Fugaku retains title as world's fastest supercomputer

The supercomputer Fugaku, which is being developed jointly by RIKEN and Fujitsu based on Arm technology, has once again taken out the top spot on the TOP500 list — a ranking of the world’s fastest supercomputers that is published twice a year.

Currently installed at the RIKEN Center for Computational Science (R-CCS) in Kobe, Fugaku is being developed under a national plan to design Japan’s next-generation flagship supercomputer and to carry out a wide range of applications that will address high-priority social and scientific issues. Scheduled to begin full operation in fiscal 2021, it will be put to use in applications aimed at achieving the Society 5.0 plan, by running applications in areas such as drug discovery; personalised and preventative medicine; simulations of natural disasters; weather and climate forecasting; energy creation, storage and use; development of clean energy; new material development; new design and production processes; and elucidation of the fundamental laws and evolution of the universe.

In addition to the TOP500 list, Fugaku also took first place on the HPCG benchmark, a ranking of supercomputers running real-world applications; the HPL-AI benchmark, which ranks supercomputers based on their performance capabilities for tasks typically used in artificial intelligence applications; and Graph 500, which ranks systems based on data-intensive loads. The results were achieved with Fugaku’s full complement of 158,976 nodes fit into 432 racks.

On the TOP500, the supercomputer achieved a High-Performance Linpack (HPL) score of 442.01 petaflops, an improvement over the 415.53 petaflops set in June. On HPCG, it scored 16 petaflops, up from 13.40 last time, and on HPL-AI it gained a score of 2 exaflops, over 1.42 in June. Its top ranking on Graph 500 was won by a collaboration involving RIKEN, Kyushu University, Fixstars and Fujitsu, with the supercomputer solving a breadth-first search of an enormous graph with 1.1 trillion nodes and 17.6 trillion edges in approximately 0.25 seconds, earning it a score of 102,955 gigaTEPS — a major advance over its score of 70,980 gigaTEPS in June and more than four times the score attained by its nearest competitor, China’s Sunway TaihuLight, which performed at 23,756 gigaTEPS.

“We are thrilled that we were able to take the top spot on the major benchmarks for the second consecutive term,” said Naoki Shinjo, Corporate Executive Officer at Fujitsu. “This time, we were able to use Fugaku in its completed form and were able to make significant improvements upon the performance we attained in June, to prove the system’s high potential. In the future, we anticipate that Fugaku will demonstrate high application performance and will be widely used as a supercomputer contributing to the realisation of Society 5.0. We would like to once again express our sincere gratitude to RIKEN and others for their great cooperation and support.”

The top 10 systems on the TOP500 list are as follows:

1. Fugaku remains at the top spot, growing its Arm A64FX capacity from 7,299,072 cores to 7,630,848 cores. The additional hardware enabled its new world record 442 petaflops result on HPL. This puts it three times ahead of the number two system in the list.
2. Summit, an IBM-built system at the Oak Ridge National Laboratory (ORNL) in Tennessee, remains the fastest system in the US with a performance of 148.8 petaflops. Summit has 4356 nodes, each one housing two 22-core Power9 CPUs and six NVIDIA Tesla V100 GPUs.
3. Sierra, a system at the Lawrence Livermore National Laboratory in California, has an HPL mark of 94.6 petaflops. Its architecture is very similar to that of Summit, with each of its 4320 nodes equipped with two Power9 CPUs and four NVIDIA Tesla V100 GPUs.
4. Sunway TaihuLight, a system developed by China’s National Research Center of Parallel Computer Engineering & Technology (NRCPC) and installed at the National Supercomputing Center, is powered by Sunway SW26010 processors and achieves 93 petaflops on HPL.
5. Selene, an NVIDIA DGX A100 SuperPOD installed in-house at NVIDIA Corp, was listed as number seven in June but has doubled in size, allowing it to move up the list. The system is based on AMD EPYC processors with NVIDIA’s new A100 GPUs for acceleration. Selene achieved 63.4 petaflops on HPL as a result of the upgrade.
6. Tianhe-2A (Milky Way-2A), a system developed by China’s National University of Defense Technology (NUDT) and deployed at the National Supercomputer Center, is powered by Intel Xeon CPUs and NUDT’s Matrix-2000 DSP accelerators and achieves 61.4 petaflops on HPL.
7. A new supercomputer, known as the JUWELS Booster Module, debuts at number seven on the list and was recently installed at the Forschungszentrum Jülich (FZJ) in Germany. It is part of a modular system architecture but running by itself was able to achieve 44.1 HPL petaflops, which makes it the most powerful system in Europe.
8. HPC5, a Dell PowerEdge system installed by the Italian company Eni S.p.A., achieves a performance of 35.5 petaflops using Intel Xeon Gold CPUs and NVIDIA Tesla V100 GPUs. It is the most powerful system in the list used for commercial purposes at a customer site.
9. Frontera, a Dell C6420 system that was installed at the Texas Advanced Computing Center of the University of Texas last year, achieves 23.5 petaflops using 448,448 of its Intel Platinum Xeon cores.
10. Dammam-7, installed at Saudi Aramco in Saudi Arabia, is the second commercial supercomputer in the current top 10. The HPE Cray CS-Storm system uses Intel Gold Xeon CPUs and NVIDIA Tesla V100 GPUs. It reached 22.4 petaflops on the HPL benchmark.

The awards were announced on 16 November at the SC20 High-Performance Computing Conference, held as an online event. The full list is available here.

Top image: The Japanese supercomputer Fugaku.

Please follow us and share on Twitter and Facebook. You can also subscribe for FREE to our weekly newsletter and bimonthly magazine.