The new TOP500 list of the world’s most powerful supercomputers, released this morning at the ISC 2025 conference in Germany, shows an expanding European presence among the top 20 supercomputers while the top three systems, all of them American, remained the same.
The El Capitan system at the Lawrence Livermore National Laboratory in California remains the number one system. An HPE Cray EX255a supercomputer, it was measured with 1.742 exaFlop/s on the high performance LINPACK (HPL) benchmark. LLNL also submitted a measurement for the high performance gradient (HPCG) benchmark, measuring performance solving sparse linear systems using a conjugate gradient method, achieving 17.41 Petaflop/s, making it the new number one on this ranking.
El Capitan, which debuted on the list last November, has 11,039,616 cores and is based on AMD 4th generation EPYC processors with 24 cores at 1.8 GHz and AMD Instinct MI300A GPU accelerators. It uses the HPE Slingshot interconnect and achieved an energy efficiency of 60.3 gigaflops/watt. El Capitan is the 3rd system exceeding the exascale mark on the HPL benchmark.
The 65th edition of the twice-annual TOP500 showed that the Frontier system at the Oak Ridge National Laboratory is the number two system on the TOP500. Frontier has been remeasured with an HPL score of 1.353 exaFlop/s. The HPE Cray system became the first certified exascale-class system on the June 2022 TOP500 list.
At number three on the list, the Intel-HPE Aurora supercomputer at the Argonne Leadership Computing Facility was submitted with 1.012 EFlop/s on the HPL benchmark.
Beyond the top three, European systems made a strong showing in the top 20, capturing five of the remaining eight spot in the top 10 and 11 of the remaining top 20.
At number four and new on the list (the only new system in the top 10) is the JUPITER Booster system at the EuroHPC / Jülich Supercomputing Centre in Germany.
JUPITER – JU Pioneer for Innovative and Transformative Exascale Research – is expected to be the first European exascale supercomputer and has achieved a preliminary HPL value of 793.4 petaFLOP/s on a partial system. It is based on Eviden’s BullSequana XH3000 direct liquid-cooled architecture.
Summary of the Other Top 10
- Eagle the number five system is installed by Microsoft in its Azure cloud. This Microsoft NDv5 system is based on Intel Xeon Platinum 8480C processors and NVIDIA H100 GPU accelerators and achieved an HPL score of 561 Petaflop/s.
- The number six system is HPC6 and installed at Eni S.p.A center in Ferrera Erbognone in Italy. It is another HPE Cray EX235a system with 3rd Gen AMD EPYC CPUs with AMD Instinct 250X accelerators, and a Slingshot interconnect. It achieved 477.9 petaFLOP/s.
- Fugaku, the number seven system, is installed at the RIKEN Center for Computational Science (R-CCS) in Kobe, Japan. It has 7,630,848 cores achieved an HPL benchmark score of 442 Petaflop/s. It is now the second fastest system on the HPCG benchmark with 16 Teraflop/s.
- The Alps system installed at the Swiss National Supercomputing Centre (CSCS) is at number eight. It is an HPE Cray EX254n system with NVIDIA Grace 72C and NVIDIA GH200 Superchip and a Slingshot interconnect. It achieved 434.9 Petaflop/s.
- The LUMI system, another HPE Cray EX system installed at EuroHPC center at CSC in Finland, is number nine with a performance of 380 Petaflop/s and is a European High-Performance Computing Joint Undertaking (EuroHPC JU) system.
- The number 10 system Leonardo is installed at a EuroHPC site in CINECA, Italy. It is an Atos BullSequana XH2000 system with Xeon Platinum 8358 32C 2.6GHz as main processors, NVIDIA A100 SXM4 40 GB as accelerators, and Quad-rail NVIDIA HDR100 Infiniband as interconnect. It achieved a HPL performance of 241.2 Petaflop/s.
Top 10 Highlights
JUPITER Supercomputer
AMD and Intel processors are the preferred option for systems in the Top 10, according to the TOP500 organization. Five systems use AMD processors (El Capitan, Frontier, HPC6, LUMI, and Alps), while three systems use Intel (Aurora, Eagle, Leonardo). JUPITER Booster relies on an NVIDIA Grace Hopper Superchip, and Fugaku continues to use a proprietary ARM-based Fujitsu A64FX.
Seven of the computers in the Top 10 use the Slingshot interconnect (El Capitan, Frontier, Aurora, HPC6, Alps, LUMI, and JUPITER Booster), while two others use Infiniband (Eagle and Leonardo). Fugaku retains its proprietary Tofu interconnect.
While China and the United States were once again the countries that earned the most entries on the entire TOP500 list, China’s stopped submitting results for the TOP500 several years ago. The United States added two systems to the list, bringing its total number of systems to 173. China dropped in its number of machines from 63 to 46. Germany continues to close the gap, now with 43 machines on the list. By continent, North America leads with 187 systems, followed by Europe with 163 systems, and Asia with 135 systems.
GREEN500 Results
This edition of the GREEN500 saw no changes since November 2024.
JEDI again claimed the number one spot – JUPITER Exascale Development Instrument, a system from EuroHPC/FZJ in Germany. JEDI repeated its energy efficiency rating from the last list at 72.73 GFlops/Watt while producing an HPL score of 4.5 PFlop/s. JEDI is a BullSequana XH3000 machine with a Grace Hopper Superchip, an NVIDIA GH200 Superchip, Quad-Rail NVIDIA InfiniBand NDR200, and 19,584 total cores.
The ROMEO-2025 system claimed the number two spot at the ROMEO HPC Center in Champagne-Ardenne, France. This system premiered with an energy efficiency rating of 70.91 GFlops/Watt and an HPL benchmark of 9.863 PFlop/s. The architecture is identical to JEDI but is twice as large, resulting in slightly lower energy efficiency.
Number three was taken by the Adastra 2 system at the Grand Equipement National de Calcul Intensif – Centre Informatique National de Enseignement Suprieur (GENCI-CINES) in France. This is a HPE Cray EX255a system with AMD 4th Gen EPYC 24 core 1.8GHz processors, AMD Instinct MI300A accelerator, and HPE Slingshot interconnect, running RHEL. With 16,128 cores total it achieved 2.529 PFlop/s HPL performance and an efficiency of 69.1 GFlops/Watt.
El Capitan secured the number 26 spot on the GREEN500 with an energy efficiency score of 58.89 GFlops/Watt. Frontier produced an energy efficiency score of 54.98 GFlops/Watt.
HPCG Results
The TOP500 list has incorporated the High-Performance Conjugate Gradient (HPCG) benchmark results, which provide an alternative metric for assessing supercomputer performance. This score is meant to complement the HPL measurement to give a fuller understanding of the machine.
- El Capitan is the new leader on the HPCG benchmark with 17.1 HPCG-PFlop/s.
- Supercomputer Fugaku, the long-time leader, is now in second position with 16 HPCG-PFlop/s.
- Frontier remains in the third position with 14.05 HPCG-PFlop/s.
- Aurora is now in fourth position with 5.6 HPCG-PFlop/s.
HPL-MxP Results (Formerly HPL-AI)
The HPL-MxP benchmark seeks to highlight the use of mixed precision computations – 32- 16- and 8-bit.
- This year’s winner is the El Capitan system with 16.7 EFlop/s.
- The second spot was captured by Aurora with a score of 11.6 EFlop/s.
- The number three spot goes to Frontier with 11.4 EFLOP/s.
Speak Your Mind