Shulou(Shulou.com)11/24 Report--

CTOnews.com The 62nd edition of the TOP500 supercomputer list has been revealed, Frontier still maintains its No. 1 position and is still the only exascale supercomputer in the current public model.

In addition, the top 10 landscape has changed significantly due to five new or upgraded supercomputers, with Fugaku Japan dropping to fourth place, while Aurora from Argonne National Laboratory and Eagle installed in the Microsoft Azure cloud are in second and third place.

Frontier, based at Oak Ridge National Laboratory (ORNL) in Tennessee, USA, held the lead with an HPL score of 1.194 EFlop / s. Frontier uses AMD EPYC 64C 2GHz processors based on the latest HPE Cray EX 235a architecture. The system has a total of 8,699,904 CPU and GPU cores. Frontier also has an energy efficiency rating of 52.59 GFlops / watt and can transfer data via HPE's Slingshot 11 network.

Aurora Systems, located at the Leadership Computing Center in Argonne, Illinois, USA, came in second on the list (previously occupied by Fugaku) with an HPL score of 585.34 PFlop / s. However, Aurora's value is submitted in its current state and is only half of the final size in the plan. Aurora is said to exceed Frontier when construction is complete, with peak performance of 2 EFlop / s.

Aurora is based on HPE Cray EX, Intel Xeon CPU Max series processors and Intel Data Center GPU Max series accelerators, and communicates over HPE's Slingshot-11 network.

There are 20 new supercomputing systems using Intel Sapphire Rapids CPUs in the entire list, bringing the total to 25, making Intel Sapphire Rapids CPUs the most widely used CPU products in new systems. However, of the 45 new systems currently on the list, only four use the corresponding Intel GPU, with Aurora being the largest by far.

Eagle, ranked third, is installed in Microsoft's US Azure cloud, which is also the highest ranking cloud system on the TOP 500 list. In fact, just two years ago, one of Azure's predecessor systems was in the TOP 10, but only ranked 10th. Today, the system, called Microsoft NDv5, has an HPL score of 561.2 PFlop / s and is based on Intel Xeon Platinum 8480C processors and NVIDIA H100 accelerators.

Fugaku slipped from No. 2 to No. 4, after remaining No. 1 between June 2020 and November 2021. The system, located in Kobe, Japan, has an HPL score of 442.01 PFlop / s. Of course, it is still the highest-ranked open supercomputer system outside the United States.

The LUMI system at the European High Performance Computing Centre (Euro HPC / CSC) in Kaani, Finland, finished fifth with an HPL score of 379.70 PFlop / s. The system is also Europe's largest supercomputing system and has undergone several upgrades, so it remains at the top of the list, this time up from 309.10 PFlop / s when the list was last published.

It is worth mentioning that the European High Performance Computing Joint Organization (EuroHPC JU) is pooling European resources to develop the top exascale supercomputers for processing big data. LUMI is one of the pan-European top exascale supercomputers located at the CSC data center in Kaani, Finland.

Overall, China and the United States dominate the TOP500 list, with the United States increasing its lead from 150 systems in the previous list to 161 in this list, while China dropped from 134 to 104 (CTOnews.com note: China has not submitted benchmarks for stronger systems for a long time).

In addition, the number one position on the GREEN500 remains Henri of the Flatiron Institute in New York, USA. The system has an efficiency rating of 65.40 GFlops / Watt and an HPL score of 2.88 PFlops / s. Henri is a Lenovo ThinkSystem SR670 equipped with Intel Xeon Platinum and NVIDIA H100, with a total of 8,288 cores, ranking 293rd on the TOP500 list.

Here is a summary of the top 10 TOP 500:

Frontier, ranked first, is the first system in the United States to exceed 1 Exaflop/s performance, and is located at Oak Ridge National Laboratory (ORNL) in Tennessee, USA, which uses 8,699,904 cores and currently achieves 1.194 Exaflop/s performance. It is based on the HPE Cray EX architecture and combines third-generation AMD EPYC CPUs optimized for HPC and AI, AMD Instinct 250X accelerators, and Slingshot-11 interconnects.

Aurora came in second with an HPL score of 585 Pflop/s. Installed at the Leadership Computing Center in Argonne, Illinois, USA, it is based on HPE Cray EX and uses Intel Xeon CPU Max family processors, Intel Data Center GPU Max family accelerators, and Slingshot-11 interconnects.

Eagle, ranked third, is installed in Microsoft Azure Cloud, a Microsoft NDv5 system based on Xeon Platinum 8480C processors and NVIDIA H100 accelerators, achieving an HPL score of 561 Pflop/s.

Fugaku ranked fourth, installed at the RIKEN Computational Science Center (R-CCS) in Kobe, Japan. It has 7,630,848 cores and is capable of achieving an HPL benchmark score of 442 Pflop/s.

The LUMI system is an upgraded HPE Cray EX system installed at the EuroHPC Center at the EuroCSC HPC Center in Finland and is now ranked fifth with a performance of 380 Pflop/s.

The sixth-ranked Leonardo system is installed at a EuroHPC site at CINECA in Italy. It is an Atos BullSequana XH2000 system with a Xeon Platinum 8358 32C 2.6GHz main processor, NVIDIA A100 SXM4 40GB accelerator, four-track NVIDIA HDR100 Infiniband as interconnect, Linpack performance of 238.7 Pflop/s.

Summit, built by IBM and located at Oak Ridge National Laboratory (ORNL) in Tennessee, is currently ranked seventh in the world in the HPL benchmark with performance of 148.8 Pflop/s and has 4,356 nodes, each equipped with two POWER9 CPUs (each with 22 cores) and six NVIDIA Tesla V100 GPUs (each with 80 SMs), connected via the Mellanox dual-track EDR InfiniBand network.

MareNostrum 5 ACC, ranked eighth, is a newly installed supercomputing system located at the EuroHPC /Barcelona Supercomputing Center in Spain. The BullSequana XH3000 system uses Xeon Platinum 8460Y processors along with NVIDIA H100 and Infiniband NDR200 to achieve HPL performance of 183.2 Pflop/s.

The new Eos system ranked ninth is NVIDIA's DGX SuperPOD-based system, based on NVIDIA's DGX H100 with Xeon Platinum 8480C processor, NVIDIA H100 accelerator and Infiniband NDR400, achieving 121.4 Pflop/s performance.

Sierra Systems, located at Lawrence Livermore National Laboratory in California, ranked 10th. Its architecture is very similar to Summit, the seventh-ranked system. It consists of 4,320 nodes, each equipped with two POWER9 CPUs and four NVIDIA Tesla V100 GPUs, achieving 94.6 Pflop/s performance.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.