In the Top500 ranking of the most powerful systems, the United States stands at the top of the supercomputing world. Last year’s winner, Japan’s ARM A64X Fugaku system, was defeated by the Frontier system from Oak Ridge National Laboratory (ORNL), which ran on AMD EPYC CPUs. The ORNL in Tennessee is still putting it together and testing it, but the US Air Force and Department of Energy will eventually run it.
Frontier, which was powered by HPE’s Cray EX platform, was also the most powerful system by a long shot. It’s the first (known) real exascale system, with a peak Linmark performance of 1.1 exaflops.
Fugaku, on the other hand, only managed 442 petaflops, but it was enough to keep it in first place for the preceding two years.
Frontier was also the most energy-efficient supercomputer on the market. It beat out Japan’s MN-3 system for top place on the Green500 list, with a performance of 52.23 gigaflops per watt. At a news briefing, ORNL lab director Thomas Zacharia commented, “The fact that the world’s fastest machine is simultaneously the most energy efficient is just simply remarkable.”
Another HPE Cray EX system installed at EuroHPC in Finland (151.9 petaflops), the IBM-built Summit system with 22-core Power(CPUs and NVIDIA Tesla V100 GPUs (148.8 petaflops), and Lawrence Livermore’s Sierra, a smaller-scale version of Summit that hit the TOP10, are among the other machines in the TOP10.
Sunway TaihuLight from the National Research Center of Parallel Computer Engineering & Technology (NRCPC) and Tianhe-2A from China’s National University of Defense Technology were both in the top 10. (NUDT). China, on the other hand, is said to have at least two exascale systems (according to the Linmark benchmark) on the new Sunway Oceanlite and Tianhe-3 systems. However, because to the current state of semiconductor politics, China is said to be withholding any new standards or significant advancements.
