The mission to save and improve lives just entered a new era.
This week, Eli Lilly and Company unveiled the most powerful AI factory ever fully owned and operated by a pharmaceutical company. Designed to accelerate drug discovery and development, the system — called LillyPod — represents a leap forward in computational medicine.
At its core is the world’s first NVIDIA DGX SuperPOD built with NVIDIA DGX B300 systems.
This isn’t just an infrastructure upgrade. It’s a redefinition of what’s possible in biology.
In This Content
A Supercomputer Built for Science
LillyPod is powered by 1,016 NVIDIA Blackwell Ultra GPUs, delivering more than 9,000 petaflops of AI performance. To put that in perspective, computational capacity that once required millions of legacy supercomputers now fits inside a single GPU — and LillyPod contains more than a thousand.
Even more impressive? The entire AI factory was assembled in just four months.
“It’s a big day for us,” said Diogo Rau, executive vice president and chief information and digital officer at Lilly. “But it’s also 150 years in the making. LillyPod reflects who we are — and our commitment to advancing medicine in ways never before possible.”
Powering the Next Era of Biology
Computation has become central to modern life sciences. From genomic sequencing to molecular simulation, biology today runs on data.
LillyPod enables the company’s genomics teams to analyze:
-
700 terabytes of genomic data
-
More than 290 terabytes of high-bandwidth GPU memory
“Computation is at the heart of biology,” said Thomas Fuchs, senior vice president and chief AI officer at Lilly. “Scaling compute isn’t optional — it’s essential. We’re building the computational backbone of future medicine.”
With this AI factory, Lilly can train:
-
Protein diffusion models
-
Small-molecule graph neural networks
-
Genomics foundation models
These advanced AI systems can dramatically accelerate how scientists identify therapeutic targets, design drug candidates, and predict outcomes — reducing time and uncertainty in development pipelines.
Built on NVIDIA’s Full-Stack AI Architecture
LillyPod leverages NVIDIA’s complete AI factory stack, including:
-
Accelerated computing systems
-
NVIDIA Spectrum-X networking
-
Optimized AI software
-
NVIDIA Mission Control for orchestration and monitoring
Mission Control enables Lilly to securely manage workloads, automate AI operations, and monitor performance — critical in the highly regulated pharmaceutical industry.
The system’s physical infrastructure is equally remarkable:
-
Nearly 5,000 high-speed connections
-
Over 1,000 pounds of fiber optic cabling
-
Advanced liquid cooling for energy efficiency
Designed With Sustainability in Mind
High-performance computing demands significant energy — but Lilly is planning for long-term sustainability.
The company aims to operate its AI supercomputing infrastructure on 100% renewable electricity by 2030, combining efficient liquid cooling with minimal incremental energy impact.
It’s a reminder that innovation and responsibility can evolve together.
Why This Matters
Drug discovery is traditionally slow, expensive, and complex. Each breakthrough requires years of research and billions of dollars in investment.
By deploying one of the most advanced AI infrastructures in the pharmaceutical industry, Lilly is shifting that equation.
LillyPod represents:
-
Faster hypothesis testing
-
More accurate molecular modeling
-
Scalable AI-driven experimentation
-
Reduced development timelines
Most importantly, it means the potential to bring life-saving treatments to patients sooner.
The Future of Medicine Is Computational
The unveiling of LillyPod signals something larger than a hardware milestone. It marks a turning point in how biology and AI converge.
The world’s most powerful pharmaceutical AI factory is now live.
And with it, the future of medicine just got faster, smarter — and more scalable than ever before.