Skip to Content

Doudna Supercomputer: Revolutionizing Science with Dell and NVIDIA

Pushing the Boundaries of Scientific Computing

Get All The Latest Research & News!

Thanks for registering!

Imagine unlocking scientific insights in weeks that once took years. That’s the promise behind the U.S. Department of Energy’s latest investment in high-performance computing: a next-generation supercomputer named Doudna

Slated for launch at Lawrence Berkeley National Laboratory’s NERSC in 2026, Doudna, powered by Dell Technologies and NVIDIA, will reshape how researchers confront some of society’s toughest challenges, from climate modeling to quantum breakthroughs.

Transformative Performance and Efficiency

At the heart of Doudna’s design lies a commitment to both performance and sustainability. This system will deliver an impressive tenfold increase in computational power over its predecessor, Perlmutter. 

It’s engineered with Dell’s advanced liquid-cooled servers and NVIDIA’s new Vera Rubin CPU-GPU platform, a combination optimized for high-performance computing, artificial intelligence, and massive-scale data analysis.

  • Ten times faster than current NERSC supercomputers
  • Direct liquid cooling slashes energy usage
  • NVIDIA Vera Rubin enables integrated, flexible CPU-GPU workflows

Enabling Rapid Discovery Across Fields

Doudna isn’t just about raw power, it’s about accelerating discovery across disciplines. Supporting over 11,000 users and 800 projects each year, it’s designed to be a “time machine for science.” 

By connecting directly to the Energy Sciences Network (ESnet), Doudna will allow seamless data streaming and near-real-time analytics for researchers nationwide. This means faster breakthroughs in areas including fusion energy, molecular science, and materials research.

  • Real-time data streaming from DOE facilities
  • Interactive computing and automated AI workflows
  • Unified platform for simulation, data analysis, and AI

Fueling Advances in AI and Quantum Research

DOE leaders view Doudna as a linchpin in maintaining U.S. leadership in technology. The supercomputer’s architecture, powered by NVIDIA GPUs, is built to seamlessly incorporate large-scale AI into scientific research. 

From advanced material discovery to biomolecular modeling and fundamental physics, Doudna will drive innovation. Its support for the NVIDIA CUDA-Q platform brings quantum simulation into the mix, offering a playground for scalable quantum algorithm development and quantum-HPC co-design.

  • AI and quantum computing as engines of discovery
  • Platform for quantum simulation and hybrid workflows
  • Adaptable infrastructure for evolving scientific needs

Robust Infrastructure for Next-Gen Research

Doudna’s infrastructure features Dell’s Integrated Rack Scalable Systems and PowerEdge servers, all accelerated by NVIDIA for both AI and compute-heavy tasks. High-speed NVIDIA Quantum-X800 InfiniBand networking ensures fast, reliable data transfer, while a high-performance parallel file system manages the vast data output of modern science. 

This robust design guarantees that Doudna is ready for complex, collaborative, and multi-institutional research at an unprecedented scale.

  • High-throughput networking for rapid data movement
  • Reliable, scalable storage for massive scientific datasets
  • Support for multi-institutional, integrated scientific workflows

Looking Ahead: A New Era for Discovery

As the world faces ever more complex scientific questions, Doudna represents a major leap forward. By combining visionary technology with strategic partnerships, the DOE is ensuring that researchers have the tools to drive transformative discoveries in energy, health, and beyond. Doudna is more than a supercomputer, it’s a launchpad for the next era of scientific innovation.

Source: Lawrence Berkeley National Laboratory News Center


Doudna Supercomputer: Revolutionizing Science with Dell and NVIDIA
Joshua Berkowitz May 31, 2025
Share this post
Sign in to leave a comment
Demystifying AI: Open-Source Circuit Tracing Tools Illuminate Neural Networks
Peering Inside the Black Box of AI