Hewlett Packard Enterprise (HPE) (NYSE: HPE) has been chosen by the US Department of Energy’s Oak Ridge National Laboratory (ORNL) to design and build two advanced computing systems — the next-generation exascale supercomputer “Discovery” and the dedicated AI cluster “Lux.” These systems will advance U.S. capabilities in artificial intelligence, supercomputing, and scientific research supporting energy, innovation, and national security.
Discovery, the successor to ORNL’s record-breaking “Frontier,” will be powered by the HPE Cray Supercomputing GX5000, a next-generation platform engineered for leadership-class systems. The machine will combine AI, high-performance computing (HPC), and quantum capabilities within a unified architecture. Equipped with the DAOS-based HPE Cray Supercomputing Storage Systems K3000, Discovery is expected to enhance application performance by up to ten times, accelerating breakthroughs in areas such as cancer research, precision medicine, aerospace, and clean energy.
Antonio Neri, President and CEO of HPE, remarked, “Frontier marked a historic milestone for the U.S. in supercomputing. With Discovery and Lux, we are extending that legacy, working alongside the Department of Energy, ORNL, and AMD to redefine scientific and AI innovation.”
The Lux system will serve as a dedicated AI infrastructure built with HPE ProLiant Compute XD685 servers, featuring AMD Instinct MI355X GPUs, AMD EPYC processors, and AMD Pensando networking. Fully liquid-cooled, Lux will provide a cloud-like environment that gives researchers nationwide seamless access to high-performance AI training and inference resources — effectively functioning as a sovereign AI factory for American innovation.
According to Bronson Messer, Director of Science at the Oak Ridge Leadership Computing Facility, “Discovery and Lux will dramatically expand the range and scale of research conducted at Oak Ridge. Discovery’s convergence of HPC, AI, and quantum capabilities will enable deeper insight across scientific domains, while Lux will significantly enhance access to powerful AI systems.”
Dr Lisa Su, Chair and CEO of AMD, added, “Our long-standing partnership with HPE continues to push the frontiers of high-performance computing. Together with ORNL, we’re enabling the next wave of AI-driven discoveries that will reinforce America’s leadership in science and technology.”
The Discovery supercomputer will feature next-generation AMD EPYC “Venice” processors and AMD Instinct MI430X GPUs, delivering unprecedented accuracy and performance for data-intensive workloads. With the HPE Slingshot interconnect, Discovery will achieve ultra-low latency and high bandwidth for HPC and machine learning applications.
The system’s liquid-cooled architecture will further enhance energy efficiency and reduce operational costs, setting new benchmarks for sustainable exascale computing. As a global leader in HPC and AI infrastructure, HPE continues to enable governments, research institutions, and enterprises worldwide with cutting-edge technology that transforms data into insight.
Hewlett Packard Enterprise (NYSE: HPE) is a global technology company advancing enterprise transformation through AI, cloud computing, and intelligent networking. By empowering organizations to optimize performance and unlock new opportunities, HPE helps drive innovation and sustainability across industries worldwide.
