分子動力学シミュレーション GROMACS 公式ページで紹介されているベンチマークテスト ADH_bench_systems の adh_cubic を実施した結果を掲載します。adh_cubic は、134,177原子のベンチマークです。表中の「ns/day」の値が大きいほど、パフォーマンスが高いことを示します。最新 GPU を搭載したベンチマーク結果も随時掲載します。
ベンチマークデータは、以下のURLからダウンロード可能です。
https://ftp.gromacs.org/pub/benchmarks/
ADH_bench_systems adh_cubic pme_verlet.mdp
| specification | nodes | procs | threads | GPUs | ns/day |
| [GPUあり構成/2024-08] Machine : HPC-ProServer DPeR7625 CPU:(2) AMD EPYC 9354 3.25G 32C/64T Mem : 384GB (24) 16GB RDIMM, 4800MT/s OS : RockyLinux8.10 GPU : (1) AMD Instinct MI210 ROCm : 6.2.0 Gromacs 2022.3.amd1_174 (Docker) SIMD_Option : AVX2_256 |
1 | 1 | 64 | 1 | 111.98 |
| [GPUあり構成/2024-01] Machine : HPC-ProServer DPrT7960 CPU:(2) Intel Xeon w9-3495X 1.9-4.8G, 56C/112T Mem : 256GB (8) 32GB RDIMM, 4800MT/s OS : Ubuntu 22.04 GPU : (2) NVIDIA GeForce 6000 Ada CUDA : 12.3, Driver : 545.23.08 Gromacs 2023.3 (build from source) Compiler : gcc 11.4.0 SIMD_Option : AVX_512 |
1 | 4 | 32 | 2 | 293.95 |
| 1 | 8 | 32 | 2 | 265.34 | |
| 1 | 16 | 32 | 2 | 115.14 | |
| [GPUあり構成/2023-12] Machine : HPC-ProServer DPeR760 CPU:(2) Intel Xeon Gold 6426Y 2.5G, 16C/32T Mem : 256GB (8) 32GB RDIMM, 4800MT/s OS : RockyLinux 8.8 GPU : (2) NVIDIA GeForce RTX A6000 CUDA : 12.3, Driver : 545.23.08 Gromacs 2023.3 (build from source) Compiler : gcc 11.4.0 SIMD_Option : AVX_512 |
1 | 4 | 4 | 2 | 143.80 |
| 1 | 4 | 8 | 2 | 203.52 | |
| 1 | 4 | 16 | 2 | 216.59 | |
| 1 | 8 | 8 | 2 | 170.93 | |
| 1 | 8 | 16 | 2 | 131.42 | |
| 1 | 8 | 32 | 2 | 134.67 | |
| 1 | 16 | 16 | 2 | 209.18 | |
| 1 | 16 | 32 | 2 | 211.27 | |
| [GPUあり構成/2023-09] Machine : HPC-ProServer DPrT5860 CPU:(1) Intel Xeon w5-2445 3.10G 10T/20T Mem : 128GB (2) 64GB DDR5 OS : Ubuntu 22.04.03 GPU : (1) NVIDIA GeForce RTX A6000 CUDA : 12.2, Driver : 535.104.05 Gromacs 2023.2 (build from source) Compiler : gcc 11.4.0 SIMD_Option : AVX_512 |
1 | 2 | 2 | 1 | 103.29 |
| 1 | 2 | 4 | 1 | 127.89 | |
| 1 | 2 | 8 | 1 | 140.03 | |
| 1 | 4 | 4 | 1 | 145.56 | |
| 1 | 4 | 8 | 1 | 126.07 | |
| [GPUあり構成/2023-05] Machine : HPC-ProServer DPeR7920 CPU:(2) Intel Xeon Silver 4210R 2.40G 8T/16T Mem : 32GB (2) 16GB DDR4 OS : RockyLinux 8.7 GPU : (1) NVIDIA GeForce RTX 4080 CUDA : 12.1, Driver : 530.30.02 Gromacs 2023.1 (build from source) Compiler : gcc 12.1.1 SIMD_Option : AVX2_256 |
1 | 2 | 2 | 1 | 63.64 |
| 1 | 2 | 4 | 1 | 82.88 | |
| 1 | 2 | 8 | 1 | 94.22 | |
| 1 | 4 | 4 | 1 | 112.25 | |
| 1 | 4 | 8 | 1 | 128.28 | |
| 1 | 8 | 8 | 1 | 127.05 | |
| 【GPUあり構成/2023-05】 HPC-ProServer DPeR7525 CPU : (2) AMD EPYC 7713 2.0GHz 64C/128T Mem : 1024GB (32) 32GB ECC DDR4 3200MHz OS : RockyLinux 8.7 GPU : (1) NVIDIA GeForce RTX 4090 GROMACS 2023.1 (build from source) Compiler : gcc 12.1.1 (gcc-toolset-12) NVIDIA Driver : 530.30.02 CUDA : 12.1 |
1 | 2 | 4 | 1 | 93.42 |
| 1 | 2 | 8 | 1 | 99.46 | |
| 1 | 2 | 16 | 1 | 119.90 | |
| 1 | 2 | 32 | 1 | 112.85 | |
| 1 | 2 | 64 | 1 | 72.82 | |
| 1 | 2 | 128 | 1 | 41.92 | |
| 1 | 4 | 8 | 1 | 120.61 | |
| 1 | 4 | 16 | 1 | 121.35 | |
| 1 | 4 | 32 | 1 | 141.03 | |
| 1 | 4 | 64 | 1 | 127.78 | |
| 1 | 4 | 128 | 1 | 92.76 | |
| 1 | 8 | 16 | 1 | 116.21 | |
| 1 | 8 | 32 | 1 | 163.33 | |
| 1 | 8 | 64 | 1 | 158.95 | |
| 1 | 8 | 128 | 1 | 122.39 | |
| 1 | 16 | 32 | 1 | 112.55 | |
| 1 | 16 | 64 | 1 | 122.77 | |
| 1 | 16 | 128 | 1 | 86.08 | |
| 【GPUあり構成/2023-04】 HPC-ProServer DPeR7525 CPU : (2) AMD EPYC 7713 2.0GHz 64C/128T Mem : 512GB (16) 32GB ECC DDR4 3200MHz OS : RockyLinux 8.7 GPU : (2) NVIDIA RTX A6000 GROMACS 2023 (build from source) Compiler : gcc 8.5.0 NVIDIA Driver : 525.85.12 CUDA : 12.0 |
1 | 2 | 4 | 2 | 69.34 |
| 1 | 2 | 8 | 2 | 74.14 | |
| 1 | 2 | 16 | 2 | 77.61 | |
| 1 | 2 | 32 | 2 | 76.79 | |
| 1 | 2 | 64 | 2 | 76.08 | |
| 1 | 2 | 128 | 2 | 50.37 | |
| 1 | 4 | 8 | 2 | 55.66 | |
| 1 | 4 | 16 | 2 | 63.09 | |
| 1 | 4 | 32 | 2 | 47.91 | |
| 1 | 4 | 64 | 2 | 66.61 | |
| 1 | 4 | 128 | 2 | 95.61 | |
| 1 | 8 | 16 | 2 | 113.72 | |
| 1 | 8 | 32 | 2 | 130.02 | |
| 1 | 8 | 64 | 2 | 150.13 | |
| 1 | 8 | 128 | 2 | 201.13 | |
| 1 | 16 | 32 | 2 | 93.95 | |
| 1 | 16 | 64 | 2 | 128.36 | |
| 1 | 16 | 128 | 2 | 123.04 | |
| 【GPUなし構成/2023-04】 HPC-ProServer DPeR7525 CPU : (2) AMD EPYC 7713 2.0GHz 64C/128T Mem : 512GB (16) 32GB ECC DDR4 3200MHz OS : RockyLinux 8.7 GPU : なし GROMACS 2023 (build from source) Compiler : gcc 8.5.0 |
1 | 2 | 4 | 0 | 7.91 |
| 1 | 2 | 8 | 0 | 13.91 | |
| 1 | 2 | 16 | 0 | 25.09 | |
| 1 | 2 | 32 | 0 | 34.26 | |
| 1 | 2 | 64 | 0 | 32.94 | |
| 1 | 2 | 128 | 0 | 23.92 | |
| 1 | 4 | 8 | 0 | 14.43 | |
| 1 | 4 | 16 | 0 | 27.02 | |
| 1 | 4 | 32 | 0 | 47.01 | |
| 1 | 4 | 64 | 0 | 48.08 | |
| 1 | 4 | 128 | 0 | 36.64 | |
| 1 | 8 | 16 | 0 | 27.31 | |
| 1 | 8 | 32 | 0 | 47.91 | |
| 1 | 8 | 64 | 0 | 65.64 | |
| 1 | 8 | 128 | 0 | 47.01 | |
| 1 | 16 | 32 | 0 | 48.41 | |
| 1 | 16 | 64 | 0 | 65.77 | |
| 1 | 16 | 128 | 0 | 60.85 | |
| 1 | 32 | 64 | 0 | 57.87 | |
| 1 | 32 | 128 | 0 | 54.24 | |
| 1 | 64 | 128 | 0 | 59.4 | |
| 1 | 128 | 128 | 0 | 54.69 | |
| 【GPUあり構成/2023-03】 HPC-ProServer DPeR750 CPU:(2) Intel Xeon Gold 6326 2.90GHz 16C/32T Mem : 256GB (16) 16GB DDR4 3200MHz OS : RockyLinux 8.7 GPU : (2) NVIDIA RTX A6000 CUDA12.1 Driver: 530.30.02 Gromacs 2023 (build from source) |
1 | 16 | 32 | 2 | 209.13 |
| 1 | 8 | 32 | 2 | 227.45 | |
| [GPUあり構成/2023-02] Machine : HPC-ProServer DPrT7865 CPU:(1) AMD Ryzen Threadripper PRO 5995WX 256 MB cache, 64C/128T, 2.7GHz‐4.5GHz Mem : 256GB (8) 32GB DDR4 3200MHz OS : Ubuntu 22.04 5.15.0-58-generic GPU : (2) NVIDIA RTX A4000 CUDA : 12,0, Driver : 525.85.12 Gromacs 2022.3 (NGC Docker) SIMD_Option : AVX2_256 |
1 | 8 | 64 | 2 | 121.32 |
| 1 | 16 | 64 | 2 | 179.75 | |
| [GPUあり構成/2023-01] Machine : HPC-ProServer DPeR7920 CPU:(2) Intel Xeon Gold 6240R 2.4GHz 24C/48T Mem : 192GB (12) 16GB DDR4 OS : CentOS 7.9.2009 3.10.0-1160.80.1.el7.x86_64 GPU : (2) NVIDIA RTX A6000 CUDA : 12,0, Driver : 525.60.13 Gromacs 2022.3 (NGC Docker) SIMD_Option : AVX2_256 |
1 | 2 | 48 | 2 | 96.96 |
| 1 | 4 | 48 | 2 | 91.62 | |
| 1 | 8 | 48 | 2 | 91.93 | |
| 1 | 12 | 48 | 2 | 149.18 | |
| 1 | 16 | 48 | 2 | 151.21 | |
| [GPUあり構成/2022-09] Machine : HPC-ProServer DPeR7920 CPU:(2) Intel Xeon Silver 4110 2.10G 8T/16T Mem : 256GB (16) 16GB DDR4 OS : RockyLinux 8.6 GPU : (2) NVIDIA RTX A4000 CUDA : 11.7, Driver : 515.65.01 Gromacs 2022.2 (build from source) Compiler : gcc 8.5.0 , MPI : Intel MPI 2021.6 SIMD_Option : AVX2_256 |
1 | 2 | 16 | 2 | 79.41 |
| 1 | 4 | 16 | 2 | 81.89 | |
| 1 | 8 | 16 | 2 | 125.98 | |
| [GPUあり構成/2022-02] Machine:HPC-ProServer DPeR750XA CPU:(2) Intel Xeon Gold 6326 2.90G 16C/32T Mem:512GB((16)32GB) OS:Ubuntu 20.04 GPU:(4) NVIDIA A40 CUDA11.6 Driver: 510.47.03 NVLINK3 Gromacs 2021.3 (NGC Docker ) SIMD_Option : AVX2_256 |
1 | 8 | 32 | 1 | 115.49 |
| 1 | 8 | 32 | 2 | 122.93 | |
| 1 | 8 | 32 | 4 | 227.08 | |
| [GPUあり構成/2022-01] Machine:HPC-ProServer DPeR7920 CPU:(2) Intel Xeon Gold 6240R 2.4GHz 24C/48T Mem:384GB((12)32GB) OS:RockyLinux8.5 GPU:(1) NVIDIA RTX A5000 CUDA11.6 Driver: 510.39.01 Gromacs 2021.3 (NGC Docker) SIMD_Option : AVX2_256 |
1 | 8 | 32 | 1 | 86.93 |
| [GPUなし構成/2021-09] Machine:HPC-ProServer DPeR6525 CPU : (2) AMD EPYC 7543 2.8GHz/3.7GHz 32C/64T Mem : 1024GB (16) 64GB RDIMM 3200 Mhz OS : CentOS 7.9.2009 3.10.0-1160.21.1.el7.x86_64 Gromacs 2021.2 Compiler :AOCC 3.1.0 MPI : OpenMPI 4.1.1 SIMD_Option : AVX2_256 |
1 | 64 | 64 | 0 | 66.65 |
| [GPUあり構成/2021-04] Machine:HPC-ProServer DPeR740 CPU:(2) Intel Xeon Gold 6242 2.8G 16C/32T Mem:192GB((12)16GB) OS:CentOS 7.9 GPU:(2) Tesla V100S CUDA11.2 Driver: 460.32.03 Gromacs 2020.4 Compiler : devtoolset-8 MPI : OpenMPI 4.0.5 SIMD_Option : AVX2_256 |
1 | 1 | 1 | 1 | 42.82 |
| 1 | 8 | 32 | 1 | 91.28 | |
| 1 | 8 | 32 | 2 | 139.94 | |
ADH_bench_systems_old-allbonds adh_cubic pme_verlet.mdp
| specification | nodes | procs | threads | GPUs | Core t (s) | Wall t (s) | ns/day |
[GPUあり構成/2021-04] Machine:HPC-ProServer DPeR740 CPU:(2) Intel Xeon Gold 6242 2.8G 16C/32T Mem:192GB((12)16GB) OS:CentOS 7.9 GPU:(2) Tesla V100S CUDA11.2 Driver: 460.32.03 Gromacs 2020.4 Compiler : devtoolset-8 MPI : OpenMPI 4.0.5 SIMD_Option : AVX2_256 |
1 | 1 | 1 | 1 | 48.496 | 48.496 | 35.635 |
| 1 | 8 | 32 | 1 | 492.671 | 15.398 | 112.236 | |
| 1 | 8 | 32 | 2 | 425.852 | 13.311 | 129.834 | |
[GPUあり構成/2021-02] Machine:HPC-ProServer DPeT640 CPU:(2) Intel Xeon Gold 6226R 2.9G 16C/32T Mem:384GB((12)32GB) OS:CentOS 7.9 GPU:(2) GeForce RTX 3090 CUDA11.2 Driver: 460.32.03 Gromacs 2021 Compiler : devtoolset-8 MPI : OpenMPI 4.1.0 SIMD_Option : AVX2_256 |
1 | 1 | 1 | 1 | 46.221 | 46.221 | 37.389 |
| 1 | 1 | 2 | 2 | 64.979 | 32.49 | 53.191 | |
| 1 | 1 | 4 | 2 | 92.326 | 23.082 | 74.872 | |
| 1 | 1 | 8 | 2 | 151.924 | 18.991 | 91.001 | |
| 1 | 1 | 16 | 2 | 279.133 | 17.446 | 99.058 | |
| 1 | 2 | 32 | 2 | 397.011 | 12.409 | 139.273 | |
【GPUなし構成/2020-12】 Machine:HPC-ProServer DPeR6525 CPU : (2) AMD EPYC 7402 2.8GHz/3.35GHz 24C/48T Mem : 512GB (16) 32GB RDIMM 3200 Mhz OS : CentOS 7.8.2003 Kernel 3.10.0-1127.18.2.el7.x86_64 Gromacs 2020.4 Compiler :gcc 8.3.1 MPI : OpenMPI 4.0.5 SIMD:AVX2_256 |
1 | 4 | 4 | 0 | 1024.538 | 256.135 | 6.747 |
| 1 | 8 | 8 | 0 | 1128.556 | 141.07 | 12.25 | |
| 1 | 16 | 16 | 0 | 1238.864 | 77.429 | 22.319 | |
| 1 | 32 | 32 | 0 | 1457.109 | 45.535 | 37.953 | |
| 1 | 48 | 48 | 0 | 1470.604 | 30.638 | 56.406 | |
【GPUなし構成/2019-12】 Machine:HPC-ProServer DPeR7515 CPU : (1) AMD EPYC 7702P 2.0GHz/3.4GHz 64C/128T Mem : 256GB (8) 32GB RDIMM 3200 Mhz OS : CentOS 7.7.1908 Kernel 3.10.0-1062.9.1.el7.x86_64 Gromacs 2019.4 Compiler :gcc 8.3.1 MPI : OpenMPI 4.0.2 |
1 | 1 | 1 | 0 | 902.619 | 902.62 | 1.915 |
| 1 | 2 | 2 | 0 | 979.972 | 489.986 | 3.527 | |
| 1 | 4 | 4 | 0 | 1072.638 | 268.16 | 6.445 | |
| 1 | 8 | 8 | 0 | 1251.652 | 156.457 | 11.046 | |
| 1 | 16 | 16 | 0 | 1459.192 | 91.2 | 18.949 | |
| 1 | 32 | 32 | 0 | 1543.919 | 48.248 | 35.819 | |
| 1 | 64 | 64 | 0 | 1984.912 | 31.015 | 55.721 | |
[GPUあり構成/2019-09] Machine:HPC-ProServer DPrT7920 CPU:(2) Intel Xeon Gold 6242 2.8G 16C/32T Mem:96GB((12)8GB) OS:CentOS 7.6 GPU:(2) GeForce RTX 2080Ti CUDA10.1 Driver: 418.87 Gromacs 2019.3 Compiler : devtoolset-7 MPI : OpenMPI 4.0.1 SIMD_Option : AVX2_256 |
1 | 8 | 32 | 2 | 853.979 | 26.687 | 64.757 |
[GPUあり構成/2019-05] Machine:HPC-ProServer DPeR740 CPU:(2) Intel Xeon Gold 6142 2.6G 16C/32T Mem:96GB((12)8GB) OS:CentOS 7.6 Network : InfiniBand FDR GPU:(2) GeForce RTX 2080Ti CUDA10.1 Driver: 418.56 Gromacs 2019.2 Compiler : devtoolset-7 MPI : OpenMPI 4.0.1 SIMD_Option : AVX2_256 |
1 | 8 | 32 | 2 | 853.269 | 26.665 | 64.811 |
[GPUあり構成/2019-03] Machine:HPC-ProServer DPeR740 CPU:(2) Intel Xeon Gold 6142 2.6G 16C/32T Mem:96GB((12)8GB) OS:CentOS 7.4 Network : InfiniBand FDR GPU:(2) GeForce RTX 2080Ti CUDA10.0 Driver: 410.48 Gromacs 2019 Compiler : devtoolset-7 MPI : OpenMPI 4.0.0 SIMD_Option : AVX2_256 |
1 | 8 | 32 | 2 | 875.563 | 27.362 | 63.16 |
| 2 | 16 | 64 | 4 | 1495.73 | 23.371 | 73.945 | |
| 4 | 32 | 128 | 8 | 2218.516 | 17.333 | 99.707 | |
[GPUあり構成/2018-12] Machine:HPC-ProServer DPeT640 CPU:(2) Intel Xeon Gold 6126 CPU 2.60GHz 12C/24T Mem:96GB((12)8GB) OS:CentOS 7.5 GPU:(4) GeForce GTX 1080Ti CUDA10.0 Driver: 410.48 Gromacs 2019Beta Compiler : devtoolset-7 MPI : OpenMPI 4.0.0 SIMD_Option : AVX2_256 |
1 | 4 | 24 | 4 | 715.105 | 29.797 | 57.999 |
| 1 | 24 | 24 | 4 | 718.434 | 29.935 | 57.73 | |
| 1 | 12 | 24 | 4 | 809.215 | 33.718 | 51.254 | |
| 1 | 8 | 24 | 4 | 792.226 | 33.01 | 52.353 | |
[GPUなし構成/2018-11] Machine:HPC-ProServer DPeR7425 CPU:(2) AMD EPYC 7601 2.2GHz/2.7GHz 32C/64T Mem:1024GB (32) 32GB 2133 RDIMMs OS:CentOS v7.5 GPU:なし Gromacs:2018.3 Compiler :devtoolset-4 MPI : OpenMPI v4.0.0 SIMD_Option:AVX2_128 |
1 | 4 | 8 | 0 | 2028.481 | 253.56 | 6.816 |
| 1 | 8 | 16 | 0 | 2269.613 | 141.851 | 12.183 | |
| 1 | 16 | 32 | 0 | 2636.195 | 82.381 | 20.978 | |
| 1 | 32 | 64 | 0 | 3415.126 | 53.361 | 32.386 | |
[GPUあり構成/2018-10] Machine:HPC-ProServer DPeR740 CPU:(2) Intel Xeon Gold 6142 CPU 2.60GHz 16C/32T Mem : 96GB 2666MT/s DDR4 OS:CentOS7.4 GPU:(2) GeForce RTX 2080 CUDA10.0 Driver: 410.48 Gromacs:2018.3 Compiler :devtoolset-4, MPI : OpenMPI 3.1.2 SIMD_Option:AVX2_256 |
1 | 8 | 32 | 2 | 973.471 | 30.421 | 56.809 |
[GPUあり構成/2018-09] Machine:HPC-ProServer DPeR740 CPU:(2) Intel Xeon Gold 6142 CPU 2.60GHz 16C/32T Mem : 144GB 2666MT/s DDR4 OS:CentOS7.5 GPU:(2) TITAN V CUDA10.0 Driver: 410.48 Gromacs:2018.3 Compiler :devtoolset-4, Intel C/C++ v2018.2 MPI : Intel MPI v2018.2 SIMD_Option:AVX2_256 |
1 | 8 | 32 | 2 | 883.384 | 27.606 | 62.602 |
[GPUあり構成/2018-09] Machine:HPC-ProServer DPeR740 CPU:(2) Intel Xeon Gold 6142 CPU 2.60GHz 16C/32T Mem : 144GB 2666MT/s DDR4 OS:CentOS7.5 GPU:(2) TITAN V CUDA10.0 Driver: 410.48 Gromacs:2018.3 Compiler :devtoolset-4, MPI : OpenMPI:3.1.2 SIMD_Option:AVX2_256 |
1 | 8 | 32 | 2 | 962.733 | 30.085 | 57.442 |
[GPUあり構成/2018-09] Machine:HPC-ProServer DPeR740 CPU:(2) Intel Xeon Gold 6126 CPU 2.60GHz 12C/24T Mem:96GB((12)8GB) OS:CentOS7.5 GPU:(2) TITAN V CUDA9.2 Driver: 396.44 Gromacs:2018.3 Compiler :devtoolset-4, MPI : OpenMPI:3.1.2 SIMD_Option:AVX2_256 |
1 | 6 | 24 | 2 | 778.072 | 32.42 | 53.306 |
[GPUあり構成/2018-08] Machine:HPC-ProServer DPeR730 CPU:(2) Intel Xeon E5-2690 v4 2.60GHz 14C/28T Mem:64GB((8)8GB), Network : InfiniBand FDR OS:CentOS7.5 GPU:(2) Geforce 1080Ti CUDA9.2 Driver: 396.37 Gromacs:2018.2 Compiler :devtoolset-4, MPI : OpenMPI:3.1.2 SIMD_Option:AVX2_256 |
1 | 6 | 24 | 2 | 786.996 | 32.792 | 52.702 |
| 2 | 12 | 48 | 4 | 1093.147 | 22.774 | 75.884 | |
| 4 | 24 | 96 | 8 | 1892.254 | 19.711 | 87.676 | |
| 8 | 48 | 192 | 16 | 3063.581 | 15.956 | 108.308 | |
[GPUあり構成/2018-07] Machine:HPC-ProServer DPeR740 CPU:(2) Intel Xeon Gold 6142 2.60GHz 16C/32T Mem:192GB (12)16GB 2666 RDIMMs Network : InfiniBand EDR OS:CentOS v7.4 GPU:(2) Geforce 1080Ti CUDA9.2 Driver: 396.37 Gromacs:2018.2 Compiler :devtoolset-4, MPI : OpenMPI 3.0.0 SIMD_Option:AVX_512 |
1 | 8 | 32 | 2 | 1106.985 | 34.593 | 49.957 |
| 2 | 16 | 64 | 4 | 1602.765 | 25.043 | 69.008 | |
[GPUあり構成/2018-03] Machine:HPC-ProServer DPeR740 CPU:(2) Intel Xeon Gold 6142 2.60GHz 16C/32T Mem:192GB (12)16GB 2666 RDIMMs Network : InfiniBand EDR OS:CentOS v7.4 GPU:(2) Geforce 1080Ti CUDA9.1 Driver: 387.26 Gromacs:2018 Compiler :devtoolset-4, Intel Compiler v2017.4 MPI : OpenMPI 3.0.0 SIMD_Option:AVX_512 |
1 | 4 | 8 | 2 | 478.784 | 59.848 | 28.876 |
| 1 | 8 | 16 | 2 | 682.274 | 42.642 | 40.527 | |
| 1 | 16 | 32 | 2 | 921.121 | 28.785 | 60.037 | |
[GPUあり構成/2018-03] Machine:HPC-ProServer DPeR740 CPU:(2) Intel Xeon Gold 6142 2.60GHz 16C/32T Mem:192GB (12)16GB 2666 RDIMMs Network : InfiniBand EDR OS:CentOS v7.4 GPU:(2) Geforce 1080Ti CUDA9.1 Driver: 387.26 Gromacs:2018 Compiler :devtoolset-4, MPI : OpenMPI 3.0.0 SIMD_Option:AVX_512 |
1 | 4 | 8 | 2 | 555.866 | 69.483 | 24.872 |
| 1 | 8 | 16 | 2 | 770.893 | 48.181 | 35.868 | |
| 1 | 16 | 32 | 2 | 1047.457 | 32.733 | 52.796 | |
[GPUなし構成/2018-03] Machine:HPC-ProServer DPeR740 CPU:(2) Intel Xeon Gold 6142 2.60GHz 16C/32T Mem:192GB (12)16GB 2666 RDIMMs Network : InfiniBand EDR OS:CentOS v7.4 Gromacs:2018 Compiler :devtoolset-4, Intel Compiler v2018.1 MPI : OpenMPI 3.0.0 SIMD_Option:AVX_512 |
1 | 4 | 8 | 0 | 1325.839 | 165.73 | 10.428 |
| 1 | 8 | 16 | 0 | 1511.06 | 94.441 | 18.299 | |
| 1 | 16 | 32 | 0 | 1730.362 | 54.074 | 31.96 | |
[GPUなし構成/2017-10] Machine:HPC-ProServer DPeR640 CPU:(2) Intel Xeon Gold 6142 2.60GHz 16C/32T Mem:192GB (12)16GB 2666 RDIMMs Network : InfiniBand EDR OS:Red Hat Enterprise Linux Server release 7.3 Gromacs:2016.4 Compiler :devtoolset-4, MPI : Intel MPI v2017 SIMD_Option:AVX_512 |
1 | 8 | 32 | 0 | 2033.593 | 63.55 | 27.194 |
| 2 | 16 | 64 | 0 | 2383.647 | 37.244 | 46.401 | |
| 4 | 32 | 128 | 0 | 3401.443 | 26.574 | 65.033 | |
| 8 | 64 | 256 | 0 | 4492.564 | 17.549 | 98.477 | |
[GPUあり構成/2017-10] Machine:HPC-ProServer DPeR730 CPU:(2) Intel Xeon E5-2690 v4 2.60GHz 14C/28T Mem:64GB((8)8GB), Network : InfiniBand FDR OS:CentOS7.3 GPU:(2) Geforce 1080Ti CUDA8.0 Driver: 381.22 Gromacs:2016.4 Compiler :devtoolset-4, MPI : OpenMPI:3.0.0 SIMD_Option:AVX2_256 |
1 | 6 | 24 | 2 | 810.919 | 33.788 | 51.147 |
| 2 | 12 | 48 | 4 | 1621.443 | 33.78 | 51.16 | |
| 4 | 24 | 96 | 8 | 2382.345 | 24.816 | 69.639 | |
| 8 | 48 | 192 | 16 | 3787.483 | 19.726 | 87.607 | |
| 16 | 96 | 384 | 32 | 6497.366 | 16.92 | 102.136 | |
[GPUなし構成/2017-09] Machine:HPC-ProServer DPeR640 CPU:(2) Intel Xeon Gold 6142 CPU 2.60GHz 16C/32T Mem:192GB((12)16GB) OS:Red Hat Enterprise Linux ComputeNode release 6.9 Compiler:Intel Parallel Studio XE 2018 Cluster Edition devtoolset-4 MPI:Intel MPI v2018.0.128 Gromacs:2016.4 SIMD_Option:AVX_512 |
1 | 1 | 2 | 0 | 1652.626 | 826.313 | 2.091 |
| 1 | 2 | 4 | 0 | 1729.831 | 432.458 | 3.996 | |
| 1 | 4 | 8 | 0 | 1847.976 | 230.997 | 7.481 | |
| 1 | 8 | 16 | 0 | 2122.576 | 132.661 | 13.027 | |
| 1 | 16 | 32 | 0 | 2289.058 | 71.533 | 24.159 | |
| 1 | 1 | 4 | 0 | 1709.529 | 427.382 | 4.044 | |
| 1 | 2 | 8 | 0 | 1864.974 | 233.122 | 7.413 | |
| 1 | 4 | 16 | 0 | 2077.692 | 129.856 | 13.308 | |
| 1 | 8 | 32 | 0 | 2360.731 | 73.773 | 23.426 | |
[GPUあり構成/2017-08] Machine:HPC-ProServer DPrT7910 CPU:(2) Intel Xeon E5-2680v3 2.5GHz 12C/24T Mem:64G((8)8GB) GPU:(1) Radeon Vega Frontier Edition Driver:17.20-445420(OpenCL 2.0 AMD-APP (2416.3)) Gromacs:2016.3, Compiler : gcc:5.4.0 MPI : OpenMPI:1.10.2 OS:Ubuntu 16.04.2 LTS , SIMD_Option:AVX2_256 |
1 | 1 | 1 | 1 | 269.892 | 269.892 | 6.403 |
| 1 | 1 | 24 | 1 | 1031.65 | 42.985 | 40.204 | |
[GPUあり構成/2017-07] Machine:HPC-ProServer DPrT7910 CPU:(2) Intel Xeon E5-2680v3 2.5GHz 12C/24T Mem:64G((8)8GB) OS:CentOS7.3 GPU:(2) Geforce 1080Ti CUDA8.0 Driver: 381.22 Gromacs:2016.3 MPI: OpenMPI:2.1.1 Compiler :devtoolset-4(gcc 5.3.1) SIMD_Option:AVX2_256 |
1 | 1 | 1 | 1 | 285.037 | 285.037 | 6.063 |
| 1 | 2 | 4 | 2 | 421.517 | 105.379 | 16.4 | |
| 1 | 2 | 8 | 2 | 539.208 | 67.401 | 25.64 | |
| 1 | 2 | 16 | 2 | 4762.945 | 297.684 | 5.805 | |
| 1 | 4 | 8 | 2 | 537.286 | 67.161 | 25.732 | |
| 1 | 4 | 16 | 2 | 716.805 | 44.8 | 38.575 | |
| 1 | 6 | 24 | 2 | 912.349 | 38.015 | 45.461 | |
| 1 | 8 | 16 | 2 | 796.225 | 49.764 | 34.727 | |
[GPUなし構成/2017-07] Machine:HPC-ProServer DPrT7910 CPU:(2) Intel Xeon E5-2680v3 2.5GHz 12C/24T Mem:64G((8)8GB) OS:CentOS7.3 Gromacs:2016.3 MPI : OpenMPI:2.1.1 Compiler :devtoolset-4(gcc 5.3.1) SIMD_Option:AVX2_256 |
1 | 1 | 1 | 0 | 1494.199 | 1494.199 | 1.157 |
| 1 | 2 | 2 | 0 | 1573.98 | 786.99 | 2.196 | |
| 1 | 2 | 4 | 0 | 1682.819 | 420.705 | 4.108 | |
| 1 | 2 | 8 | 0 | 1848.183 | 231.023 | 7.481 | |
| 1 | 2 | 16 | 0 | 7819.342 | 488.709 | 3.536 | |
| 1 | 4 | 8 | 0 | 1781.366 | 222.671 | 7.761 | |
| 1 | 4 | 16 | 0 | 2033.039 | 127.065 | 13.601 | |
| 1 | 6 | 24 | 0 | 2228.476 | 92.853 | 18.612 | |
| 1 | 8 | 16 | 0 | 2025.357 | 126.585 | 13.652 | |
[GPUあり構成/2017-06] Machine:HPC-ProServer DPeR730 CPU:(2) Intel Xeon E5-2698v4 2.2GHz 20C/40T Mem:1024T((16)64GB) OS:CentOS6.9 GPU:(2) Geforce 1080Ti CUDA8.0 Driver: 381.22 Gromacs:5.1.4 Compiler :devtoolset-3, MPI : OpenMPI:2.1.1 SIMD_Option:AVX2_256 |
1 | 4 | 32 | 2 | 1334.93 | 42.351 | 40.806 |
| 1 | 8 | 8 | 2 | error | error | error | |
| 1 | 8 | 16 | 2 | 872.987 | 54.641 | 31.628 | |
| 1 | 8 | 32 | 2 | 1221.567 | 38.232 | 45.202 | |
| 1 | 10 | 20 | 2 | 1066.245 | 53.39 | 32.369 | |
| 1 | 10 | 40 | 2 | 1574.564 | 39.441 | 43.817 | |
| 1 | 20 | 40 | 2 | 1506.762 | 37.857 | 45.65 | |
[GPUなし構成/2017-06] Machine:HPC-ProServer DPeR730 CPU:(2) Intel Xeon E5-2698v4 2.2GHz 20C/40T Mem:1024T((16)64GB) OS:CentOS6.9 Gromacs:5.1.4 Compiler :devtoolset-3, MPI : OpenMPI:2.1.1 SIMD_Option:AVX2_256 |
1 | 4 | 32 | 0 | 2545.684 | 79.635 | 21.701 |
| 1 | 8 | 8 | 0 | 1754.183 | 219.495 | 7.873 | |
| 1 | 8 | 16 | 0 | 2026.862 | 126.807 | 13.628 | |
| 1 | 8 | 32 | 0 | 2546.572 | 79.662 | 21.694 | |
| 1 | 10 | 20 | 0 | 2112.517 | 105.734 | 16.344 | |
| 1 | 10 | 40 | 0 | 2615.156 | 66.146 | 26.127 | |
| 1 | 20 | 40 | 0 | 2564.54 | 64.185 | 26.925 |
