ORNL, AMD and HPE to deliver DOE’s newest AI supercomputers: Discovery and Lux

ORNL's Discovery system will be based on the new HPE Cray Supercomputing GX5000, a next generation supercomputer built for the AI era, which is powered by a combination of upcoming next generation AMD EPYC™ processors, codenamed "Venice," and AMD Instinct™ MI430X GPUs. The system will be augmented by a new Distributed Asynchronous Object Storage (DAOS)-based HPE Cray Supercomputing Storage Systems K3000, the first factory-built storage system with embedded open source software.

"AMD has a long history of collaboration with the Department of Energy and Oak Ridge National Laboratory, powering some of the world's most advanced supercomputers," said Lisa Su, chair and CEO, AMD. "Building on that legacy - from Frontier to El Capitan - Discovery represents the next major leap in AI and high-performance computing, driving breakthroughs that will strengthen U.S. leadership in science and energy."

Discovery's enhanced computing speeds and bandwidth will directly translate into accomplishing more scientific discoveries faster. With Discovery's computational engine, researchers will be able to generate and analyze data at unprecedented speeds, which will expand scientific understanding and accelerate training of AI models built for science.

Insights made possible by Discovery will help realize the full potential of many innovations, including:

• AI modeling to make nuclear energy safer, cheaper and more available
• AI-driven digital twins for precision medicine and patient health
• Aerospace design cycles shortened from years to months for faster, more fuel-efficient aircraft

Discovery will be delivered in 2028 with significantly greater performance than Frontier, ORNL's current flagship supercomputer, across every part of the system. Discovery will continue to advance scientific discoveries via high-performance modeling and simulation integrated with AI, while charting the path to the convergence of high-performance computing (HPC), AI and quantum computing.

"We're proud to continue our collaboration with ORNL and contribute to its legacy of AI and HPC excellence," said Antonio Neri, HPE CEO. "Building on the groundbreaking achievements of Frontier, these new systems will drive new levels of scientific productivity, enable complex simulations, and establish new standards for sovereign AI."

Discovery continues ORNL's and DOE's proven record of HPC leadership. The lab is home to the Oak Ridge Leadership Computing Facility, a DOE Office of Science user facility, which has deployed a total of seven flagship supercomputers since 2004. Each of the last four machines - Jaguar, Titan, Summit and Frontier - were recognized as the world's fastest system of its time. These leadership-class systems support scientists in achieving cutting-edge breakthroughs and have provided unprecedented computational science capabilities for the nation.

Lux speeds up AI deployment for national priorities

The Lux AI Cluster, to be deployed at ORNL in early 2026, will expand DOE's near-term AI capacity and accelerate progress on critical problems, including fusion, fission, materials, quantum, advanced manufacturing, and the grid. It marks the start of an ambitious scientific journey that draws on deep multi-domain expertise and unique data assets across the entire DOE national laboratory system.

Lux will employ AMD Instinct™ MI355X GPUs, AMD EPYC™ CPUs, AMD Pensando™ advanced networking, and HPE ProLiant Compute XD685 to deliver a secure, open, and efficient AI software stack for U.S. innovation, supporting large-scale AI training and distributed inference.

"We are proud to partner with the Department of Energy and Oak Ridge National Laboratory to expand America's AI infrastructure," said Su. "Lux will accelerate progress on critical U.S. research priorities - from fusion and materials to quantum, advanced manufacturing, and is a great example of public-private partnership at its best."

DOE's investment in Lux reflects the partners' trusted and proven leadership in advanced computing, including delivery of Frontier, the first supercomputer to break the exascale barrier.

"The Discovery and Lux systems will drive scientific innovation faster and farther than ever before," said Streiffer. "ORNL's leadership in supercomputing has dramatically shortened researchers' time from problem to solution across a host of fields and industries. With Discovery and Lux, the integration of high-performance computing and AI promises breakthroughs at the accelerated speed and scale necessary for continued U.S. leadership in an increasingly competitive global environment."

UT-Battelle manages ORNL for DOE's Office of Science, the single largest supporter of basic research in the physical sciences in the United States. The Office of Science is working to address some of the most pressing challenges of our time. For more information, please visit energy.gov/science.