The Exascale Moment Finally Arrives

pcmag.com:

There was big news on this week's update of the Top 500 list(Opens in a new window) of the world's fastest supercomputers at the International Supercomputer Conference (ISC) in Germany: The Frontier supercomputer at the Oak Ridge National Laboratory not only tops the list, but is the first machine to register one exaflop. In other words, it can perform a billion billion 64-bit floating point operations per second. We’ve been expecting this milestone for a long time,and it's good to see it finally arrive.

The Frontier system, which is based on the HPE Cray EX235a architecture, has 8,730,112 total cores based on the AMD's 3rd Generation EPYC 7A53s 64-core CPUs running at 2GHz along with AMD's Instinct 250X accelerators and the Slingshot-11 interconnect. It scored 1.102 exaflops on the High Performance Linpack (HPL) benchmark used in determining the rankings. It uses 21 megawatts of power.

For the past two years, the top spot has been held by the Fugaku system at the RIKEN Center for Computational Science (R-CCS) in Kobe, Japan. Sticking with its previous HPL benchmark score of 442 petaflops, Fugaku uses Fujitsu A64FX 48C 2.2GHz processors, with 7,630,848 cores, with a theoretical peak about one exaflop, though it hasn't scored that high on the benchmark. This has now fallen to second place, though it's still three times ahead of the number three system.

That system is also new; it's the LUMI system at the EuroHPC center at SCS in Finland, scoring 151.9 petaflops. This is part of the European High-Performance Computing Joint Undertaking, in which the various European countries are working together to create exascale machines. This is also built by HPE Cray and has a similar architecture to Frontier.

The top five are rounded