BFS results, June 2024
Rank | previous rank | Machine | Vendor | Type | Network | Installation Site | Location | Country | Year | Application | Usage | Number of nodes | Number of cores | Memory | Implementation | Scale | GTEPS | C_time | Power |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1 | new | Supercomputer Fugaku | Fujitsu | Fujitsu A64FX | Tofu Interconnect D | RIKEN Center for Computational Science (R-CCS) | Kobe Hyogo | Japan | 2020 | Various scientific and industrial fields | Academic and industry | 152064 | 7299072 | 4866048 | Custom | 42 | 166029 | 4350.38 | |
2 | 2 | Wuhan Supercomputer | HUST | Kunpeng 920+Tesla A100 | Custom | Wuhan Supercomputing Center | Wuhan | China | 2023 | Research | Academic | 252 | 6999552 | 584640 | Custom (Graph Processing System "YITU" implemented by Jin Zhao Yu Zhang et al.) | 41 | 115357.6 | 16284.8 | |
3 | 3 | Frontier | HPE | HPE Cray EX235a | Slingshot-11 | DOE/SC/Oak Ridge National Laboratory | Oak Ridge TN | United States | 2021 | government | open research | 9248 | 8730112 | 9469952 | Fugaku Derivative | 40 | 29654.6 | 29.7099 | NA |
4 | 4 | Pengcheng Cloudbrain-II | HUST-Pengcheng Lab-HUAWEI | Kunpeng 920+Ascend 910 | Custom | Pengcheng Lab | ShenZhen | China | 2022 | Research | Academic | 488 | 93696 | 999424 | Custom (Graph Processing System "YITU" implemented by Jin Zhao Yujian Liao Yu Zhang et al.) | 40 | 28463.1 | 57218.3 | |
5 | new | Aurora | Intel/HPE | HPE Cray EX - Intel Exascale Compute Blade | HPE Slingshot-11 | DOE/SC/Argonne National Laboratory | Argonne IL | United States | 2023 | Open research | government | 4096 | 25591808 | 4718592 | Custom "Fugaku Derivative" | 40 | 24250.2 | 240.394 | |
6 | 5 | Sunway TaihuLight | NRCPC | Sunway MPP | Sunway | National Supercomputing Center in Wuxi | Wuxi | China | 2015 | research | research | 40768 | 10599680 | 1304580 gigabytes | Custom | 40 | 23755.7 | 961.455 seconds | |
7 | 6 | Wisteria/BDEC-01 (Odyssey) | Fujitsu | PRIMEHPC FX1000 | Tofu interconnect D | Information Technology Center The University of Tokyo | Kashiwa Chiba | Japan | 2021 | University | Research | 7680 | 368640 | 245760 | Custom | 37 | 16118 | 90.7395 | |
8 | new | MareNostrum 5 ACC | Eviden | BullSequana XH3000 XEON PLATINUM 8460Y+ 40C 2.3GHZ NVIDIA H100 64GB INFINIBAND NDR200 | NDR200 | Barcelona Supercomputing Center | Barcelona | Spain | 2024 | Research and innovation | all open R&D | 1120 | 680960 | 573400 | BFS2D-NVIDIA | 35 | 15737.43 | 86.65 | |
9 | 7 | TOKI-SORA | Fujitsu | PRIMEHPC FX1000 | Tofu interconnect D | Japan Aerospace eXploration Agency (JAXA) | Tokyo | Japan | 2020 | Research | CFD | 5760 | 276480 | 184320 | Custom | 36 | 10813 | 70.5226 | |
10 | 8 | NAPS-FX1000 | Fujitsu | PRIMEHPC FX1000 | Tofu interconnect D | Japan Meteorological Agency | Tokyo | Japan | 2022 | Research | Weather and Climate Research | 4608 | 221184 | 147456 | Custom | 36 | 10158 | 78.5127 | |
11 | 9 | LUMI-C | HPE | HPE Cray EX | HPE Slingshot-10 | EuroHPC/CSC | Kajaani | Finland | 2021 | Research | Various | 1492 | 190976 | 381952 | custom | 38 | 8467.71 | 434.819 | |
12 | 10 | OLCF Summit (CPU-Only) | IBM | IBM POWER9 | Oak Ridge National Laboratory | Oak Ridge TN | United States | 2018 | Government | Scientific Research | 2048 | 86016 | 1048576 | Custom | 40 | 7665.7 | 370.647 | ||
13 | 11 | SuperMUC-NG | Lenovo | ThinkSystem SD530 Xeon Platinum 8174 24C 3.1GHz Intel Omni-Path | Leibniz Rechenzentrum | Garching | Germany | 2018 | Academic | Research | 4096 | 196608 | 393216 | custom-aml-heavy-diropt | 39 | 6279.47 | 1547.10 | ||
14 | 12 | Lise | Atos | Bull Intel Cluster Intel Xeon Platinum 9242 48C 2.3GHz Intel Omni-Path | Intel Omni-Path | Zuse Institute Berlin (ZIB) | Berlin | Germany | 2019 | Research | Academic | 1270 | 121920 | 502272 | custom the same code used on Fugaku | 38 | 5423.94 | 228.899 | |
15 | 13 | NERSC Cori - 1024 haswell partition | Cray | XC40 | Aries | NERSC/LBNL | DOE/SC/LBNL/NERSC | United States | 2017 | Government | basic science and simulation | 1024 | 32768 | 133376 | custom-aml-heavy-diropt | 37 | 2562.16 | 1369.29 | |
16 | 14 | Tianhe-2 (MilkyWay-2) | National University of Defense Technology | MPP | TH Express-2 | Changsha China | Changsha China | China | 2013 | Science research Big data | Research | 8192 | 196608 | 524288 gigabytes | custom | 36 | 2061.48 | 456.071 seconds | 0 Watts |
17 | 15 | RAZIEL | Technion | RAZIEL | custom | Haifa | Haifa | Israel | 2022 | Academic | Academic | 1 | 16 | 1024 | custom | 28 | 1679.21 | 107.872 | 76 |
18 | 16 | MeluXina - Cluster Module | Atos | BullSequana XH2000 | Mellanxo HDR200 | LuxProvide | Bissen | Luxembourg | 2021 | Research and Industry | Research and Industry | 384 | 49152 | 196608 | Custom running on system deployed with ParaStation Modulo | 36 | 1527.39623691 | 135.777 | |
19 | 17 | Nurion | Cray | CS500 Intel Xeon Phi 7250 68C 1.4GHz Intel Omni-Path | Intel Omni-Path | Korea Institute of Science and Technology Information | Daejeon | Korea Republic Of | 2018 | Research | Various scientific and instudrial fields | 1024 | 65536 | 98304 | custom-aml-heavy-diropt | 37 | 1456.46 | 3519 | |
20 | 18 | Turing | IBM | BlueGene/Q Power BQC 16C 1.60GHz | custom | CNRS/IDRIS-GENCI | Orsay | France | 2012 | Environment Fluid Dynamics Biomedical Astrophysics Physics Plasmas Computer Science Mathematics Chemistry | Academic | 4096 | 65536 | 65536 gigabytes | custom | 36 | 1427 | 1348 seconds | 0 Watts |
20 | 18 | Blue Joule | IBM | BlueGene/Q Power BQC 16C 1.60 GHz | Custom | Science and Technology Facilities Council - Daresbury Laboratory | UK | 2012 | Scientific Research | Research | 4096 | 65536 | 65536 gigabytes | Custom | 36 | 1427 | 1348 seconds | 0 Watts | |
20 | 18 | DIRAC | IBM | BlueGene/Q Power BQC 16C 1.60 GHz | Custom | University of Edinburgh | UK | 2012 | Scientific Research | Academic | 4096 | 65536 | 65536 gigabytes | Custom | 36 | 1427 | 1348 seconds | 0 Watts | |
20 | 18 | Zumbrota | IBM | BlueGene/Q Power BQC 16C 1.60 GHz | Custom | EDF R&D | France | 2012 | Energy Research | Industry | 4096 | 65536 | 65536 gigabytes | Custom | 36 | 1427 | 1348 seconds | 0 Watts | |
20 | 18 | Avoca | IBM | BlueGene/Q Power BQC 16C 1.60 GHz | Custom | Victorian Life Sciences Computation Initiative | Australia | 2012 | Scientific Research | Research | 4096 | 65536 | 65536 gigabytes | Custom | 36 | 1427 | 1348 seconds | 0 Watts | |
25 | 23 | ALCF Cetus - 2048 | IBM | IBM - BlueGene/Q Power BQC 16C 1.60 GHz | DOE/ALCF | Chicago IL | United States | 2012 | Scientific Research | Research | 2048 | 32768 | 32768 gigabytes | Custom | 34 | 1384 | 421 seconds | 0 Watts | |
26 | 24 | TSUBAME 2.5 | HP | Cluster Platform SL390s G7 | QDR InfiniBand x 2 (80Gbps) | Global Scientific Information and Computing Center Tokyo Institute of Technology | Tokyo | Japan | 2013 | Research | University | 1024 | 12288 | 51200 gigabytes | Custom (CPU only) | 36 | 1344.95 | 172.487 seconds | 0 Watts |
27 | 25 | JURECA Data Centric Module | Atos | BullSequana XH2000 | Mellanox HDR InfiniBand/ParTec ParaStation Modulo Software Suite | Forschungszentrum Juelich (FZJ) | Juelich | Germany | 2021 | Research | Research | 189 | 24192 | 96768 | Custom | 37 | 1242.74996482 | 244.309 | |
28 | 26 | DARPA Trial Subset | IBM | Power 775 POWER7 8C 3.836GHz | Custom | Endicott | Endicott | USA | 2012 | Research | Research | 1960 | 62720 | 250880 gigabytes | Custom 2D partitioned | 36 | 1171.91 | 326 seconds | 0 Watts |
29 | 27 | Tianhe Exa-node Prototype@Newbenchmarking@GreenBFS | National University of Defense Technology | FT-2000V | TH Express-3 | National Supercomputer Center in Tianjin | Tianjin | China | 2022 | Industry | Research | 1 | 64 | 256 | Custom(Gan xinbiao's implementation) | 27 | 1165.74 | 1201.82 | 135 |
30 | 28 | Microsoft Azure HBv2 | Microsoft Azure | Microsoft Azure HBv2 | InfiniBand | United States | South Central US | United States | 2020 | Industry | High Performance Computing | 640 | 76800 | 480 | Custom | 36 | 1152 | 808.023 | |
31 | 29 | AI Bridging Cloud Infrastructure (ABCI) | Fujitsu | PRIMERGY CX2570 M4 | Infiniband EDR | National Institute of Advanced Industrial Science and Technology (AIST) | Kashiwa Chiba | Japan | 2018 | Research | Artificial Intelligence Research | 256 | 10240 | 98304 | Custom | 32 | 1063.6 | 34.236 | |
32 | 30 | Tianhe Exa-node Prototype@Newbenchmarking@SuperCSR | National University of Defense Technology | FT-2000V | TH Express-3 | National Supercomputer Center in Tianjin | Tianjin | China | 2022 | Industry | Research | 1 | 64 | 512 | Custom(Gan xinbiao's implementation) | 30 | 1055.48 | 1652.42 | 167 |
33 | new | YITU | BDTS-SCTS-CGCL | YITU | Custom | BDTS-SCTS-CGCL | Wuhan | China | 2021 | Research | Academic | 1 | 28 | 256 | Custom | 27 | 1010.59 | 162.572 | 158 |
34 | 31 | Oakleaf-FX | Fujitsu | Fujitsu - Fujitsu PRIMEHPC FX 10 | Tofu | Information Technology Center The University of Tokyo | Tokyo | Japan | 2011 | Varioss scientific applications | University | 4800 | 76800 | Cutom | 36 | 1002.72 | 1302.84 seconds | ||
35 | 32 | Aion | Atos | BullSequana X2410 ARM blades w. AMD Epyc ROME 7H12 | IB HDR 100 | University of Luxembourg Centre de Calcul | 2 avenue de l'Université L-4365 Esch-sur-Alzette | Luxembourg | 2021 | university | Computational sciences physics Engineering Life Science Economy System Bio-medicine Digital history Security and Trust | 256 | 32767 | 65536 | reference sc21 release | 36 | 975.347220596 | 156.006 | |
36 | 33 | Tianhe Exa-node Prototype@GraphV | National University of Defense Technology | FT-2000V | National Supercomputer Center in Tianjin | Tianjin | China | 2021 | Industry | AI Research | 1 | 64 | 512 | ustom-TianheGraph (Gan Xinbiao's implementation) | 26 | 884.443 | 614.547 | 145 | |
37 | 34 | NERSC Edison | Cray | XC30 | Cray Aries | NERSC/LBNL | DOE/SC/LBNL/NERSC | USA | 2013 | basic science and simulation | Government | 4817 | 115600 | 308288 gigabytes | Custom | 36 | 865.3 | 1884.3 seconds | 0 Watts |
38 | 35 | Tianhe Exa-node Prototype | National University of Defense Technology | FT-2000V | National Supercomputer Center in Tianjin | Tianjin | China | 2021 | Industry | Research | 1 | 64 | 512 | Custom (Gan Xinbiao's implementation) | 30 | 811.27 | 15848.6 | 185 | |
39 | 36 | ENIAD | UPenn | Xilinx Ultrascale VU37P | 0 | UPenn | Philadelphia | United States | 2020 | University | AI | 1 | 16 | 1024 | Custom | 26 | 783.7503 | 3 | 130 |
40 | 37 | SAKURA | IBM | BlueGene/Q Power BQC 16C 1.60 GHz | Custom | High Energy Accelerator Research Organization /KEK | Japan | 2012 | Scientific Research | Research | 2048 | 32768 | 32768 gigabytes | Custom | 35 | 769 | 784 seconds | 0 Watts | |
40 | 37 | DOE/NNSA/LLNL Vulcan | IBM | BlueGene/Q Power BQC 16C 1.60 GHz | Custom | Lawrence Livermore National Laboratory | Livermore CA | USA | 2012 | Scientific Research | Government | 2048 | 32768 | 32768 gigabytes | Custom | 35 | 769 | 784 seconds | 0 Watts |
40 | 37 | IBM Rochester 1 | IBM | BlueGene/Q Power BQC 16C 1.60 GHz | Custom | IBM Rochester | Rochester MN | USA | 2012 | Scientific Research | Industry | 2048 | 32768 | 32768 gigabytes | Custom | 35 | 769 | 784 seconds | 0 Watts |
40 | 37 | IBM Rochester 2 | IBM | BlueGene/Q Power BQC 16C 1.60 GHz | Custom | IBM Rochester | Rochester MN | USA | 2012 | Scientific Research | Industry | 2048 | 32768 | 32768 gigabytes | Custom | 35 | 769 | 784 seconds | 0 Watts |
40 | 37 | HIMAWARI | IBM | BlueGene/Q Power BQC 16C 1.60 GHz | Custom | High Energy Accelerator Research Organization /KEK | Japan | 2012 | Scientific Research | Research | 2048 | 32768 | 32768 gigabytes | Custom | 35 | 769 | 784 seconds | 0 Watts | |
40 | 37 | BGQ | IBM | BlueGene/Q Power BQC 16C 1.60 GHz | Custom | Southern Ontario Smart Comuputing Innovation Consortium/Univeristy of Toronto | Canada | 2012 | Scientific Research | Academic | 2048 | 32768 | 32768 gigabytes | Custom | 35 | 769 | 784 seconds | 0 Watts | |
46 | 43 | tatara | Fujitsu | PRIMERGY CX400 | InfiniBand FDR×1 | Kyushu University | Fukuoka | Japan | 2012 | Scientific Research | University | 1024 | 16384 | 131072 gigabytes | Custom | 37 | 736.54 | 311.453 seconds | 0 Watts |
47 | 44 | NERSC Cori - 512 KNL partition | Cray | XC40 | Aries | NERSC/LBNL | DOE/SC/LBNL/NERSC | United States | 2017 | Government | basic science and simulation | 512 | 32768 | 57344 | custom-aml-heavy-diropt | 35 | 678.933 | 1188 | |
48 | 45 | Zeus | IBM | Power 775 POWER7 8C 3.836 GHz | Custom | DARPA Trial Subset IBM Development Engineering | USA | 2012 | Research | Research | 1024 | 62720 | 131072 gigabytes | Custom-IBMRes-Austin | 35 | 508.05 | 353.247 seconds | 0 Watts | |
49 | 46 | TSUBAME 2.0 | HP | Cluster Platform SL390s G7 w/ three Tesla cards per node | QDR InfiniBand x 2 (80Gbps) | GSIC Center Tokyo Institute of Technology | Tokyo | Japan | 2010 | Research | University | 1366 | 16392 | 71032 gigabytes | Custom (GPU) | 35 | 462.25 | 107.08 seconds | 0 Watts |
50 | 47 | JiFeng | HUST && ZHEJIANG LAB | JiFeng (FPGA Prototype) | HUST && ZHEJIANG LAB | Wuhan | China | 2022 | Research | Research | 1 | 512 | 16 | Custom | 20 | 461.222 | 7.587779 | 36.86 | |
51 | 48 | ENIAD | UPenn | FPGA | University of Pennsylvania | Philadelphia PA | United States | 2020 | University | AI | 1 | 8 | 256 | Custom | 30 | 436.10 | 9 | 212 | |
52 | 49 | High Throughput Computer + (4 * NVidia V100-SXM2) | Institute of Computing Technology Chinese Academy of Science | High Throughput Computer | SMARCO && ICT | Beijing | China | 2020 | Scientific Research | High Throughput Big Data Processing | 1 | 48 | 384 | Custom GPU | 26 | 392.041 | 430.485 | 0 | |
53 | 50 | BGQ | IBM | BlueGene/Q Power BQC 16C 1.60 GHz | Custom | Ecole Polytechnique Federale de Lausanne | Switzerland | 2012 | Scientific Research | Academic | 1024 | 16384 | 16384 gigabytes | Custom | 34 | 382 | 754 seconds | 0 Watts | |
53 | 50 | BGQ | IBM | BlueGene/Q Power BQC 16C 1.60 GHz | Custom | Interdisciplinary Centre for Mathematical and Computational Modelling University of Warsaw | Poland | 2012 | Scientific Research | Academic | 1024 | 16384 | 16384 gigabytes | Custom | 34 | 382 | 754 seconds | 0 Watts | |
53 | 50 | Vesta | IBM | BlueGene/Q Power BQC 16C 1.60 GHz | Custom | DOE/SC/Argonne National Laboratory | Chicago IL | USA | 2012 | Scientific Research | Government | 1024 | 16384 | 16384 gigabytes | Custom | 34 | 382 | 754 seconds | 0 Watts |
53 | 50 | Grotius | IBM | BlueGene/Q Power BQC 16C 1.60 GHz | Custom | IBM Thomas J. Watson Research Center | Yorktown Heights NY | USA | 2012 | Scientific Research | Industry | 1024 | 16384 | 16384 gigabytes | Custom | 34 | 382 | 754 seconds | 0 Watts |
53 | 50 | BGQ | IBM | BlueGene/Q Power BQC 16C 1.60 GHz | Custom | Rensselaer Polytechnic Institute | Troy NY | USA | 2012 | Scientific Research | Academic | 1024 | 16384 | 16384 gigabytes | Custom | 34 | 382 | 784 seconds | 0 Watts |
53 | 50 | BGQ | IBM | BlueGene/Q Power BQC 16C 1.60 GHz | Custom | University of Rochester | USA | 2012 | Scientific Research | Academic | 1024 | 16384 | 16384 gigabytes | Custom | 34 | 382 | 784 seconds | 0 Watts | |
59 | 56 | Polymer2 | SJTU && Huawei | Intel(R) Xeon(R) Gold 6330 CPU @ 2.00GHz | IPADS@SJTU && Huawei | IPADS@SJTU && Huawei | China | 2022 | research | research | 1 | 56 | 512 | Custom(Ding Baorong’s implementation) CPU-only | 29 | 356.179 | 134.255 | 510 | |
60 | 57 | Polymer2 | SJTU && Huawei | Intel(R) Xeon(R) Gold 6330 CPU @ 2.00GHz | Shanghai Jiao Tong University | IPADS SJTU | China | 2022 | research | research | 1 | 56 | 512 | Custom | 30 | 350.46 | 284.904 | 510 | |
60 | 57 | Polymer2 | SJTU && Huawei | Intel(R) Xeon(R) Gold 6330 CPU @ 2.00GHz | IPADS Shanghai Jiao Tong University | IPADS@SJTU && Huawei | China | 2022 | research | research | 1 | 56 | 512 | Custom(Ding Baorong’s implementation) CPU-only | 30 | 350.46 | 284.904 | 510 | |
62 | 59 | High Throughput Computer (+Nvidia V100) | Institute of Computing Technology Chinese Academy of Science | High Throughput Computer | SMARCO && ICT | Beijing | China | 2020 | Scientific Research | High Throughput Big Data Processing | 1 | 24 | 32 | Custom GPU | 26 | 319.061 | 129.952 | 152 | |
63 | 60 | Cluster Platform SL390s G7 w/ three Tesla cards per node | NEC/HP | Cluster Platform SL390s G7 w/ three Tesla cards per node | GSIC Center Tokyo Institute of Technology | Tokyo | Japan | 2010 | Research | University | 1366 | 16392 | 73764 gigabytes | unknown | 35 | 317.09 | 536.357 seconds | 0 Watts | |
64 | 61 | snq | IBM | BlueGene/Q | Brookhaven National Laboratory | Upton NY 11973 | USA | 2012 | physics geophysics materials biology chemistry comutational fluid dynamics and others | government industryuniversity | 1024 | 16384 | 16384 gigabytes | custom | 34 | 294.293 | 405.953 seconds | 0 Watts | |
65 | 62 | Pleiades | HPE/SGI | ICE-X dual plane hypercube FDR infiniband E5-2670 sandybridge | Infiniband dual rail FDR | NASA-Ames / Parallel Computing Lab Intel Labs | Mountain View CA | USA | 2012 | Geophysics automotive CFS Climate modeling Astrophysics | Government university industry and Government research | 1024 | 16384 | 32768 gigabytes | custom | 34 | 270.33 | 1550.12 seconds | 0 Watts |
66 | 63 | GMC6KA | AMD and NVIDIA | AMD Ryzen 9 7950X + NVIDIA RTX 6000 Ada | NTT Musashino R-and-D Center | Tokyo | Japan | 2023 | Industry | Research | 1 | 16 | 128 | Custom | 23 | 265.46 | 19.063 | 185.52 | |
67 | 64 | Microway GPU Test Drive | Microway | 14-core Xeon E5-2690v4 CPUs Two NVIDIA Tesla P100 | University of Massachusetts Lowell | Microway | United States | 2016 | Graph Traversal on new GPUs | Graph Traversal | 1 | 14 | 256 gigabytes | Several Developed | 20 | 257.225 | 49.8944 seconds | Watts | |
68 | 65 | DOE/SC Hopper | Cray | XE6 | NERSC/Lawrence Berkeley National Laboratory | DOE/SC/LBNL/NERSC | USA | 2010 | Basic science and simulation | governmen | 4817 | 115600 | 154144 gigabytes | Custom | 35 | 254.074 | 4639.4 seconds | 0 Watts | |
69 | 66 | High Throughput Computing Cluster | Institute of Computing Technology Chinese Academy of Science | High Throughput Computer | InfiniBand 100Gbps | SMARCO && ICT | Beijing | China | 2019 | High Throughput Big Data Processing | Scientific Research | 72 | 2880 | 18432 | Custom BFS implementation | 35 | 249.73 | 2664.26 | 0 |
70 | 67 | Polymer2-Arm | SJTU && Huawei | Kunpeng 920 | IPADS@SJTU && Huawei | IPADS@SJTU && Huawei | China | 2022 | research | research | 1 | 96 | 512 | Custom(Ding Baorong’s implementation) CPU-only | 29 | 243.518 | 247.316 | 340 | |
71 | 68 | High Throughput Computer(+GPU) | Institute of Computing Technology Chinese Academy of Science | High Throughput Computer | SMARCO | Beijing | China | 2019 | High Throughput big data processing | Graph Computing | 1 | 58 | 32 | Custom GPU | 26 | 237.94 | 155.55 | 130 | |
72 | 69 | HA-PACS | Appro International / Cray Inc. | Xtream-X GreenBlade 8204 Xeon E5-2670 8C 2.600GHz Infiniband QDR NVIDIA 2090 | Infiniband QDR | Center for Computational Sciences University of Tsukuba | Tsukuba | Japan | 2012 | Research | University | 128 | 2048 | 16384 gigabytes | Custom (GPU) | 32 | 223.634 | 191.805 seconds | 0 Watts |
73 | 70 | SGI UV 300 | HPE/SGI | UV 300 (32-socket Intel(R) Xeon(R) CPU E7-8890 v3 16 TB RAM) | NUMAlink7 | SGI | Chippewa Falls WI | USA | 2015 | Vendor | Vendor | 1 | 576 | 16384 gigabytes | Custom (NUMA-aware OpenMP with ULIBC SGI & Kyushu University) | 34 | 219.428 | 3422.83 seconds | 0 Watts |
74 | 71 | IBM POWER8+ Tesla P100 | IBM && Nvidia | IBM POWER8+ Tesla P100 | IBM Moscow && MSU Alex Kolganov | Moscow | Russian Federation | 2016 | research | research | 1 | 66 | 256 gigabytes | GPU Custom | 26 | 204 | 150 sec seconds | 175 watts Watts | |
75 | 72 | SHOUBU | PEZY Computing / Exascaler Inc. | ZettaScaler-1.4 80Brick | FDR Infiniband | Advanced Center for Computing and Communication RIKEN | Wako Saitama | Japan | 2015 | Research | Government | 256 | 2048 | 32768 gigabytes | Custom | 31 | 184.725 | 11.5564 seconds | 0 Watts |
76 | 73 | ismuv2k | HPE/SGI | SGI UV 2000 | NUMAlink6 | The Institute of Statistical Mathematics | Tokyo | Japan | 2014 | Data assimilation | National research institute | 1 | 1280 | 32768 gigabytes | Custom (OpenMP + NUMA-aware) | 33 | 174.704 | 20240 seconds | 0 Watts |
77 | 74 | BGQdev | IBM | BlueGene/Q Power BQC 16C 1.60 GHz | Custom | Southern Ontario Smart Computing Innovation Consortium/Univeristy of Toronto | Canada | 2012 | Scientific Research | Academic | 512 | 8192 | 8192 gigabytes | Custom | 33 | 172 | 717 seconds | 0 Watts | |
77 | 74 | BGQdev | IBM | BlueGene/Q Power BQC 16C 1.60 GHz | Custom | IBM Thomas J. Watson Research Center | Yorktown Heights NY | USA | 2012 | Scientific Research | Industry | 512 | 8192 | 8192 gigabytes | Custom | 33 | 172 | 717 seconds | 0 Watts |
77 | 74 | DOE/NNSA rzuseq | IBM | BlueGene/Q Power BQC 16C 1.60 GHz | Custom | Lawrence Livermore National Laboratory | Livermore CA | USA | 2012 | Scientific Research | Government | 512 | 8192 | 8192 gigabytes | Custom | 33 | 172 | 717 seconds | 0 Watts |
80 | 77 | Bem | SuperMicro/WASKO | Cluster | infiniband FDR | Wrocław University of Technology | Plac Grunwaldzki 11 50-377 Wrocław | Poland | 2015 | general purpose | University | 720 | 17280 | 47000 gigabytes | Tuned MPI implementation with 2-D data distribution | 33 | 168.288 | 544 seconds | 0 Watts |
81 | 78 | High Throughput Computer (+ NVidia V100) | Institute of Computing Technology Chinese Academy of Science | High Throughput Computer | SMARCO && ICT | Beijing | China | 2020 | Scientific Research | High Throughput Big Data Processing | 1 | 48 | 384 | Custom GPU | 30 | 163.461 | 528.223 | 241 | |
82 | 79 | NEDO-AIRC-NAIROBI-UV | HPE/SGI | UV300 (Intel Xeon E7-8867v3 x16) | National Institute of Advanced Industrial Science and Technology (AIST) | Tokyo | Japan | 2016 | Artificial Intelligence | Government | 1 | 256 | 11775.7 gigabytes | Custom (graph500_hpgp2016_2016-05-27) | 33 | 161.215 | 2759.46 seconds | 4663.56 Watts | |
83 | 80 | ismuv2k2 | HPE/SGI | SGI UV 2000 | NUMAlink6 | The Institute of Statistical Mathematics | Tokyo | Japan | 2014 | Data assimilation | National research institute | 1 | 640 | 16384 gigabytes | Custom (OpenMP NUMA Degree-aware BFS) | 33 | 149.8 | 7654.7 seconds | 0 Watts |
84 | 81 | High Throughput Computer(+Nvidia V100) | Institute of Computing Technology Chinese Academy of Science | High Throughput Computer | SMARCO && ICT | Beijing | China | 2019 | High Throughput Big Data Processing | Scientific Research | 1 | Custom GPU | 30 | 146.79 | 478.65 | 230 watts | |||
85 | 82 | Hawk028os1 | Hewlett Packard Enterprise | HPE Integrity Superdome X Gen9 | N/A | HPE Ft. Collins | Ft. Collins CO | USA | 2015 | Multi-discipline | Vendor Benchmarking | 1 | 288 | 12288 gigabytes | Custom (Kyushu University Japan) | 33 | 141.9 | 5242 seconds | 9508 Watts |
86 | 83 | TitanX_forsite | Intel(R) Server | CPU Intel(R) Core i5 4th Gen (4 cores) && NVidia GTX Titan X | none | forsite-company.ru && NVidia LTD | Moscow State University && NVidia LTD | Russia | 2015 | Scientific Research | University | 1 | 28 | 32 gigabytes | Custom GPU | 26 | 132.14 | 305 seconds | 162 Watts |
87 | 84 | Xeon(R) Gold 5118+Tesla P100 | Intel | Xeon(R) Gold 5118+Tesla P100 | SMARCO | Beijing | China | 2018 | High Performance computing | High Performance computing | 1 | 24 | 192 | Custom GPU | 26 | 131.53 | 155.5 | 220 | |
88 | 85 | hawk040os1 | Hewlett-Packard | HP Integrity Superdome X | N/A | HP Ft. Collins | Ft Collins CO | USA | 2014 | Benchmarking | Vendor Benchmarking | 1 | 240 | 12288 gigabytes | Custom (Kyushu's implementation) | 33 | 128.336 | 6382 seconds | 6264 Watts |
89 | 86 | Colonial | Supermicro | Intel(R) Xeon(R) CPU E5-2620 x 2 NVIDIA K40c x 2 | George Washington University | Washington DC | USA | 2014 | Computer Engineering | University | 1 | 12 | 20 gigabytes | Custom | 20 | 122.182 | 72.8705 seconds | 274 Watts | |
90 | 87 | GMC6K | Intel and NVIDIA | Intel(R) Core(TM) i9-10940X CPU @ 3.30GHz + NVIDIA RTX A6000 | NTT Musashino R-and-D Center | Tokyo | Japan | 2021 | Industry | Research | 1 | 28 | 256 | Custom | 26 | 117.24 | 599.123 | 288 | |
91 | 88 | Endeavor | Intel | Dual-socket 2.6GHz SNB-EP | Mellanox Infiniband Febric | Intel Dupont / Parallel Computing Lab Intel Labs | Dupont Washington | USA | 2012 | Geophysics automotive and commercial benchmarking for all fields | Government university industry and commercial benchmarking | 320 | 5120 | 20480 gigabytes | custom | 32 | 115.935 | 44.6482 seconds | 0 Watts |
92 | 89 | Titan_Keldysh | Intel(R) Server | CPU Intel(R) Xeon(R) E5-1660v2 (6 cores 3.7GHz) && GPU NVidia GTX Titan | none | Keldysh Institute of Applied Mathematics (Russian Academy of Sciences) | Moscow State University | Russia | 2012 | Scientific Research | University | 1 | 20 | 40 gigabytes | Custom GPU | 25 | 114.68 | 58.6 seconds | 212 Watts |
93 | 90 | TSUBAME-KFC | NEC | self-made | Infiniband FDR | Tokyo Institute of Technology | Tokyo | Japan | 2013 | Research | University | 32 | 384 | 2048 gigabytes | custom | 32 | 104.31 | 133.436 seconds | 7419.85 Watts |
94 | 91 | Lomonosov | T-Platforms | MPP | Moscow State University | Russia | 4096 | 32768 | 0 gigabytes | Custom | 37 | 103.079 | 0 seconds | 0 Watts | |||||
95 | 92 | Jugene | IBM | BlueGene/P | Forschungszentrum Juelich | Germany | 64000 | 256000 | 0 gigabytes | Custom | 37 | 92.3418 | 0 seconds | 0 Watts | |||||
96 | 93 | Big Red 2 | Cray Inc. | XE6/XK7 | Cray Gemini | Indiana University | Bloomington Indiana | USA | 2013 | Academic research | University | 1020 | 21824 | 43648 gigabytes | Custom | 32 | 88.8 | 8.61388 seconds | 0 Watts |
97 | 94 | HYPATIA | Data Vortex Technologies | DV206 | Data Vortex Proprietary | Plexus Engineering | Neenah Wi | USA | 2015 | Scientific Computing Graph Analytics | Development | 32 | 256 | 8192 gigabytes | Custom | 33 | 88.112 | 4344.97 seconds | 10498.2 Watts |
98 | 95 | Intrepid | IBM | BlueGene/P | DOE/SC/Argonne National Laboratory | Argonne National Laboratory | USA | 2009 | Science | Government | 32768 | 131071 | 65536 gigabytes | Custom | 35 | 80.64 | 26.73 seconds | 0 Watts | |
99 | 96 | Graph-WICIL-Node0 | MICRON | PicoComputing AC520 | NA | University of Wisconsin-Madison | Madison WI | United States | 2018 | Research | Research | 1 | 18 | 4 | Custom | 23 | 79.531 | 29.6908 | 56.8 |
100 | 97 | SALOMON | HPE/SGI | SGI ICE X (2016 Intel Xeon E5-2680 v3 12-core processors @ 2.5 GHz (24192 cores) + 864 Intel Xeon Phi 7120P 61-core accelerators @ 1.238 GHz (52704 cores) | FDR Infiniband | IT4Innovations national supercomputing center VSB-Technical University of Ostrava | Ostrava | Czech Republic | 2015 | Research | University Industry | 722 | 17328 | 0 gigabytes | Tuned MPI Implementation with 2-D data distribution from http: //www.graph500.org/referencecode | 33 | 70.7969 | 23.5858 seconds | 0 Watts |
101 | 98 | Sierra | IBM / NVIDIA / Mellanox | IBM Power System S922LC IBM POWER9 22C 3.1GHz NVIDIA Volta GV100 Dual-rail Mellanox EDR Infiniband | Dual-rail Mellanox EDR Infiniband | DOE/NNSA/LLNL | Livermore CA | United States | 2018 | Government | Scientific Research | 2048 | 81920 | 524288 | Custom NVRAM out-of-core | 42 | 67.258 | 11572.795 | |
102 | 99 | GraphCREST-Huawei | Huawei | Huawei RH5885H V3 (4-way Intel Xeon E7-4890 v2 and 2TB RAM) | Kyushu University | Fukuoka | Japan | 2014 | Graph analysis | University | 1 | 60 | 2048 gigabytes | Custom (OpenMP + NUMA-aware + Degree-aware) | 32 | 59.9 | 3209.9 seconds | 1153.1 Watts | |
103 | 100 | JunGPU | Nvidia | E5-2683 v3 Titan X | Tsinghua University | Beijing | China | 2016 | High performance computing | University | 1 | 3072 | 12 gigabytes | Custom | 26 | 48.295 | 22.3719 seconds | 0 Watts | |
104 | 101 | altix | HPE/SGI | SGI Altix UV 1000 | Research Center for Advanced Computing Infrastructure Japan Advanced Institute of Science and Technology | Ishikawa | Japan | 2012 | Research | University | 16 | 256 | 2048 gigabytes | Custom (OpenMP NUMA Degree-aware BFS) | 31 | 47.071 | 9906.29 seconds | 0 Watts | |
105 | 102 | GraphCREST-Sandybridge-EP-2.7GHz | HPCTECH Corporation | Intel(R) Xeon(R) CPU E5-4650 @ 2.70GHz (4 sockets) | Kyushu University | Fukuoka | Japan | 2013 | Graph analysis | University | 1 | 32 | 512 gigabytes | Custom (OpenMP NUMA Degree-aware BFS) | 27 | 45.71 | 234.545 seconds | 665 Watts | |
106 | 103 | GraphCREST-SandybridgeEP-2.4GHz | HPCTECH Corporation | Intel Xeon CPU E5-4640 2.40GHz (4 sockets) | Kyushu University | Fukuoka | Japan | 2013 | Graph analysis | University | 1 | 32 | 512 gigabytes | Custom (OpenMP NUMA Degree-aware BFS) | 27 | 42.45 | 260.017 seconds | 520.9 Watts | |
107 | 104 | GraphCREST-Westmere-EX-2.4GHz | HPCTECH Corporation | Intel(R) Xeon(R) CPU E7 4870 2.40GHz (4 sockets) | Kyushu University | Fukuoka | Japan | 2012 | Graph analysis | University | 1 | 40 | 512 gigabytes | Custom (OpenMP NUMA Degree-aware BFS) | 27 | 42.114 | 238.947 seconds | 0 Watts | |
108 | 105 | I950-G | Sugon | 8x Intel Xeon X7550 CPUs (8 core 2.0GHz) | Institute of Computing Technology Chinese Academy of Sciences | Beijing | China | 2011 | Research | University | 16 | 1024 | 4096 gigabytes | Optimized | 32 | 41.9924 | 423.075 seconds | 0 Watts | |
109 | 106 | IBM Power8+ Tesla P100 | IBM && Nvidia | IBM Power8+ Tesla P100 | IBM Moscow && MSU Alex Kolganov | Moscow | Russian Federation | 2016 | research | research | 1 | 66 | 256 gigabytes | GPU Custom | 30 | 41.7 | 2713 sec seconds | 235 Watts | |
110 | 107 | High Throughput Computer | Institute of Computing Technology Chinese Academy of Science | Single Node ARM Server | SMARCO | Beijing | China | 2020 | High Throughput big data processingroug | Big Data | 1 | 40 | 384 | custom | 30 | 39.83 | 2000 | 122 | |
111 | 108 | Monty Pi-thon | Newark Element14 | Raspberry Pi 2 Model B | 10/100 Fast Ethernet driven through onboard USB 2.0 bus connected to Dell PowerConnect 6248 Switch via flat topology | CODE@TACC (Texas Advanced Computing Center Summer High School STEM Program) | Austin Texas | USA | 2015 | Education Outreach | University | 32 | 128 | 32 gigabytes | graph500-2.1.4-energy | 26 | 35.8254 | 403.705 seconds | 147.175 Watts |
112 | 109 | Super2PiR | Newark Element14 | Raspberry Pi 2 Model B | 10/100 Fast Ethernet driven through onboard USB 2.0 bus connected to Dell PowerConnect 6248 Switch via flat topology | CODE@TACC (Texas Advanced Computing Center Summer High School STEM Program) | Austin Texas | USA | 2015 | Education Outreach | University | 32 | 128 | 32 gigabytes | Reference 2.1.4 with *validation disabled* | 26 | 35.5917 | 552.526 seconds | 0 Watts |
113 | 110 | kitty6 | SYSTEM WORKS | Xeon E5-2650 v2 GeForce GTX TITAN Black | University of Tsukuba | Tsukuba | Japan | 2014 | Research | University | 1 | 8 | 32 gigabytes | Custom (GPU) | 25 | 35.18 | 335.485 seconds | 206.164 Watts | |
114 | 111 | GraphCREST-Sandybridge-EP-2.7GHz | Chuo University | 64 | 27 | 31.648 | 771.71 Watts | ||||||||||||
115 | 112 | RSC KNL | RSC && Intel | RSC Intel Xeon KNL 7250 | RSC Moscow && MSU Alex Kolganov | Moscow | Russian Federation | 2017 | research | research | 1 | 68 | 98 gigabytes | Custom | 29 | 31.3 | 1742 sec seconds | 215 Watts | |
116 | 113 | GraphCREST-4Way-SandybridgeEP-2.4GHz | Chuo University | 64 | 27 | 29.034 | 639.1 Watts | ||||||||||||
117 | 114 | NTHU Quanta Cluster | Quanta Computer | Quanta - Intel(R) Xeon(R) E5-2670 2.60GHz | InfiniBand | National Tsing Hua University | Hsinchu | Taiwan | 2012 | HPC applications | university | 8 | 128 | 0 gigabytes | reference | 8 | 28.3746 | 67.4514 seconds | 0 Watts |
118 | 115 | GraphCREST-Magny-Cours | Dell | AMD Opteron(tm) Processor 6174 2.2 GHz (4 sockets) | Kyushu University | Fukuoka | Japan | 2011 | Graph analysis | University | 1 | 48 | 256 gigabytes | Custom (OpenMP NUMA Degree-aware BFS) | 27 | 25.243 | 359.16 seconds | 0 Watts | |
119 | 116 | momcom-max | Apple | MacBook Pro with M1 Max | NTT Musashino R-and-D Center | Tokyo | Japan | 2022 | Industry | Research | 1 | 10 | 64 | Custom | 27 | 25.16 | 672.891 | 27.37 | |
120 | 117 | rc1.umb.edu | Microway | Microway NumberSmasher with two Intel Xeon 6150 | UMass Boston | Boston | United States | 1 | university | research | 1 | 36 | 192 | custom | 27 | 23.77 | 398 | ||
121 | 118 | Tellico | IBM | IBM Power System AC922 | University of Tennessee Knoxville | Knoxville | United States | 2019 | University | Research | 2 | 64 | 256 | Custom | 28 | 23.6756079674 | 44.471 | ||
122 | 119 | Phoenix | Penguin Computing | Relion XE2142 | Infiniband HDR100 | Georgia Institute of Technology | Atlanta GA | United States | 2020 | University | Physics chemistry engineering | 1275 | 30600 | 361800 | Reference | 35 | 23.152 | 61.3598 | |
123 | 120 | UV 2000 | HPE/SGI | SGI UV 2000 | NUMAlink6 | SGI | Chippewa Falls WI | USA | 2012 | Vendor | Vendor | 128 | 2048 | 32768 gigabytes | Reference | 34 | 19.62 | 1607.04 seconds | 0 Watts |
124 | 121 | Alkindi27-hybrid | Dell | Single node commodity server | The University of British Columbia | Vancouver | Canada | 2015 | University | Research | 1 | 28 | 256 gigabytes | custom | 29 | 19.4685 | 634.11 seconds | 579.6 Watts | |
125 | 122 | Franklin | Cray | XT4 | Lawrence Berkeley National Laboratory | USA | 4000 | 16000 | 0 gigabytes | Custom | 36 | 19.3274 | 0 seconds | 0 Watts | |||||
126 | 123 | Tryton | Action/WASKO | Actina SOLAR 820 S6 | InfiniBand | Academic Computer Centre in Gdansk (CI TASK) | Gdańsk | Poland | 2015 | BiochemistryOceanagraphyCFD | University | 147 | 3528 | 0 gigabytes | mpi-tuned-2d | 29 | 18.9 | 95.4345 seconds | 0 Watts |
127 | 124 | PowerfullServer | Intel | Intel Xeon E5 2699 v3 | Alex Kolganov | Moscow | Russian Federation | 2016 | research | research | 1 | 18 | 384 gigabytes | Custom | 30 | 16.3 | 2700 sec seconds | 145 Watts Watts | |
128 | 125 | Alkindi28 | Dell | Commodity Machine / Intel Xeon E7-4870 v2 (4 Sockets 60 cores 120 hardware threads) | The University of British Columbia | Vancouver BC | Canada | 2015 | University | Research | 1 | 60 | 1536 | custom NUMA-aware - Joint work of UBC and RIKEN-AIP | 31 | 15.98 | 3874 | 0 | |
129 | 126 | Todi | Cray Inc. | Xk7 | Gemini | Swiss National Supercomputing Center | Lugano | Switzerland | 2011 | General purpose scientific computing | University | 272 | 8160 | 10336 gigabytes | Custom developed by M. Fatica group at Nvidia | 29 | 15.5983 | 382.819 seconds | 79012 Watts |
130 | 127 | Nano-Graphybrid++ | AMD | 4 AMD Radeon R9 Nano GPUs | AMD Research | Austin TX | USA | 2015 | Architecture Testbed | Industry | 1 | 16 | 128 gigabytes | Custom (Hybrid++) | 25 | 15.58 | 22.9 seconds | 0 Watts | |
131 | 128 | EBD-RH2288V3 | Huawei | RH2288/V3 (Intel®SSD 750 series 1.2TB crucial®m4 mSATA 256GB x8) | Tokyo Institute of Technology | Tokyo | Japan | 2015 | Research | University | 1 | 36 | 256 gigabytes | Custom (netalx-v7.7fz-sg OpenMP NUMA-optimized NVM) | 32 | 14.8566 | 23556.3 seconds | 334.985 Watts | |
132 | 129 | fox6 | Convey | MX-100 host-210 | n/a | Convey Computer Corporation | Richardson TX | USA | 2012 | big data | vendor | 1 | 16 | 384 gigabytes | custom | 29 | 14.56 | 7682 seconds | 0 Watts |
133 | 130 | Altix ICE 8400EX | HPE/SGI | Altix ICE 8400EX | SGI | USA | 256 | 1024 | 0 gigabytes | Reference | 31 | 13.9586 | 0 seconds | 0 Watts | |||||
134 | 131 | GraphCREST-Custom #1 | Custom | Intel® Xeon™ E5-2690 crucial@m4 mSATA 256GB x16 | Tokyo Institute of Technology | Tokyo | Japan | 2014 | Research | University | 1 | 32 | 256 gigabytes | Custom (netalx-v7.7fz-sg OpenMP NUMA-optimized NVM) | 31 | 13.7963 | 90234.5 seconds | 391.825 Watts | |
135 | 132 | GraphCREST-Bulldozer | HPCTECH Corporation | AMD Opteron(tm) Processor 6386 SE (4 sockets) | Kyushu University | Fukuoka | Japan | 2013 | Graph analysis | University | 1 | 64 | 1024 gigabytes | Custom (OpenMP NUMA Degree-aware BFS) | 31 | 13.633 | 3218.79 seconds | 790.8 Watts | |
136 | 133 | IBM POWER8+ | IBM | IBM POWER8+ | IBM Moscow & MSU Alex Kolganov | Moscow | Russian Federation | 2016 | research | research | 1 | 10 | 256 gigabytes | Custom | 30 | 13.2 | 2713 sec seconds | 200 watts Watts | |
137 | 134 | Colonial0 | Microway | Intel(R) Xeon(R) CPU E5-2683 v3 @ 2.00GHz | George Washington University | Washington DC | USA | 2015 | High-Performance Computing Big Data Computing | Computer Engineering | 1 | 28 | 320 gigabytes | Custom (Self-develop code) | 30 | 12.9 | 532.52 seconds | 350 Watts | |
138 | 135 | MIC | Intel | Xeon Phi 7120P | Tsinghua University | Tsinghua University | China | 2014 | High performance computing | University | 1 | 61 | 16 gigabytes | custom | 25 | 12.87 | 14.7 seconds | 0 Watts | |
139 | 136 | Nebulae | Sugon | TC3600 Blade System Xeon X5650 6C 2.66GHz | InfiniBand QDR | National Supercomputing Centre in Shenzhen | Shenzhen | China | 2010 | Multiphysics Fluid mechanics | Industry | 512 | 6144 | 12288 gigabytes | Custom | 33 | 12.119 | 94.3242 seconds | 0 Watts |
140 | 137 | ROMEO HPC Center | Bull Atos | BullX R421 | Infiniband QDR | University of Reims Champagne-Ardenne | Reims | France | 2013 | All scientific domains | Academic and Industry | 130 | 2080 | 4160 gigabytes | Custom | 27 | 11.7806 | 203 seconds | |
141 | 138 | thunder4 | Convey | HC-2ex host-110 | n/a | Convey Computer Corporation | Richardson TX | USA | 2012 | big data | vendor | 1 | 12 | 256 gigabytes | custom | 27 | 11.614 | 1432 seconds | 0 Watts |
142 | 139 | Celero | Convey | HC-2ex host-210 | n/a | Argonne National Laboratory | Illinois | USA | 2012 | Bioinformatics | DOE | 1 | 8 | 192 gigabytes | custom | 27 | 11.448 | 1368 seconds | 0 Watts |
143 | 140 | Convey HC-2ex | Convey | HC-2ex host-210 | UC Riverside | Riverside CA | USA | 2012 | HPC | University | 1 | 8 | 200 gigabytes | custom | 27 | 11.447 | 1414 seconds | 0 Watts | |
144 | 141 | coconino | Convey | HC-2ex host-210 | n/a | Convey Computer Corporation | Richardson TX | USA | 2012 | big data | vendor | 1 | 8 | 448 gigabytes | custom | 27 | 11.444 | 1470 seconds | 0 Watts |
145 | 142 | Intel-8180 | Intel | Intel | Shenzhen Institutes of Advanced Technology(SIAT) Chinese Academy of Science | Shenzhen | China | 2019 | Scientific Research | Deep Learning | 1 | 56 | 96 | Custom | 23 | 11.3 | 117 | 0 | |
146 | 143 | Alkindi28-CPU | Dell | Commodity Machine / Intel Xeon E7-4870 v2 (4 Sockets 60 cores 120 hardware threads) | The University of British Columbia | Vancouver BC | Canada | 2015 | University | Research | 1 | 60 | 1536 gigabytes | Custom | 32 | 11.1678 | 31889.4 seconds | 591.024 Watts | |
147 | 144 | GraphCREST-4waySBep2.4 | Chuo University | 32 | 26 | 11.115 | 639 Watts | ||||||||||||
148 | 145 | GraphCREST-Wex40 | Chuo University | 40 | 26 | 11.061 | 1010.13 Watts | ||||||||||||
149 | 146 | Grace | Cray | XMT2 | Cray SeaStar | Mayo Clinic Rochester | Rochester MN | USA | 2011 | Semantic Data Analysis and Bioinformatics Number of Processors | Research | 64 | 64 | 2048 gigabytes | custom | 31 | 10.9415 | 4035.8 seconds | 0 Watts |
150 | 147 | RAIDEN MMS Subsystem | Fujitsu | single node commodity server (Intel(R) Xeon(R) Platinum 8160M CPU @ 2.10GHz) | RIKEN AIP | Tokyo | Japan | 2018 | Research | scientific research | 1 | 96 | 6144 | custom (NUMA-aware Totem from the University of British Columbia) | 34 | 10.63 | 55582 | ||
151 | 148 | grace | Cray | XMT-2 | Mayo Clinic | Rochester MN | USA | 2011 | Medicine | Research | 64 | 64 | 2048 gigabytes | Custom | 31 | 10.319 | 4306 seconds | 14100 Watts | |
152 | 149 | Alkindi26-hybrid | Dell | Commodity Machine / Intel Xeon E5-2670 (2 Sockets 20 cores 40 hardware threads) Nvidia Kepler K40 (2 GPUs) | The University of British Columbia | Vancouver BC | Canada | 2014 | Research | University | 1 | 20 | 512 gigabytes | Custom | 30 | 10.3079 | 7040 seconds | 461 Watts | |
153 | 150 | Altix UV 1000 | HPE/SGI | Altix UV 1000 | SGI | USA | 128 | 2048 | 0 gigabytes | Reference | 32 | 10.0932 | 0 seconds | 0 Watts | |||||
154 | 151 | TH-IVB-FEP/C | National University of Defense Technology | SMP | Changsha China | Changsha China | China | 2013 | Science research Big data | Research | 1 | 24 | 64 gigabytes | custom | 26 | 9.74402 | 180.843 seconds | 248 Watts | |
155 | 152 | Xeon Server | Dell | 4 x Intel(R) Xeon(R) Gold 5115 CPU @ 2.40GHz | Industry | Boise ID | United States | 2017 | Research | Research | 1 | 32 | heavy-newreference | 21 | 9.68 | 3.78 | |||
156 | 153 | Red Sky | Sun Microsystems | Sun Blade x6275 Xeon X55xx 2.93 Ghz | Sandia National Laboratories | USA | 512 | 4096 | 0 gigabytes | Custom/ETI SWARM | 33 | 9.44893 | 0 seconds | 0 Watts | |||||
157 | 154 | Lonestar | Dell | PowerEdge 1955 2.66 GHz Infiniband | TACC | USA | 1024 | 12288 | 0 gigabytes | Custom/ETI SWARM | 35 | 9.23418 | 0 seconds | 0 Watts | |||||
158 | 155 | PowerEdge M610 | Dell | PowerEdge M610 | TACC | USA | 512 | 6144 | 0 gigabytes | 34 | 8.05306 | 0 seconds | 0 Watts | ||||||
159 | 156 | MEM-CREST Node #2 | Supermicro | 2027GR-TRF w/ FusionIO ioDrive2 1.2TB x2 | Tokyo Institute of Technology | Tokyo | Japan | 2012 | Research | University | 1 | 16 | 128 gigabytes | Custom (netalx-v7.7fz-sg OpenMP NUMA-optimized NVRAM) | 30 | 7.98417 | 52984.9 seconds | 276.507 Watts | |
160 | 157 | DAS-4/VU | SuperMicro | VU University | Netherlands | 128 | 1024 | 0 gigabytes | Reference MPI/OpenMP | 30 | 7.0867 | 0 seconds | 0 Watts | ||||||
161 | 158 | IBM BlueGene/P | IBM | BlueGene/P | Moscow State University | Russia | 2048 | 8192 | 0 gigabytes | Custom | 32 | 6.87195 | 0 seconds | 0 Watts | |||||
162 | 159 | Lucata Pathfinder-I | Lucata Corporation | Pathfinder-I | Lucata Corporation | New York NY | United States | 2023 | Research | Research | 32 | 512 | 4096 | Custom | 32 | 6.8343 | 0 | 4650 | |
163 | 160 | GraphCREST-SBep2.9 | Chuo University | 16 | 25 | 6.777 | 362.79 Watts | ||||||||||||
164 | 161 | Intel Westmere X5670 | Intel | Westmere X5670 | Intel | USA | 0 | 3072 | 0 gigabytes | 33 | 6.76457 | 0 seconds | 0 Watts | ||||||
165 | 162 | Vortex | Convey | HC-1ex | Convey Computer Corporation | Richardson TX | USA | 2012 | Geophysics automotive etc. | Government university industry etc. | 1 | 4 | 192 gigabytes | custom | 27 | 6.63583 | 3747.82 seconds | 0 Watts | |
166 | 163 | URAN | HP | Cluster/16 nodes 32 CPU Intel Xeon X5675 | IB DDR | Institute of Mathematics and Mechanics UrB RAS | Yekaterinburg | Russia | 2012 | Mathematical and physical modelling | Science education industry | 16 | 192 | 1904 gigabytes | graph500_custom_imm | 26 | 6.25902 | 18.6411 seconds | 0 Watts |
167 | 164 | JaguarPF | Cray | XT5 | Oak Ridge National Laboratory | USA | 512 | 1024 | 0 gigabytes | Custom/ETI SWARM | 33 | 6.2277 | 0 seconds | 0 Watts | |||||
168 | 165 | GraphCREST-SBep2.0 | Chuo University | 16 | 25 | 6.142 | 298.44 Watts | ||||||||||||
169 | 166 | Convey HC-1ex | Convey | Computer Corporation HC-1ex | Infiniband | Virginia Bioinformatics Institute | Blacksburg VA | USA | 2011 | Bioinformatics | University | 1 | 4 | 160 gigabytes | Custom (Convey) | 27 | 5.9962 | 3588.6 seconds | 0 Watts |
170 | 167 | Cray XT4 | Cray | XT4 | NERSC | USA | 9544 | 38176 | 0 gigabytes | Optimized | 32 | 5.60493 | 0 seconds | 0 Watts | |||||
171 | 168 | UND Cluster | Dell | Cluster | University of Notre Dame | Notre Dame IN | USA | 2016 | Research | University | 8 | 128 | 2048 gigabytes | New Reference | 27 | 5.55522 | 60.0327 | 0 watts | |
171 | 168 | University of Notre Dame cluster | Dell | cluster | Ethernet | University of Notre Dame | University of Notre Dame | United States | 2016 | University | Research | 8 | 128 | 2048 gigabytes | newreference+heavy | 27 | 5.55522 | 60.0327 seconds | Watts |
173 | 170 | ULP2012 ultra-green oil-cooled server : GraphCREST | self-made | Xeon E5-2620 2.0GHz x 2 (12 cores 24 threads) (oil cooling) / 1 node | Tokyo Institute of Technology | Tokyo | Japan | 2012 | high-performance computing | university | 1 | 12 | 32 gigabytes | custom (OpenMP) | 24 | 5.12372 | 25.8703 seconds | 168.5 Watts | |
174 | 171 | mirasol | Intel | Westmere-EX | Georgia Tech | USA | 1 | 10 | 0 gigabytes | 28 | 5.04659 | 0 seconds | 0 Watts | ||||||
175 | 172 | SuperMicro | SuperMicro | VU University | Netherlands | 128 | 0 | 0 gigabytes | 31 | 4.61709 | 0 seconds | 0 Watts | |||||||
176 | 173 | MIC | Intel | Xeon Phi 7120P | PACMAN Group | Tsinghua University Beijing | China | 2014 | High performance computing | University | 1 | 61 | 16 gigabytes | custom | 25 | 4.53221 | 19.3895 seconds | 0 Watts | |
177 | 174 | Blacklight | HPE/SGI | Altix UV1000 | PSC | USA | 32 | 512 | 0 gigabytes | Custom | 32 | 4.40234 | 0 seconds | 0 Watts | |||||
178 | 175 | MIC | Intel | Xeon Phi 7120P | DISLab.org NICEVT / svetcorp.net | Moscow | Russia | 2013 | Computer engineering | Industry | 1 | 61 | 16 gigabytes | Custom | 25 | 4.36578 | 491.4 seconds | 0 Watts | |
179 | 176 | Alkindi26-CPU | Dell | Commodity Machine / Intel Xeon E5-2670 (2 Sockets 20 cores 40 hardware threads) | The University of British Columbia | Vancouver BC | Canada | 2014 | Research | University | 1 | 20 | 512 gigabytes | Custom | 30 | 4.24712 | 3247 seconds | 391 Watts | |
180 | 177 | Catalyst | Cray / Intel | Cray CS300 Supercomputer dual Intel E5-2695v2 processors 128GB memory Intel 910 800GB SSD | Intel QDR-80 dual rail Infiniband | Lawrence Livermore National Laboratory | Livermore CA | USA | 2014 | Research | Government | 300 | 7200 | 38400 gigabytes | Custom | 40 | 4.175 | 172800 seconds | 0 Watts |
181 | 178 | GPU-based cluster | Self-assembled | GPU-based cluster | Seoul National University Korea | Seoul | South Korea | 2010 | Computer Science | University | 8 | 192 | 384 gigabytes | custom | 26 | 4.06453 | 512.861 seconds | 0 Watts | |
182 | 179 | Georgia Tech CRNCH Pathfinder | Lucata | Pathfinder-S | Georgia Institute of Technology | Atlanta GA | United States | 2022 | Research | University | 32 | 512 | 2048 | Custom | 30 | 3.873 | 0 | 2868 | |
183 | 180 | Mustang | Appro | Xtreme-X cluster with dual-socket 12-core AMD MagnyCours processors | Mellanox InfiniScale IV 4x-QDR InfiniBand | Los Alamos National Laboratory | Los Alamos New Mexico | USA | 2012 | open science (any field) | government | 86 | 2064 | 5504 gigabytes | graph500_mpi_replicated_csc from v2.1.4 | 31 | 3.79502 | 455.079 seconds | 0 Watts |
184 | 181 | I840r-GP | Sugon | 4x Intel Xeon E7-4830 CPU (8 core 2.13GHz) | Institute of Computing Technology Chinese Academy of Sciences | Beijing | China | 2012 | Research | University | 1 | 32 | 1024 gigabytes | Custom | 30 | 3.69 | 15400.8 seconds | 0 Watts | |
185 | 182 | EBD-RH5885v2 | Huawei | Tecal RH5885 V2 Server w/ Tecal ES3000 PCIe SSD 800GBx2 1.2Tx2 | Tokyo Institute of Technology | Tokyo | Japan | 2013 | Research | University | 1 | 48 | 512 gigabytes | Custom (OpenMP NUMA-optimized Hybrid Double Graph NVRAM) | 30 | 3.67475 | 2822.84 seconds | 844.337 Watts | |
186 | 183 | MEM-CREST Node #1 | Supermicro | Intel® Xeon™ E5-2660 x 2socets FusionIO ioDrive2 1.2TB | Tokyo Institute of Technology | Tokyo | Japan | 2012 | Research | University | 1 | 16 | 64 gigabytes | Custom (OpenMP NUMA-optimized Hybrid Single Graph NVRAM) | 29 | 3.11171 | 6686.73 seconds | 303.6 Watts | |
187 | 184 | Todi | Cray | XK7 | CSCS | Switzerland | 0 | 0 | 0 gigabytes | Custom GPU Result | 28 | 3.00648 | 0 seconds | 0 Watts | |||||
188 | 185 | GraphCREST-M48 | Dell | AMD Opteron 6174 2.2GHz | Chuo University | Tokyo | Japan | 2011 | Mathematical optimization | University | 1 | 48 | 256 gigabytes | custom | 29 | 2.42145 | 276.558 seconds | 0 Watts | |
189 | 186 | APU | AMD | Kaveri(AMD A10-7850K Radeon R7) | Tsinghua University | Beijing | China | 2014 | Research | University | 1 | 4 | 32 gigabytes | Custom | 26 | 2.07 | 14.6525 seconds | 0 Watts | |
190 | 187 | GraphCREST-Retina15 | Chuo University | 4 | 24 | 1.987 | 38.5 Watts | ||||||||||||
191 | 188 | GraphCREST-Mac-mini | Chuo University | 4 | 24 | 1.941 | 36.3 Watts | ||||||||||||
192 | 189 | B2-mic0 | Intel | Xeon Phi 5110P | DISLab.org NICEVT | Moscow | Russia | 2013 | Computer engineering | Industry | 1 | 60 | 8 gigabytes | Custom | 23 | 1.80143 | 13.78 seconds | 0 Watts | |
193 | 190 | GraphCREST-7048GR-TR | Supermicro | 7048GR-TR (Huawei ES3000 2.4TB) | Tokyo Institute of Technology | Tokyo | Japan | 2015 | Research | University | 1 | 36 | 128 gigabytes | Custom (netalx-v7.7fz-sg OpenMP NUMA-optimized NVM) | 31 | 1.77788 | 11561.8 seconds | 336.018 Watts | |
194 | 191 | GLADOS | Convey | Computer Corporation HC-1 | Imperial College London | South Kensington Campus London | UK | 2010 | Research on acceleration of data-intensive applications | University | 1 | 6 | 128 gigabytes | unknown | 24 | 1.72364 | 393 seconds | 0 Watts | |
195 | 192 | Sho's Computer | Intel | Intel Core i9-9900K | Sho's house | Tokyo | Japan | 2020 | Desktop Computer | Desktop Computer | 1 | 16 | 32 | custom https://github.com/suzumura/graph500 | 20 | 1.72 | 1.89596 | ||
196 | 193 | Dingus | Convey | HC-1 | Sandia National Laboratories | USA | 1 | 4 | 0 gigabytes | Convey Custom | 28 | 1.71799 | 0 seconds | 0 Watts | |||||
196 | 193 | Wingus | Convey | HC-1 | Sandia National Laboratories | USA | 1 | 4 | 0 gigabytes | Convey Custom | 27 | 1.71799 | 0 seconds | 0 Watts | |||||
198 | 195 | Graphybrid++ | AMD | AMD A10-7850K APU + AMD W9100 Firepro GPU | AMD Research | Austin TX | USA | 2014 | Architecture Testbed | Industry | 1 | 4 | 24 gigabytes | Custom (Hybrid++) | 25 | 1.68555 | 205.9 seconds | 0 Watts | |
199 | 196 | EBD-GoldenBox-Prototype | NVIDIA | Jetson TK1 | Tokyo Institute of Technology | Tokyo | Japan | 2014 | Research | University | 1 | 5 | 2 gigabytes | Custom (OpenMP Degree-aware BFS) | 21 | 1.63503 | 40.153 seconds | 8 Watts | |
200 | 197 | Convey01 | Convey | HC-1 | Bielefeld University CeBiTec | Germany | 1 | 4 | 0 gigabytes | Convey Custom | 28 | 1.61061 | 0 seconds | 0 Watts | |||||
200 | 197 | HC-1d | Convey | Computer Corporation HC-1 | Convey Computer Corporation | USA | 1 | 4 | 0 gigabytes | Convey Custom | 28 | 1.61061 | 0 seconds | 0 Watts | |||||
202 | 199 | Convey2 | Convey | HC-1 | Lawrence Berkeley National Laboratory/NERSC | USA | 1 | 4 | 0 gigabytes | Convey Custom | 28 | 1.50324 | 0 seconds | 0 Watts | |||||
202 | 199 | IBM with ScaleMP/vSMP Foundation | IBM | ScaleMP/vSMP Fundation | LArgonne National Laboratory | USA | 0 | 128 | 0 gigabytes | Reference | 29 | 1.50324 | 0 seconds | 0 Watts | |||||
204 | 201 | Anton's Macbook | Apple | Macbook Pro 2017 LPDDR3 | personal | Boise ID | United States | 2017 | laptop | laptop | 1 | 2 | 16 | custom heavy+dir opt | 20 | 1.4153 | 9.06 | 39 | |
205 | 202 | SuperDragon-1 | unknown | Institute of Computing Technology | Beijing | China | 32 | 384 | 0 gigabytes | Custom | 30 | 1.39586 | 0 seconds | 0 Watts | |||||
206 | 203 | GraphNode | Intel Sever | Intel(R) Xeon(R) CPU E5-2683 v3 | ICT | China | United States | 2015 | University | Scientific research | 1 | 28 | 384 | custom | 25 | 1.3358 | 162 | ||
207 | 204 | Cray XMT | Cray | XMT | Pacific Northwest National Laboratory | USA | 128 | 128 | 0 gigabytes | Optimized | 29 | 1.30997 | 0 seconds | 0 Watts | |||||
208 | 205 | PLX | IBM | PLX | CINECA | Bologna | Italy | 32 | 64 | 0 gigabytes | GPU Customized | 26 | 1.28849 | 0 seconds | 0 Watts | ||||
209 | 206 | EBD-prototype-m | Handmade | Custom (Intel® Core™ i7 @ 3.50GHz crucial@m4 msata 256GB * 16) | Tokyo Institute of Technology | Tokyo | Japan | 2013 | Research | University | 1 | 4 | 16 gigabytes | Custom (OpenMP NUMA-optimized Hybrid Single Graph NVRAM) | 28 | 1.28359 | 10674 seconds | 0 Watts | |
210 | 207 | Cray XMT | Cray | XMT | Sandia National Laboratories | USA | 128 | 128 | 0 gigabytes | Optimized | 29 | 1.25628 | 0 seconds | 0 Watts | |||||
211 | 208 | Anton's Macbook Pro | Apple | laptop | personal | Boise ID | United States | 2017 | personal | research | 1 | 2 | custom_aml_heavy_diropt | 17 | 1.22818 | 0.602644 | |||
212 | 209 | GraphCREST-MBA13 | Apple | Macbook Air 13inch (Intel(R) Core(TM) i5-2557M CPU @ 1.70GHz) / 1node | Chuo University | Tokyo | Japan | 2012 | mathematical optimization | university | 1 | 2 | 8 gigabytes | custom (OpenMP) | 23 | 1.22761 | 69.7941 seconds | 23.6 Watts | |
213 | 210 | Alkindi | Supermicro workstation | Commodity Machine / Intel Xeon E5-2650 (2 Sockets 32 cores) Nvidia Kepler K20c (2 GPUs) | The University of British Columbia | Vancouver BC | Canada | 2012 | Generic | University | 1 | 0 | 256 gigabytes | custom | 28 | 1.20576 | 2860 seconds | 496 Watts | |
214 | 211 | cougarxmt | Cray | XMT | Pacific Northwest National Laboratory | USA | 128 | 128 | 0 gigabytes | Reference | 29 | 1.18112 | 0 seconds | 0 Watts | |||||
214 | 211 | GraphCREST-W12 | AMD | Opteron 6174 2.2GHz | Chuo University | Tokyo | Japan | 1 | 12 | 0 gigabytes | 0 | 1.18112 | 0 seconds | 0 Watts | |||||
216 | 213 | PowerEdge R815 Opteron 6174 | Dell | PowerEdge R815 Opteron 6174 | Infiniband | STE Lab Nagoya University | Nagoya | Japan | 2010 | Geophysics and astrophysics | University | 4 | 192 | 384 gigabytes | unknown | 22 | 1.1623 | 2898.1 seconds | 0 Watts |
217 | 214 | GraphCREST-Intel-NUC | Intel | Intel® NUC Kit DC3217IYE / 1 node | Chuo University | Tokyo | Japan | 2013 | mathematical optimization | university | 1 | 2 | 8 gigabytes | custom (OpenMP) | 23 | 1.08175 | 71.1322 seconds | 20.1 Watts | |
218 | 215 | graphstorm | Cray | XMT | Sandia National Laboratories | USA | 128 | 128 | 0 gigabytes | Reference | 29 | 1.07374 | 0 seconds | 0 Watts | |||||
219 | 216 | GraphCREST-Xperia-Z1-SO-01F | Sony | Android Smartphone (Qualcomm Snapdragon 800 MSM8974 2.2GHz) | Kyushu University | Fukuoka | Japan | 2013 | Smartphone | Smartphone | 1 | 4 | 2 gigabytes | Custom (OpenMP Android-NDK Degree-aware BFS) | 20 | 1.033 | 19.149 seconds | 4.393 Watts | |
220 | 217 | ScaleMP/vSMP Foundation | IBM | ScaleMP/vSMP Fundation | LArgonne National Laboratory | USA | 0 | 128 | 0 gigabytes | Reference | 30 | 1.03079 | 0 seconds | 0 Watts | |||||
221 | 218 | Hyperion | Dell Intel Supermicro QLogic Cisco Mellanox DDN Sun LSI and RedHat | Lawrence Livermore National Laboratory | USA | 64 | 512 | 0 gigabytes | Custom | 31 | 0.99858 | 0 seconds | 0 Watts | ||||||
222 | 219 | Cray XT5-HE | Cray | XT5-HE | Oak Ridge National Laboratory | USA | 18688 | 224256 | 0 gigabytes | 30 | 0.966368 | 0 seconds | 0 Watts | ||||||
223 | 220 | tara | IBM | Server x iDataPlex | InfiniBand (QDR) | University of Maryland Baltimore County | Baltimore MD | USA | 2009 | Mathematics and statistics computer science atmospheric physics computational chemistry mechanical engineering and more | University | 86 | 688 | 2088 gigabytes | Modified reference implementation | 31 | 0.946 | 135.525 seconds | 0 Watts |
224 | 221 | Matterhorn | Cray Inc. | XMT2 | Seastar | Swiss National Supercomputing Center | Lugano | Switzerland | 2011 | General purpose scientific computing | University | 64 | 64 | 2048 gigabytes | Green500-modified Cray version | 31 | 0.897 | 2203 seconds | 19557 Watts |
225 | 222 | Matterhorn | Cray | XMT | CSCS | Switzerland | 64 | 64 | 0 gigabytes | Reference | 31 | 0.880468 | 0 seconds | 0 Watts | |||||
226 | 223 | Cray XMT2 | Cray | XMT2 | CSCS | Switzerland | 64 | 64 | 0 gigabytes | 31 | 0.858993 | 0 seconds | 0 Watts | ||||||
227 | 224 | Alkindi-Jetson | Nvidia | Jetson | The University of British columbia | Vancouver | Canada | 2015 | Research | University | 0 | 0 | 0 gigabytes | Custom | 22 | 0.850828 | 37.54 seconds | 6.346 Watts | |
228 | 225 | Minerva | Intel | Westmere 2.66 GHz | University of Warwick | UK | 1 | 12 | 0 gigabytes | Reference | 26 | 0.837519 | 0 seconds | 0 Watts | |||||
229 | 226 | AMD-APU | AMD | AMD A10-7850K APU | AMD Research | Austin TX | USA | 2014 | Architecture Testbed | Industry | 1 | 0 | 16 gigabytes | Custom (Hybrid++) | 24 | 0.827 | 89.6 seconds | 0 Watts | |
230 | 227 | GraphCREST-Xperia-A-SO-04E | Sony | Android Smartphone (Qualcomm Snapdragon S4 Pro APQ8064 1.5GHz) | Kyushu University | Fukuoka | Japan | 2013 | Smartphone | Smartphone | 1 | 4 | 2 gigabytes | Custom (OpenMP Android-NDK Degree-aware BFS) | 20 | 0.739 | 26.664 seconds | 3.2074 Watts | |
231 | 228 | Brokoli Cluster | Inspur | Intel Xeon E5-2670v3 2.3GHz | FDR Infiniband | Telkom University | Bandung | Indonesia | 2017 | Academic | Scientific computing | 1 | 48 | 128 gigabytes | Graph500-2.1.4 omp-csr | 20 | 0.7128 | 0.467 seconds | Watts |
232 | 229 | Erdos | unknown | Oak Ridge National Laboratory | USA | 0 | 0 | 0 gigabytes | Reference | 29 | 0.697932 | 0 seconds | 0 Watts | ||||||
232 | 229 | Intel Westmere E7-4870 2.4GHz | Intel | E7-4870 2.4GHz | Intel Research | USA | 1 | 40 | 0 gigabytes | 27 | 0.697932 | 0 seconds | 0 Watts | ||||||
234 | 231 | Intel Westmere X5670 2.93 | Intel | Westmere X5670 2.93 IB network | Intel Corporation | USA | 128 | 256 | 0 gigabytes | Reference | 26 | 0.660351 | 0 seconds | 0 Watts | |||||
235 | 232 | Hyperion + FusionIO | unknown | Lawrence Livermore National Laboratory | USA | 64 | 512 | 0 gigabytes | Custom/NAND FLASH | 36 | 0.601295 | 0 seconds | 0 Watts | ||||||
236 | 233 | GraphCREST-NEXUS7-2013 | ASUS | Android Tablet (Qualcomm Snapdragon S4 Pro 1.5GHz) | Tokyo Institute of Technology | Tokyo | Japan | 2013 | Research | University | 1 | 4 | 2 gigabytes | Custom (OpenMP Android NDK) | 20 | 0.534713 | 33.3676 seconds | 4.1248 Watts | |
237 | 234 | Nehalem X5570 | Intel | Nehalem X5570 | Sandia National Laboratories | USA | 0 | 512 | 0 gigabytes | Reference | 28 | 0.512175 | 0 seconds | 0 Watts | |||||
238 | 235 | Aziz | Fujitsu | Cluster | IB (QDR) Intel True Scale | king abdulaziz university | Jeddah | Saudi Arabia | 2015 | Genomics Weather | University | 0 | 0 | 0 gigabytes | MPI-Replicated | 16 | 0.478866 | 0.170506 seconds | 0 Watts |
239 | 236 | GraphCREST-Xperia-A-SO-04E | Chuo University | 4 | 20 | 0.478 | 3.12071 Watts | ||||||||||||
240 | 237 | Vedur | HP | AMD Opteron 6276 CPU (32 cores) @ 2.3GHz | 4x QDR Infiniband | University of Tartu | Tartu | Estonia | 2011 | University | General | 8 | 512 | Reference | 21 | 0.455214 | 10.284522 | ||
241 | 238 | ultraviolet | unknown | Sandia National Laboratories | USA | 0 | 32 | 0 gigabytes | Custom | 29 | 0.418759 | 0 seconds | 0 Watts | ||||||
242 | 239 | Viridis HPC@Uni.lu | Boston | Boston Viridis - ARM Microservers 2U | Ethernet 10G | University of Luxembourg | Campus Belval Esch-sur-Alzette | Luxembourg | 2013 | Computer Science research | University | 32 | 128 | 128 gigabytes | Reference implementation graph500_mpi_replicated | 21 | 0.389643 | 8.71395 seconds | 206.38 Watts |
243 | 240 | momcom-max | Apple | MacBook Pro with M1 Max | NTT Musashino R&D Center | Tokyo | Japan | 2022 | Industry | Research | 1 | 10 | 32 | Reference | 26 | 0.29082 | 21.6326 | ||
244 | 241 | Owens | unknown | The Ohio State University | USA | 0 | 0 | 0 gigabytes | Reference | 26 | 0.257698 | 0 seconds | 0 Watts | ||||||
245 | 242 | Trestles | Appro | 1U-1143H Cluster Opteron 8C 2.4Ghz | Infiniband QDR | UC San Diego | San Diego CA | USA | 144 | 4608 | 0 gigabytes | Custom/NAND FLASH | 36 | 0.236223 | 0 seconds | 0 Watts | |||
246 | 243 | Okorok | Dell | PowerEdge C1100 | Home | Yekaterinburg | Russia | 2013 | Natural language processing | Personal | 1 | 8 | 48 gigabytes | Reference OpenMP | 17 | 0.228 | 0.208 seconds | 0 Watts | |
247 | 244 | LANL-LARP-hbchen-001 | HP | Z840 | Local | Los Alamos National Lab | Los Alamos New Mexico | United States | 2016 | Government | Metadata indexing | 1 | 24 | 64 gigabytes | Reference | 27 | 0.179422 | 47.2368 seconds | Watts |
248 | 245 | Knot | unknown | UC Santa Barbara | Santa Barabara CA | USA | 0 | 64 | 0 gigabytes | Custom/NAND FLASH | 30 | 0.171799 | 0 seconds | 0 Watts | |||||
249 | 246 | GraphCREST-Tegra3 | Chuo University | 4 | 20 | 0.154 | 2.402 Watts | ||||||||||||
250 | 247 | larp-hbchen-002 | HP | Z840 | NO | UltraScale Systems research Center Los Alamos National lab | Los Alamos New Mexcio | United States | 2016 | Government Lab | Storage system metadata indexing research | 1 | 24 | gigabytes | graph500_reference_bfs_sssp | 29 | 0.13869 | 56.219773 seconds | Watts |
251 | 248 | GraphCREST-NEXUS10 | Chuo University | 2 | 20 | 0.119 | 3.69 Watts | ||||||||||||
252 | 249 | Twizzle | Dell | Optiplex 9010 Intel Ivy Bridge Core i7 3770 | Sandia National Laboratories | New Mexico USA | USA | 2012 | Architectural Testbed | Government | 1 | 4 | 16 gigabytes | Reference OpenMP-CSR | 23 | 0.0992 | 39.18 seconds | 0 Watts | |
253 | 250 | Kraken | Cray | Lawrence Livermore National Laboratory | USA | 0 | 6128 | 0 gigabytes | Custom/NAND FLASH | 31 | 0.0966368 | 0 seconds | 0 Watts | ||||||
254 | 251 | Xeon Server | Intel | Intel(R) Xeon(R) Gold 5117 CPU | Custom | 2019 | hangzhou | China | 2019 | Research | Research | 1 | 14 | 128 | Custom | 20 | 0.0758571 | 7.40217 | |
255 | 252 | Caribou | unknown | Link Analytics | USA | 0 | 0 | 0 gigabytes | Reference | 27 | 0.0751619 | 0 seconds | 0 Watts | ||||||
256 | 253 | Bertha | Supermicro / Saratoga Speed | Quad-socket Intel E7-4870v2 server w/ Saratoga Speed Altamont XP flash arrays | Lawrence Livermore National Laboratory | Livermore CA | USA | 2014 | Research | Government | 1 | 60 | 2048 gigabytes | Custom | 37 | 0.0627 | 0 seconds | 0 Watts | |
257 | 254 | Leviathan | Dell | FusionIO | Lawrence Livermore National Laboratory | USA | 1 | 40 | 0 gigabytes | Custom/NAND FLASH | 36 | 0.0536871 | 0 seconds | 0 Watts | |||||
257 | 254 | Appro | Dell | Lawrence Livermore National Laboratory | USA | 1 | 32 | 0 gigabytes | 34 | 0.0536871 | 0 seconds | 0 Watts | |||||||
259 | 256 | Simons PC | Intel | Desktop Computer | Home | Aachen | Germany | 2012 | Research | Private | 1 | 4 | 8 gigabytes | Reference omp-csr | 23 | 0.05199 | 19.12 seconds | ||
260 | 257 | Neumann | AMD | UC Santa Barbara | Santa Barabara CA | USA | 0 | 32 | 0 gigabytes | Custom | 26 | 0.0322123 | 0 seconds | 0 Watts | |||||
261 | 258 | Scott Beamer's iPad | Apple | iPad 3 | none | UC Berkeley | Berkeley CA | USA | 2012 | research | university | 1 | 2 | 1 gigabytes | direction-optimizing BFS | 14 | 0.0304 | 13.211 seconds | 0 Watts |
262 | 259 | Chaos (LABOS T-Rex) | DALCO AG | DALCO r2164i5 (2x Intel Xeon E5-2630v3 2x 6TB SATAIII) | 40G Mellanox | Ecole Polytechnique Fédérale de Lausanne | Lausanne | Switzerland | 2015 | University | Computer Science | 8 | 128 | 128 | Chaos (http://labos.epfl.ch/chaos) | 33 | 0.0264 | 46 minutes 12 seconds | 1824 |
263 | 260 | Gordon | Appro | Xtreme-X GreenBlade GB512X Xeon E5-2670 8C 2.600GHz | Infiniband QDR | San Diego Supercomputing Center | San Diego CA | USA | 7 | 84 | 0 gigabytes | Reference | 29 | 0.0214748 | 0 seconds | 0 Watts | |||
264 | 261 | Kichakato Kizito 1 | Acer | Aspire A315 Laptop | Mobile | Nairobi | Kenya | 2020 | Personal Computing | Science and Technology | 1 | 2 | 4 | Reference | 16 | 0.01587 | 0.245339 |
SSSP results, June 2024
Rank | previous rank | Machine | Vendor | Type | Network | Installation Site | Location | Country | Year | Application | Usage | Number of nodes | Number of cores | Memory | Implementation | Scale | GTEPS | construction time | Power |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1 | 1 | Wuhan Supercomputer | HUST | Kunpeng 920+Tesla A100 | Custom | Wuhan Supercomputing Center | Wuhan | China | 2023 | Research | Academic | 252 | 6999552 | 584640 | Custom (Graph Processing System "YITU" implemented by Jin Zhao Yu Zhang et al.) | 41 | 15335.9 | 16285 | |
2 | 2 | Pengcheng Cloudbrain-II | HUST-Pengcheng Lab-HUAWEI | Kunpeng 920+Ascend 910 | Custom | Pengcheng Lab | ShenZhen | China | 2022 | Research | Academic | 488 | 93696 | 999424 | Custom (Graph Processing System "YITU" implemented by Jin Zhao Yujian Liao Yu Zhang et al.) | 40 | 11529.7 | 57218 | |
3 | 3 | Supercomputer Fugaku | Fujitsu | Fujitsu A64FX | Tofu Interconnect D | RIKEN Center for Computational Science (R-CCS) | Kobe Hyogo | Japan | 2020 | Various scientific and industrial fields | Academic and industry | 82944 | 3981312 | 2654208 | Custom | 39 | 2126.45 | 322 | |
4 | 4 | Tianhe Exascale Prototype Upgrade System | National University of Defense Technology | FT-2000V | TH Express-3 | National Supercomputer Center in Tianjin | Tianjin | China | 2021 | Industry | Research | 2048 | 131072 | 262144 | Custom_MST_FT-topo_TianheGraph @(Gan Xinbiao‘s implementation) | 34 | 2054.35 | 114.72 | |
5 | 5 | SuperMUC-NG | Lenovo | ThinkSystem SD530 Xeon Platinum 8174 24C 3.1GHz Intel Omni-Path | Leibniz Rechenzentrum | Garching | Germany | 2018 | Academic | Research | 4096 | 196608 | 393216 | custom-aml-heavy-diropt | 37 | 1053.93 | 310.72 | ||
6 | 6 | NERSC Cori - 1024 haswell partition | Cray | XC40 | Aries | NERSC/LBNL | DOE/SC/LBNL/NERSC | United States | 2017 | Government | basic science and simulation | 1024 | 32768 | 133376 | custom-aml-heavy-diropt | 36 | 558.833 | 446.475 | |
7 | 7 | Nurion | Cray | CS500 Intel Xeon Phi 7250 68C 1.4GHz Intel Omni-Path | Intel Omni-Path | Korea Institute of Science and Technology Information | Daejeon | Korea Republic Of | 2018 | Research | Various scientific and instudrial fields | 1024 | 65536 | 98304 | custom-aml-heavy-diropt | 36 | 337.239 | 858 | |
8 | 8 | NERSC Cori - 512 KNL partition | Cray | XC40 | Aries | NERSC/LBNL | DOE/SC/LBNL/NERSC | United States | 2017 | Government | basic science and simulation | 512 | 32768 | 57344 | custom-aml-heavy-diropt | 35 | 229.188 | 1188 | |
9 | 9 | Lise | Atos | Bull Intel Cluster Intel Xeon Platinum 9242 48C 2.3GHz | Intel Omni-Path | Zuse Institute Berlin (ZIB) | Berlin | Germany | 2019 | Research | Academic | 1270 | 121920 | 502272 | Custom 2D-delta-stepping (ZIB) | 38 | 197.7 | 240 | |
10 | 10 | Undisclosed Cray XE6 | Cray | XE6 | Gemini | National Computing Facility | University | United States | 2013 | University | Research | 512 | 16384 | custom-aml-heavy-diropt | 34 | 134.173 | 400.206 | ||
11 | 11 | Undisclosed Cray XE6 | Cray | MPP | Gemini | National Computing Facility | University | United States | 2013 | University | Research | 512 | 8192 | 32768 | newreference+heavy | 31 | 12.88 | 27.5 | 0 |
12 | 12 | Xeon Server | Dell | 4 x Intel(R) Xeon(R) Gold 5115 CPU @ 2.40GHz | industry | BoiseID | United States | 2017 | Research | Research | 1 | 40 | heavy+newreference2 | 23 | 3.09 | 4.68 | |||
13 | 13 | RAIDEN CPU subsystem | Fujitsu | Intel Skylake cluster (Intel(R) Xeon(R) Gold 6148 CPU @ 2.40GHz) | Infiniband EDR | RIKEN AIP | Tokyo | Japan | 2016 | scientific research | Research | 32 | 1024 | 12288 | custom | 33 | 2.66 | 107 | 0 |
14 | 14 | Alkindi-CPU | Dell | Single node commodity server | The University of British Columbia | Vancouver | Canada | 2015 | University | Research | 1 | 28 | 256 | Custom | 22 | 1.23 | 3.06 | ||
15 | 15 | Xeon Server | Dell | 4 x Intel(R) Xeon(R) Gold 5115 CPU @ 2.40GHz | Industry | BoiseID | United States | 2017 | Research | Research | 1 | 32 | heavy-newreference | 21 | 0.985 | 3.78 | |||
16 | 16 | Tellico | IBM | IBM Power System AC922 | University of Tennessee Knoxville | Knoxville | United States | 2019 | University | Research | 2 | 64 | 300 | Custom | 27 | .69296 | 12 | ||
17 | 17 | University of Notre Dame cluster | Dell | cluster | Ethernet | University of Notre Dame | University of Notre Dame | United States | 2016 | University | Research | 8 | 128 | 2048 | newreference+heavy | 27 | 0.656085 | 60.0327 | |
18 | 18 | Anton's Macbook Pro | Apple | laptop | personal | Boise ID | United States | 2017 | personal | research | 1 | 2 | custom_aml_heavy_diropt | 17 | 0.42 | 0.602644 | 40 | ||
19 | 19 | Tellico | IBM | Power9 | University of Tennessee Knoxville | Knoxville | United States | 2019 | University | High Performance Computing | 2 | 55 | 300 | Reference | 28 | 0.337443 | 18 | ||
20 | 20 | Tellico | IBM | IBM Power System AC922 | University of Tennessee Knoxville | Knoxville | United States | 2019 | University | Research | 2 | 64 | 256 | Custom | 28 | 0.321862 | 21 | ||
21 | 21 | Alkindi28 | Dell | Commodity Machine / Intel Xeon E7-4870 v2 (4 Sockets 60 cores 120 hardware threads) | The University of British Columbia | Vancouver BC | Canada | 2015 | Research | University | 1 | 60 | 1536 | custom NUMA-aware - Joint work of UBC and RIKEN-AIP | 30 | 0.28 | 2173 | 0 | |
22 | 22 | Alkindi27 | Dell | Commodity Machine / Intel Xeon E5-2695 v3 (2 Sockets 28 cores 56 hardware threads) | The University of British Columbia | Vancouver BC | Canada | 2015 | University | Research | 1 | 56 | 256 | Custom | 27 | 0.1972 | 102.35 | 0 | |
23 | 23 | RAIDEN MMS Subsystem | Fujitsu | single node commodity server (Intel(R) Xeon(R) Platinum 8160M CPU @ 2.10GHz) | RIKEN AIP | Tokyo | Japan | 2018 | Research | scientific research | 1 | 96 | 6144 | custom (NUMA-aware Totem from the University of British Columbia) | 32 | 0.17 | 12109 | ||
24 | 24 | Alkindi28 | Dell | Commodity Machine / Intel Xeon E7-4870 v2 (4 Sockets 60 cores 120 hardware threads) | The University of British Columbia | Vancouver BC | Canada | 2015 | University | Research | 1 | 60 | 1536 | Custom | 29 | 0.1567 | 527.32 | 0 | |
25 | 25 | Alkindi28 | Dell | Commodity Machine / Intel Xeon E7-4870 v2 (4 Sockets 60 cores 120 hardware threads) | The University of British Columbia | Vancouver BC | Canada | 2015 | University | Research | 1 | 60 | 1536 | Custom | 30 | 0.1536 | 1086.271 | 0 | |
26 | 26 | Alkindi27 | Dell | Commodity Machine / Intel Xeon E5-2695 v3 (2 Sockets 28 cores 56 hardware threads) | The University of British Columbia | Vancouver BC | Canada | 2015 | University | Research | 1 | 28 | 256 | Custom | 28 | 0.1375 | 278.11 | 0 | |
27 | 27 | Alkindi28-CPU | Dell | Commodity Machine / Intel Xeon E7-4870 v2 (4 Sockets 60 cores) | The University of British Columbia | Vancouver | Canada | 2015 | University | Research | 1 | 60 | 1536 | Custom | 31 | 0.134 | 17585 | ||
28 | 28 | Anton's Macbook | Apple | Macbook Pro 2017 LPDDR3 | personal | Boise ID | United States | 2017 | laptop | laptop | 1 | 2 | 16 | custom heavy+dir opt | 20 | 0.123492 | 9.06 | 39 | |
29 | 29 | momcom-max | Apple | MacBook Pro with M1 Max | NTT Musashino R&D Center | Tokyo | Japan | 2022 | Industry | Research | 1 | 10 | 32 | Reference | 26 | 0.0664295 | 22 | ||
30 | 30 | LANL-LARP-hbchen-001 | HP | Z840 | Local | Los Alamos National Lab | Los Alamos New Mexico | United States | 2016 | Government | Metadata indexing | 1 | 24 | 64 | Reference | 27 | 0.0547379 | 47.2368 | |
31 | 31 | larp-hbchen-002 | HP | Z840 | NO | UltraScale Systems research Center Los Alamos National lab | Los Alamos New Mexcio | United States | 2016 | Government Lab | Storage system metadata indexing research | 1 | 24 | graph500_reference_bfs_sssp | 29 | 0.0519326 | 56.219773 | ||
32 | 32 | Chaos (LABOS T-Rex) | DALCO AG | DALCO r2164i5 (2x Intel Xeon E5-2630v3 2x 6TB SATAIII) | 40G Mellanox | Ecole Polytechnique Fédérale de Lausanne | Lausanne | Switzerland | 2015 | University | Computer Science | 8 | 128 | 128 | Chaos (http://labos.epfl.ch/chaos) | 33 | 0.0271 | 46 minutes 12 seconds | 1824 |
33 | 33 | Kichakato Kizito 1 | Acer | Aspire A315 Laptop | Mobile | Nairobi | Kenya | 2020 | Personal Computing | Science and Technology | 1 | 2 | 4 | Reference | 15 | 0.01605 | 1 | ||
34 | 34 | Brokoli Cluster | Inspur | Intel Xeon E5-2670v3 2.3GHz | FDR Infiniband | Telkom University | Bandung | Indonesia | 2017 | Academic | Scientific computing | 1 | 48 | 128 | Reference | 22 | 0.00532283 | 31.0229 | 0 |