Most  Powerful Supercomputers in the World

By Priyanka Mehta

September 28, 2022

Celebrating 97th Birthday of Seymour Cray, Father of Supercomputing

USA

Frontier 2022

Frontier is ushering in a new era of exascale computing, solving the world’s biggest scientific challenges. Also, Frontier secured first place in the Green500 list, delivering 62.68 gigaflops/watt power efficiency from a single cabinet of optimised processors.

Built by: Oak Ridge with HPE Cray and AMD Processor: AMD EPYC 64C 2GHz

Image Source: Wikipedia

Germany

JUWELS Booster Module

Meet one of the most powerful supercomputers in Europe, at 44.1 HPL petaflops. JUWELS is powered by AMD processors and NVIDIA GPUs, similar to the Selene system.

Built by: Forschungszentrum Jülich (FZJ) Processor: AMD processors and NVIDIA GPUs

Image Source: Wikipedia

Finland

Lumi

Meet one of the most powerful supercomputers in Europe, at 44.1 HPL petaflops. JUWELS is powered by AMD processors and NVIDIA GPUs, similar to the Selene system, showcases excellence and cost-efficiency.

Built by: Forschungszentrum Jülich (FZJ) Processor: AMD processors and NVIDIA GPUs

Image Source: Wikipedia

China

Sunway TaihuLight

National Supercomputing Centre in Wuxi, Sunway TaihuLight held the Number 1 spot in 2016 & 2017.

Built by: National Research Centre of Parallel Computer Engineering & Technology (NRCPC) Processor: Sunway SW26010

Image Source: Wikipedia

Japan

Fugaku

Fugaku was installed at the RIKEN Centre for Computational Science (R-CCS) in Kobe, Japan. Alongside the additional hardware, Fugaku achieved a new world record of 442 petaflops results on HPL, making it three times ahead of the number two system in the list.

Built by: Fujitsu Processor:  ARM version 8.2A

Image Source:  HPC @ LLNL

USA

Sierra

Sierra functions at an HPL mark of 94.6 petaflops. Its 4,320 nodes are equipped high-power CPUs. It has an architecture similar to supercomputer Summit. Sierra also hit 15th on the Green500 List of the world’s most energy-efficient supercomputers.

Built by: Lawrence Livermore National Laboratory (LLNL), California Processor: Two Power9 CPUs and four NVIDIA Tesla V100 GPUs

Image Source:  HPC @ LLNL

China

Tianhe-2A

With 61.4 petaflops, you can find Tianhe-2A at Guangzhou's National Supercomputer Centre. Tianhe-2A is primarily used for simulation, analysis, and government security applications.

Built by: National University of Defence Technology (NUDT) Processor: Intel Xeon CPUs and NUDT's Matrix-2000 DSP accelerators

Image Source: Worldkings

USA

Summit

Summit is the fastest system the USA currently owns. Launched in 2018, Summit has a performance of 148.8 petaflops, commonly referred to as the ‘Nobel Prize of supercomputing.’

Built by: IBM Processor: Two 22-core Power9 CPUs and six NVIDIA Tesla V100 GPUs

Image Source:  HPC @ LLNL

USA

Cray-1

Cray-1 was first released in the market, in 1975, and sold 100s of Cray-1 supercomputers. Cray-1 was the first supercomputer to implement the vector processor design successfully. Cray-1’s math operations boosted memory and registers to perform a single operation on a large data set efficiently.

Built by: Cray Research Processor: 64-bit processor @ 80 MHz

Image Source: Computer History

Happy Birthday  Seymour Cray

September 28 marks the birthday of the American electrical engineer and supercomputer architect. Cray designed a series of computers that were the fastest in the world for decades and founded Cray Research. He is credited with creating the supercomputer industry, now making Cray "the Thomas Edison of the supercomputing industry."

Image Source:  Wikipedia

Father of Supercomputing