast amounts of data. With more than 700 terabytes of genomic data and nearly 300 terabytes of high-bandwidth GPU memory at their disposal, Lilly researchers can now tackle complex biological challenges with unprecedented precision.

“We are building the computational future of medicine,” said Fuchs. “This is not just about improving our current processes; it’s about creating new paradigms in how we approach drug discovery and treatment development.”

Achieving Unprecedented Scale and Accuracy

The implications for pharmaceutical research with LillyPod are profound. The facility will enable faster, more accurate medical advancements that could lead to breakthroughs in areas such as personalized medicine, rare disease treatments, and even pandemic response strategies.

“LillyPod is a powerful symbol of who we are and why we do this work: to make life better for people around the world,” Rau concluded. “We stand at an inflection point where technology can truly transform healthcare.”