The El Capitan supercomputer at the Lawrence Livermore National Laboratory (LLNL) has clinched the title of the world’s highest-performing supercomputer with a staggering 1.742 Exaflop/s. This powerhouse, announced in 2019 by Cray before its acquisition by HPE, features AMD Epyc Gen 4 24-core CPUs and AMD Instinct MI300A accelerators with a combined total of 11,039,616 cores.
In second place is Frontier, the former champion since 2022 with a performance of 1.353 Exaflops. This supercomputer, housed at Oak Ridge National Laboratory (ORNL), utilizes Epyc Gen 3 CPUs and Instinct MI250X accelerators.
Coming in at third place is Aurora, another supercomputer belonging to the US Department of Energy, boasting a performance of 1.012 exaflops. Powered by Intel Xeon CPU Max and Data Center GPU Max, this machine resides at the Argonne Leadership Computing Facility.
All three supercomputers surpass the 1 Exaflop/s mark and are manufactured by HPE/Cray. With AMD supplying two machines, and Intel contributing one, the US Department of Energy is leading the charge in supercomputing technology.
Furthermore, the newly charted HPC6 at Eni S.p.A research center in Italy, a sibling to Frontier, showcases the AMD Epic Gen 3 + MI250X configuration, also built by HPE/Cray. This supercomputer boasts a formidable performance of 477.9 Petaflop/s, ranking fifth on the charts and standing as the most powerful supercomputer in Europe.
For more rankings, refer to the TOP500 list. Stay tuned for the upcoming unveiling of the Discovery supercomputer, slated for completion in 2027 or early 2028 at ORNL, alongside Frontier.
TLDR: El Capitan, Frontier, and Aurora lead the pack in supercomputing with powerhouse performances and cutting-edge technology, setting the stage for future advancements in the field.
Leave a Comment