SANTA CLARA — At Supercomputing 2024, AMD has continued its leadership in high-performance computing (HPC) by powering the world’s fastest supercomputer, El Capitan, which has officially surpassed the Exascale barrier. El Capitan, located at Lawrence Livermore National Laboratory (LLNL) and built by Hewlett Packard Enterprise (HPE), now holds the top position on the latest Top500 list with a staggering High-Performance Linpack (HPL) score of 1.742 exaflops. This achievement marks the second time an AMD-powered supercomputer has claimed the Exascale milestone, further solidifying the company’s role in advancing HPC capabilities.

Alongside this, both El Capitan and AMD’s Frontier system at Oak Ridge National Lab have ranked among the most energy-efficient supercomputers globally, appearing on the Green500 list at positions 18 and 22, respectively. This underscores AMD’s commitment to both performance and energy efficiency, with the combination of AMD EPYC processors and AMD Instinct GPUs driving leadership in both areas.

Forrest Norrod, executive vice president and general manager at AMD, expressed excitement over this breakthrough:
“We are thrilled to see El Capitan, powered by AMD Instinct MI300A APUs, become the fastest supercomputer in the world. This achievement showcases the extraordinary performance and efficiency of our technology and reflects the tireless collaboration between AMD, LLNL, and HPE.”

The supercomputer will play a critical role in advancing national security missions, including nuclear deterrence and nonproliferation, as well as enabling significant scientific breakthroughs. Rob Neely, director of LLNL’s Advanced Simulation and Computing program, emphasized the importance of El Capitan in supporting the National Nuclear Security Administration’s (NNSA) mission through high-fidelity simulations.

In addition to the El Capitan triumph, AMD has unveiled plans to enhance its collaboration with IBM by deploying AMD Instinct MI300X accelerators as a service on IBM Cloud. This new initiative, expected to roll out in 2025, will provide enhanced performance for generative AI models and HPC applications across enterprise cloud environments. The collaboration aims to further bolster AMD’s position in delivering cutting-edge AI and HPC capabilities to a wide range of industries.

Other key AMD-powered projects include the HPC 6 supercomputer for Eni, Italy’s energy giant, and upcoming supercomputing projects in Norway and Japan. The company’s collaboration with HPE has also helped create new supercomputing solutions designed to accelerate scientific research in fields such as climate modeling, biomedical research, and AI model development.

El Capitan and AMD’s involvement in powering both El Capitan and Frontier highlight AMD’s continuing dominance in the Exascale computing era. As the only company to provide the infrastructure for multiple Exascale systems, AMD is shaping the future of computational research, AI, and national security.

• Learn more about AMD EPYC Processors
• Learn more about AMD Instinct Accelerators