The world’s first exascale supercomputer is the world’s fastest

The first true exascale supercomputer, Frontier, is now the fastest in the world, toppling Fugaku, who held the title for two years, according to the latest TOPP500 list of top performing supercomputers.

An exascale computer is a computer that can perform 1018 (one quintillion) floating point operations per second (1 exaFLOPS), and Frontier, housed at Oak Ridge National Laboratory in Tennessee, weighed in at 1.12 exaFLOPS.

Frontier also won the title of most energy-efficient supercomputer, generating 62.68 GFLOPs per watt.

Frontier’s speed slows Fugaku at the RIKEN Center for Computational Science (R-CCS) in Kobe, Japan from number 1 on the TOPP500 list last fall to number 2 now. Fugaku scored 442 peta FLOPS (PFLOPS) on the High Performance Linpack (HPL) benchmark, which measures the systems ability to solve a dense system of linear equations.

TOPP500 notes that Fugaku’s theoretical upper peak is greater than 1exaFLOP, but it did not demonstrate this speed in the HPL benchmark.

This is Frontier’s first entry into the top 10, and two other supercomputers, both in Europe, have also broken into the list. The LUMI system at EUROHPC/CSC in Finland weighed in at number 3, and the Adastra system at GENCI-CINES in France came in at number 10.

Here is the list of the top 10.

#1—Border

An HPE Cray EX system operated by the U.S. Department of Energy, Frontier features 3rd Gen AMD EPYC™ processors representing 8,730,112 cores that have been optimized for high performance computing (HPC) and AI with AMD accelerators Instinct™ 250X and Slingshot-11 interconnects.

#2—Fugaku

Fugaku has 7,630,848 cores and was built by Fujitsu.

#3—LUMI

LUMI is an HPE Cray EX system from CSC’s EuroHPC center in Kajaani, Finland, with a performance of 151.9 PFLOPS.

#4—Summit

An IBM system at Oak Ridge National Laboaratory in Tennessee, Summit scored 148.8 PFLOPS on the HPL benchmark. It has 4,356 nodes, each with two 22-core Power9 processors and six NVIDIA Tesla V100 GPUs, each with 80 streaming multiprocessors (SMs). The nodes are linked by a Mellanox dual-rail InfiniBand EDR network.

#5—Sierra

Similar in architecture to Summit, Sierra achieved 94.6 PFLOPS. It has 4,320 nodes with two Power9 processors and four NVIDIA Tesla V100 GPUs and is hosted at Lawrence Livermore National Laboratory, California. California, United States

#6—Sunway TaihuLight

Sunway TaihuLight is a machine developed by the National Research Center of Parallel Computer Engineering & Technology (NRCPC) in China and installed in Wuxi City. It reached 93PFLOPS on the HPL benchmark.

#7—Perlmutter

The perlmutter system is based on the HPE Cray Shasta platform and is a heterogeneous system with both AMD EPYC based nodes and 1536 Nvidia A100 accelerated nodes. It reaches 64.6 Pflop/s.

#8—Selene

Selene is an AMD EPYC-based Nvidia DGX A100 SuperPOD with Nvidia A100 for acceleration and a Mellanox HDR InfiniBand as the network. It reached 63.4 Pflop/s and is installed in-house at Nvidia’s facilities in the United States.

#9—Tianhe-2A (Milky Way-2A)

Powered by Intel Xeon processors and NUDT Matrix-2000 DSP accelerators, it has 4,981,760 cores in the system to achieve 61.4 Pflop/s. It was developed by the National University of Defense Technology (NUDT) of China and is deployed at the National Supercomputer Center in Guangzhou, China.

#10—Adastra

Adestra, installed at GENCI-CINES, is a new HPE Cray EX system and the second fastest system in Europe. It reached 46.1 Pflop/s.

Join the Network World communities on Facebook and LinkedIn to comment on the topics that matter to you.

Copyright © 2022 IDG Communications, Inc.


Source link