June 2021 Green: SMALL DATA
| Rank | MTEPS/W | Site | Machine | Graph500 rank | Scale | GTEPS | Nodes | Cores | Power |
|---|---|---|---|---|---|---|---|---|---|
| 1 | 6028.85 | UPenn | ENIAD | 26 | 26 | 783.750300 | 1 | 16 | 130 |
| 2 | 4724.30 | National Supercomputer Center in Tianjin | Tianhe Exa-node Prototype | -1 | 25 | 779.510000 | 1 | 64 | 165 |
| 3 | 2099.09 | SMARCO && ICT | High Throughput Computer (+Nvidia V100) | 45 | 26 | 319.061000 | 1 | 24 | 152 |
| 4 | 1926.52 | SMARCO && ICT | High Throughput Computer (+Nvidia V100) | -1 | 26 | 271.640000 | 1 | 1 | 141 |
| 5 | 1830.31 | SMARCO | High Throughput Computer(+GPU) | 52 | 26 | 237.940000 | 1 | 58 | 130 |
| 6 | 1400.19 | University of Wisconsin-Madison | Graph-WICIL-Node0 | 79 | 23 | 79.531000 | 1 | 18 | 56.8 |
| 7 | 1165.71 | IBM Moscow && MSU Alex Kolganov | IBM POWER8+ Tesla P100 | 55 | 26 | 204.000000 | 1 | 66 | 175 watts Watts |
| 8 | 815.68 | forsite-company.ru && NVidia LTD | TitanX_forsite | 67 | 26 | 132.140000 | 1 | 28 | 162 Watts |
| 9 | 597.86 | SMARCO | Xeon(R) Gold 5118+Tesla P100 | 68 | 26 | 131.530000 | 1 | 24 | 220 |
| 10 | 540.94 | Keldysh Institute of Applied Mathematics (Russian Academy of Sciences) | Titan_Keldysh | 72 | 25 | 114.680000 | 1 | 20 | 212 Watts |
| 11 | 445.92 | George Washington University | Colonial | 70 | 20 | 122.182000 | 1 | 12 | 274 Watts |
| 12 | 243.42 | CODE@TACC (Texas Advanced Computing Center Summer High School STEM Program) | Monty Pi-thon | 91 | 26 | 35.825400 | 32 | 128 | 147.175 Watts |
| 13 | 235.15 | Kyushu University | GraphCREST-Xperia-Z1-SO-01F | 195 | 20 | 1.033000 | 1 | 4 | 4.393 Watts |
| 14 | 230.40 | Kyushu University | GraphCREST-Xperia-A-SO-04E | 206 | 20 | 0.739000 | 1 | 4 | 3.2074 Watts |
| 15 | 204.38 | Tokyo Institute of Technology | EBD-GoldenBox-Prototype | 175 | 21 | 1.635030 | 1 | 5 | 8 Watts |
| 16 | 180.76 | Kyushu University | GraphCREST-Xperia-A-SO-04E | -1 | 21 | 0.588000 | 1 | 4 | 3.253 Watts |
| 17 | 171.77 | Kyushu University | GraphCREST-Xperia-Z1-SO-01F | -1 | 21 | 0.909000 | 1 | 4 | 5.292 Watts |
| 18 | 170.64 | University of Tsukuba | kitty6 | 93 | 25 | 35.180000 | 1 | 8 | 206.164 Watts |
| 19 | 153.17 | Chuo University | GraphCREST-Xperia-A-SO-04E | 215 | 20 | 0.478000 | 1 | 4 | 3.12071 Watts |
| 20 | 145.58 | RSC Moscow && MSU Alex Kolganov | RSC KNL | 95 | 29 | 31.300000 | 1 | 68 | 215 Watts |
| 21 | 134.07 | The University of British columbia | Alkindi-Jetson | 203 | 22 | 0.850828 | 0 | 0 | 6.346 Watts |
| 22 | 129.63 | Tokyo Institute of Technology | GraphCREST-NEXUS7-2013 | 212 | 20 | 0.534713 | 1 | 4 | 4.1248 Watts |
| 23 | 87.12 | Kyushu University | GraphCREST-SandybridgeEP-2.4GHz | -1 | 27 | 41.802000 | 1 | 32 | 479.8 Watts |
| 24 | 81.54 | Keldysh Institute of Applied Mathematics (Russian Academy of Sciences) | Titan_Keldysh | -1 | 27 | 10.600000 | 1 | 6 | 130 Watts |
| 25 | 81.49 | Kyushu University | GraphCREST-SandybridgeEP-2.4GHz | 86 | 27 | 42.450000 | 1 | 32 | 520.9 Watts |
| 26 | 73.57 | University of Tsukuba | kitty6 | -1 | 25 | 17.207000 | 1 | 8 | 233.894 Watts |
| 27 | 68.74 | Kyushu University | GraphCREST-Sandybridge-EP-2.7GHz | 85 | 27 | 45.710000 | 1 | 32 | 665 Watts |
| 28 | 64.11 | Chuo University | GraphCREST-Tegra3 | 224 | 20 | 0.154000 | 1 | 4 | 2.402 Watts |
| 29 | 53.82 | Chuo University | GraphCREST-Intel-NUC | 193 | 23 | 1.081750 | 1 | 2 | 20.1 Watts |
| 30 | 53.47 | Chuo University | GraphCREST-Mac-mini | 167 | 24 | 1.941000 | 1 | 4 | 36.3 Watts |
| 31 | 52.02 | Chuo University | GraphCREST-MBA13 | 188 | 23 | 1.227610 | 1 | 2 | 23.6 Watts |
| 32 | 51.61 | Chuo University | GraphCREST-Retina15 | 166 | 24 | 1.987000 | 1 | 4 | 38.5 Watts |
| 33 | 45.43 | Chuo University | GraphCREST-4Way-SandybridgeEP-2.4GHz | 96 | 27 | 29.034000 | 1 | 64 | 639.1 Watts |
| 34 | 41.01 | Chuo University | GraphCREST-Sandybridge-EP-2.7GHz | 94 | 27 | 31.648000 | 1 | 64 | 771.71 Watts |
| 35 | 39.29 | Changsha China | TH-IVB-FEP/C | 132 | 26 | 9.744020 | 1 | 24 | 248 Watts |
| 36 | 36.29 | personal | Anton's Macbook | 180 | 20 | 1.415300 | 1 | 2 | 39 |
| 37 | 33.59 | The University of British Columbia | Alkindi27-hybrid | 102 | 29 | 19.468500 | 1 | 28 | 579.6 Watts |
| 38 | 32.25 | Chuo University | GraphCREST-NEXUS10 | 226 | 20 | 0.119000 | 1 | 2 | 3.69 Watts |
| 39 | 30.95 | The University of British Columbia | Alkindi27-hybrid | -1 | 29 | 17.795900 | 0 | 0 | 575 Watts |
| 40 | 30.41 | Tokyo Institute of Technology | ULP2012 ultra-green oil-cooled server : GraphCREST | 150 | 24 | 5.123720 | 1 | 12 | 168.5 Watts |
| 41 | 29.47 | Chuo University | GraphCREST-Sandybridge-EP-2.7GHz | -1 | 29 | 22.983000 | 1 | 64 | 779.88 Watts |
| 42 | 20.58 | Chuo University | GraphCREST-SBep2.0 | 145 | 25 | 6.142000 | 1 | 16 | 298.44 Watts |
| 43 | 18.68 | Chuo University | GraphCREST-SBep2.9 | 140 | 25 | 6.777000 | 1 | 16 | 362.79 Watts |
| 44 | 17.39 | Chuo University | GraphCREST-4waySBep2.4 | 125 | 26 | 11.115000 | 1 | 32 | 639 Watts |
| 45 | 10.95 | Chuo University | GraphCREST-Wex40 | 126 | 26 | 11.061000 | 1 | 40 | 1010.13 Watts |
| 46 | 10.25 | Tokyo Institute of Technology | MEM-CREST Node #1 | 162 | 29 | 3.111710 | 1 | 16 | 303.6 Watts |
| 47 | 2.43 | The University of British Columbia | Alkindi | 189 | 28 | 1.205760 | 1 | 0 | 496 Watts |
| 48 | 1.89 | University of Luxembourg | Viridis HPC@Uni.lu | 218 | 21 | 0.389643 | 32 | 128 | 206.38 Watts |
| 49 | 0.20 | Swiss National Supercomputing Center | Todi | 107 | 29 | 15.598300 | 272 | 8160 | 79012 Watts |
June 2021 Green: BIG DATA
| Rank | MTEPS/W | Site | Machine | Graph500 rank | Scale | GTEPS | Nodes | Cores | Power |
|---|---|---|---|---|---|---|---|---|---|
| 1 | 4385.24 | National Supercomputer Center in Tianjin | Tianhe Exa-node Prototype | 25 | 30 | 811.270000 | 1 | 64 | 185 |
| 2 | 2057.08 | University of Pennsylvania | ENIAD | 37 | 30 | 436.100000 | 1 | 8 | 212 |
| 3 | 678.26 | SMARCO && ICT | High Throughput Computer (+ NVidia V100) | 62 | 30 | 163.461000 | 1 | 48 | 241 |
| 4 | 638.22 | SMARCO && ICT | High Throughput Computer(+Nvidia V100) | 65 | 30 | 146.790000 | 1 | 1 | 230 watts |
| 5 | 326.48 | SMARCO | High Throughput Computer | 90 | 30 | 39.830000 | 1 | 40 | 122 |
| 6 | 282.70 | SMARCO | High Throughput Computer | -1 | 30 | 34.490000 | 1 | 40 | 122 watts |
| 7 | 177.45 | IBM Moscow && MSU Alex Kolganov | IBM Power8+ Tesla P100 | 89 | 30 | 41.700000 | 1 | 66 | 235 Watts |
| 8 | 165.97 | SMARCO | High Throughput Computer | -1 | 30 | 22.240000 | 1 | 40 | 134 |
| 9 | 163.96 | SMARCO | High Throughput Computer | -1 | 30 | 21.970000 | 1 | 40 | 134 |
| 10 | 112.41 | Alex Kolganov | PowerfullServer | 105 | 30 | 16.300000 | 1 | 18 | 145 Watts Watts |
| 11 | 66.00 | IBM Moscow & MSU Alex Kolganov | IBM POWER8+ | 114 | 30 | 13.200000 | 1 | 10 | 200 watts Watts |
| 12 | 62.93 | Kyushu University | GraphCREST-SandybridgeEP-2.4GHz | -1 | 30 | 31.325000 | 1 | 32 | 497.8 Watts |
| 13 | 61.48 | Kyushu University | GraphCREST-SandybridgeEP-2.4GHz | -1 | 30 | 28.614000 | 1 | 32 | 465.4 Watts |
| 14 | 51.95 | Kyushu University | GraphCREST-Huawei | 82 | 32 | 59.900000 | 1 | 60 | 1153.1 Watts |
| 15 | 48.28 | Kyushu University | GraphCREST-Sandybridge-EP-2.7GHz | -1 | 30 | 31.950000 | 1 | 32 | 661.7 Watts |
| 16 | 44.42 | Kyushu University | GraphCREST-Huawei | -1 | 32 | 55.741000 | 1 | 60 | 1255 Watts |
| 17 | 44.35 | Tokyo Institute of Technology | EBD-RH2288V3 | 109 | 32 | 14.856600 | 1 | 36 | 334.985 Watts |
| 18 | 36.86 | George Washington University | Colonial0 | 115 | 30 | 12.900000 | 1 | 28 | 350 Watts |
| 19 | 35.21 | Tokyo Institute of Technology | GraphCREST-Custom #1 | 112 | 31 | 13.796300 | 1 | 32 | 391.825 Watts |
| 20 | 34.57 | National Institute of Advanced Industrial Science and Technology (AIST) | NEDO-AIRC-NAIROBI-UV | 63 | 33 | 161.215000 | 1 | 256 | 4663.56 Watts |
| 21 | 28.88 | Tokyo Institute of Technology | MEM-CREST Node #2 | 137 | 30 | 7.984170 | 1 | 16 | 276.507 Watts |
| 22 | 23.68 | SGI | SGI UV 300 | -1 | 33 | 99.794000 | 1 | 256 | 4214 Watts |
| 23 | 22.36 | The University of British Columbia | Alkindi26-hybrid | 130 | 30 | 10.307900 | 1 | 20 | 461 Watts |
| 24 | 20.81 | Kyushu University | GraphCREST-Huawei | -1 | 33 | 29.292400 | 1 | 60 | 1407.31 Watts |
| 25 | 20.49 | HP Ft. Collins | hawk040os1 | 69 | 33 | 128.336000 | 1 | 240 | 6264 Watts |
| 26 | 18.90 | The University of British Columbia | Alkindi28-CPU | 124 | 32 | 11.167800 | 1 | 60 | 591.024 Watts |
| 27 | 17.24 | Kyushu University | GraphCREST-Bulldozer | 113 | 31 | 13.633000 | 1 | 64 | 790.8 Watts |
| 28 | 14.92 | HPE Ft. Collins | Hawk028os1 | 66 | 33 | 141.900000 | 1 | 288 | 9508 Watts |
| 29 | 14.06 | Tokyo Institute of Technology | TSUBAME-KFC | 73 | 32 | 104.310000 | 32 | 384 | 7419.85 Watts |
| 30 | 13.90 | The University of British Columbia | Alkindi28 | -1 | 31 | 8.370000 | 1 | 60 | 602.06 Watts |
| 31 | 12.48 | The Institute of Statistical Mathematics | ismuv2k2 | -1 | 32 | 131.427000 | 1 | 640 | 10530 Watts |
| 32 | 10.86 | The University of British Columbia | Alkindi26-CPU | 156 | 30 | 4.247120 | 1 | 20 | 391 Watts |
| 33 | 8.55 | RIKEN Center for Computational Science (R-CCS) | Supercomputer Fugaku | -1 | 40 | 70980.000000 | 92160 | 4423680 | 8299736 |
| 34 | 8.39 | Plexus Engineering | HYPATIA | 77 | 33 | 88.112000 | 32 | 256 | 10498.2 Watts |
| 35 | 6.88 | RIKEN Center for Computational Science (R-CCS) | Supercomputer Fugaku | 1 | 41 | 102956.000000 | 158976 | 7630848 | 14961115 |
| 36 | 5.41 | Forschungszentrum Juelich (FZJ) | JUQUEEN | -1 | 38 | 5848.000000 | 16384 | 262144 | 1081340 Watts |
| 37 | 5.29 | Tokyo Institute of Technology | GraphCREST-7048GR-TR | 169 | 31 | 1.777880 | 1 | 36 | 336.018 Watts |
| 38 | 4.42 | Argonne National Laboratory | DOE/SC/Argonne National Laboratory Mira | -1 | 40 | 14328.000000 | 49152 | 786432 | 3244030 Watts |
| 39 | 4.35 | Tokyo Institute of Technology | EBD-RH5885v2 | 161 | 30 | 3.674750 | 1 | 48 | 844.337 Watts |
| 40 | 4.30 | Lawrence Livermore National Laboratory | DOE/NNSA/LLNL Sequoia | -1 | 41 | 27467.000000 | 98304 | 1572864 | 6389760 |
| 41 | 3.72 | Lawrence Livermore National Laboratory | DOE/NNSA/LLNL Sequoia | -1 | 41 | 23751.000000 | 98304 | 1572864 | 6389760 Watts |
| 42 | 3.42 | Tokyo Institute of Technology | EBD-RH5885v2 | -1 | 33 | 3.107220 | 1 | 48 | 908.22 Watts |
| 43 | 1.89 | Research Center for Advanced Computing Infrastructure Japan Advanced Institute of Science and Technology | altix | -1 | 30 | 37.659300 | 32 | 512 | 19873 Watts |
| 44 | 0.73 | Mayo Clinic | grace | 129 | 31 | 10.319000 | 64 | 64 | 14100 Watts |
| 45 | 0.05 | Institute of Mathematics and Mechanics UrB RAS | URAN | -1 | 30 | 2.796350 | 16 | 1984 | 57600 Watts |
| 46 | 0.05 | Swiss National Supercomputing Center | Matterhorn | 200 | 31 | 0.897000 | 64 | 64 | 19557 Watts |
| 47 | 0.01 | Ecole Polytechnique Fédérale de Lausanne | Chaos (LABOS T-Rex) | 236 | 33 | 0.026400 | 8 | 128 | 1824 |
| 48 | 0.01 | Ecole Polytechnique Fédérale de Lausanne | Chaos (LABOS T-Rex) | -1 | 40 | 0.025700 | 20 | 320 | 4560 Watts |