Summary:
- The United States Department of Energy announces the development of the Doudna supercomputer, set to revolutionize American science and innovation.
- Powered by NVIDIA’s Vera Rubin platform and Dell’s latest infrastructure, Doudna promises unmatched performance for a wide range of scientific applications.
- The supercomputer will empower breakthroughs in AI, quantum computing, and fusion energy research, fostering collaboration and driving the future of scientific computing.
Article:
In a groundbreaking move to elevate American science and innovation to new heights, the United States Department of Energy (DOE) has unveiled plans for the Doudna supercomputer. Scheduled for deployment in 2026 at the National Energy Research Scientific Computing Center (NERSC) at Lawrence Berkeley National Laboratory, the Doudna supercomputer is poised to redefine the boundaries of scientific discovery.
Named after Nobel Prize-winning biochemist Jennifer Doudna, co-inventor of the revolutionary CRISPR gene-editing technology, the Doudna supercomputer represents a fusion of cutting-edge computational power and groundbreaking biological research. This next-generation flagship system will succeed the Perlmutter supercomputer, offering over ten times the performance and enabling researchers to conduct vastly more complex simulations and analyses in significantly less time.
Powered by NVIDIA’s next-generation Vera Rubin platform and built on Dell’s latest liquid-cooled ORv3 server infrastructure, the Doudna supercomputer is designed to support the DOE’s most computationally demanding workloads, from molecular dynamics and high-energy physics to large-scale AI training. This integration of simulation, data analysis, and AI into a single, high-efficiency environment signifies a significant step forward in multidisciplinary science.
The infrastructure of the Doudna supercomputer combines Dell Integrated Rack Scalable Systems with PowerEdge servers and NVIDIA accelerators, optimizing performance for both AI-centric and traditional computing tasks. High-speed data movement is facilitated by the NVIDIA Quantum-X800 InfiniBand network, ensuring seamless communication across thousands of compute nodes. With storage solutions featuring a high-performance parallel file system and smart data tiering, researchers will have real-time access to massive datasets to fuel their scientific endeavors.
One of the standout capabilities of the Doudna supercomputer lies in its ability to support AI-driven science at scale. Researchers can integrate machine learning models into simulation workflows, improving accuracy and reducing computation time in fields such as materials discovery, climate modeling, and biomolecular design. Additionally, the system will advance quantum computing research, offering tools like NVIDIA’s CUDA-Q platform for simulating and developing quantum algorithms, paving the way for future hybrid quantum-HPC systems.
Through its integration with the Energy Sciences Network (ESnet), the Doudna supercomputer will enable researchers across the US to stream experimental and observational data directly into the system from DOE user facilities nationwide. This real-time data analysis capability enhances responsiveness to new scientific insights and experimental results, supporting automated, AI-driven workflows and interactive computing environments tailored for data-intensive science.
The launch of the Doudna supercomputer signals a pivotal moment in the evolution of US scientific infrastructure, combining the strengths of Dell Technologies, NVIDIA, and the DOE to propel world-changing discoveries. Whether unraveling the mysteries of the universe or pioneering next-generation energy solutions, Doudna will empower researchers to achieve more, faster, and with greater precision than ever before.
