This week at the ISC High Performance 2024 conference in Hamburg, Germany, the new TOP500 list was revealed. The big news is the announcement of a second exascale machine, the second American exascale machine. The Aurora supercomputer, located in the Argonne Leadership Computing Facility, has officially achieved exascale status, performing a billion of billion calculations per second. According to the latest TOP500 list, Aurora achieved a high-performance LINPACK (HPL) benchmark of 1.012 exaFLOPS, placing it at number 2 behind the current number 1 system, Frontier. Frontier, an AMD-HPE Cray system located at Oak Ridge National Laboratory, maintains its top position with a LINPACK score of 1.206 exaFLOPS, a title it has held since May 2022.

Aurora's Limpak benchmark improved by more than 585 petaFLOPS from November's TOP500 list, and it is just over half of its projected ultimate performance when installation and tuning at Argon are completed. The machine is based on the HPE Cray architecture, powered by Intel Xeon CPUs and Intel GPUs, previously known as Ponte Vecchio and now the Max series. This means that Hewlett Packard Enterprise has delivered the world's second exascale supercomputer as well.

Number three is the Eagle system, installed on the Microsoft Azure cloud, making it the highest-ranking cloud system on the TOP500. The ARM-based Fugaku system at Japan's RIKEN Center for Computational Science is still at number four. At number five, we have the LUMI system operated by our colleagues from CSC, Finland, which is again an HPE Cray system powered by AMD. The one new system in the top 10 at number six is Alps, the Alps machine from the Swiss National Supercomputing Center with a score of 270 petaFLOPS. The number seven system, Leonardo, is installed at Cineca in Italy. Mare Nostrum 5 at the Barcelona Supercomputing Center is at number eight. Number nine is IBM’s Summit, which has been on the list since 2018. The EOS system operated by NVIDIA is at number 10.