{
  "title": "HPC three-system comparison: all 108 main measurements",
  "source_sha256": "b46aa6cc4a3928a13c8807cffff48f7c38927d382fabff8d99318b468a3a40d3",
  "conditions": 36,
  "repetitions_per_condition": 3,
  "main_runs": 108,
  "timing": "Application-native elapsed/loop seconds; not combined with startup-to-exit process timing. Three repeats per condition, all results retained.",
  "cases": {
    "LAMMPS": {
      "name": "Cu EAM",
      "atoms": 2048000,
      "warmup_steps": 1000,
      "timed_steps": 30000,
      "timestep_ps": 0.005,
      "neighbor": "every 1 delay 0 check yes",
      "precision": "FP64",
      "input_sha256": "f2564466d5366e9e7859650496dfc727ecbcc48a9ff669c48d10ecc7d2d31b5a",
      "potential_sha256": "6d37c8e85214154c9d899fd71a95163baf8cc518c9c256007dbe37f3eab9e3fa"
    },
    "GROMACS": {
      "name": "benchPEP-h",
      "atoms": 12495503,
      "input_sha256": "6d7d38191b649132e905c4827a3ec9a1bce89bd6d792886f8cee4f53ea0c127a",
      "total_steps": 11000,
      "counter_reset_at_step": 1000,
      "nominal_timed_steps": 10000,
      "timestep_ps": 0.002,
      "nominal_timed_interval_ps": 20,
      "repeats_per_setting": 3
    },
    "Quantum ESPRESSO": {
      "name": "PSIWAT",
      "atoms": 586,
      "input_sha256": "43d70eca9989ac9106d0e8b39f7647bd468728878064d069f095af0a71e28dec",
      "ecutwfc_Ry": 25,
      "ecutrho_Ry": 200,
      "kpoints": 4,
      "bands": 2040,
      "precision": "FP64",
      "initialization": "from_scratch each repeat"
    }
  },
  "rows": [
    {
      "application": "LAMMPS",
      "host": "a100",
      "gpus": 0,
      "application_seconds": {
        "values": [
          1262.23,
          1262.91,
          1264.31
        ],
        "median": 1262.91,
        "min": 1262.23,
        "max": 1264.31
      },
      "source": "lammps-all-hosts-verified-data.json",
      "speedup_vs_same_host_CPU": 1.0,
      "speedup_vs_same_host_1GPU": null,
      "process_seconds": {
        "values": [
          1318.3769472260028,
          1318.681346982019,
          1320.1186866760254
        ],
        "median": 1318.681346982019,
        "min": 1318.3769472260028,
        "max": 1320.1186866760254
      },
      "configuration": {
        "MPI_ranks": 64,
        "threads_per_rank": 1,
        "backend": "cpu"
      }
    },
    {
      "application": "LAMMPS",
      "host": "a100",
      "gpus": 1,
      "application_seconds": {
        "values": [
          308.747,
          310.727,
          308.893
        ],
        "median": 308.893,
        "min": 308.747,
        "max": 310.727
      },
      "source": "lammps-all-hosts-verified-data.json",
      "speedup_vs_same_host_CPU": 4.088503138627292,
      "speedup_vs_same_host_1GPU": 1.0,
      "process_seconds": {
        "values": [
          323.2008565750439,
          325.3638039859943,
          323.4102929569781
        ],
        "median": 323.4102929569781,
        "min": 323.2008565750439,
        "max": 325.3638039859943
      },
      "configuration": {
        "MPI_ranks": 1,
        "threads_per_rank": 1,
        "backend": "kokkos"
      }
    },
    {
      "application": "LAMMPS",
      "host": "a100",
      "gpus": 2,
      "application_seconds": {
        "values": [
          164.787,
          165.748,
          163.502
        ],
        "median": 164.787,
        "min": 163.502,
        "max": 165.748
      },
      "source": "lammps-all-hosts-verified-data.json",
      "speedup_vs_same_host_CPU": 7.663893389648456,
      "speedup_vs_same_host_1GPU": 1.874498595156171,
      "process_seconds": {
        "values": [
          173.45987988100387,
          174.47709746303735,
          172.12770657800138
        ],
        "median": 173.45987988100387,
        "min": 172.12770657800138,
        "max": 174.47709746303735
      },
      "configuration": {
        "MPI_ranks": 2,
        "threads_per_rank": 1,
        "backend": "kokkos"
      }
    },
    {
      "application": "LAMMPS",
      "host": "a100",
      "gpus": 4,
      "application_seconds": {
        "values": [
          259.585,
          260.007,
          258.716
        ],
        "median": 259.585,
        "min": 258.716,
        "max": 260.007
      },
      "source": "lammps-all-hosts-verified-data.json",
      "speedup_vs_same_host_CPU": 4.865111620471137,
      "speedup_vs_same_host_1GPU": 1.1899493422193115,
      "process_seconds": {
        "values": [
          276.2898796930094,
          276.75160755001707,
          275.4940739260055
        ],
        "median": 276.2898796930094,
        "min": 275.4940739260055,
        "max": 276.75160755001707
      },
      "configuration": {
        "MPI_ranks": 16,
        "threads_per_rank": 1,
        "backend": "gpu"
      }
    },
    {
      "application": "LAMMPS",
      "host": "h100",
      "gpus": 0,
      "application_seconds": {
        "values": [
          1793.72,
          1792.95,
          1792.31
        ],
        "median": 1792.95,
        "min": 1792.31,
        "max": 1793.72
      },
      "source": "lammps-all-hosts-verified-data.json",
      "speedup_vs_same_host_CPU": 1.0,
      "speedup_vs_same_host_1GPU": null,
      "process_seconds": {
        "values": [
          1855.3345483949524,
          1854.4933130409918,
          1853.9565274110064
        ],
        "median": 1854.4933130409918,
        "min": 1853.9565274110064,
        "max": 1855.3345483949524
      },
      "configuration": {
        "MPI_ranks": 32,
        "threads_per_rank": 1,
        "backend": "cpu"
      }
    },
    {
      "application": "LAMMPS",
      "host": "h100",
      "gpus": 1,
      "application_seconds": {
        "values": [
          201.587,
          202.185,
          201.93
        ],
        "median": 201.93,
        "min": 201.587,
        "max": 202.185
      },
      "source": "lammps-all-hosts-verified-data.json",
      "speedup_vs_same_host_CPU": 8.879067003417026,
      "speedup_vs_same_host_1GPU": 1.0,
      "process_seconds": {
        "values": [
          212.31244346604217,
          213.01862931397045,
          212.66741031396668
        ],
        "median": 212.66741031396668,
        "min": 212.31244346604217,
        "max": 213.01862931397045
      },
      "configuration": {
        "MPI_ranks": 1,
        "threads_per_rank": 1,
        "backend": "kokkos-serial"
      }
    },
    {
      "application": "LAMMPS",
      "host": "h100",
      "gpus": 2,
      "application_seconds": {
        "values": [
          451.63,
          451.183,
          450.979
        ],
        "median": 451.183,
        "min": 450.979,
        "max": 451.63
      },
      "source": "lammps-all-hosts-verified-data.json",
      "speedup_vs_same_host_CPU": 3.9738864274584818,
      "speedup_vs_same_host_1GPU": 0.4475567563494192,
      "process_seconds": {
        "values": [
          469.85521693201736,
          469.4342591140303,
          469.14513561100466
        ],
        "median": 469.4342591140303,
        "min": 469.14513561100466,
        "max": 469.85521693201736
      },
      "configuration": {
        "MPI_ranks": 2,
        "threads_per_rank": 1,
        "backend": "kokkos"
      }
    },
    {
      "application": "LAMMPS",
      "host": "b200",
      "gpus": 0,
      "application_seconds": {
        "values": [
          404.177,
          403.91,
          404.076
        ],
        "median": 404.076,
        "min": 403.91,
        "max": 404.177
      },
      "source": "lammps-all-hosts-verified-data.json",
      "speedup_vs_same_host_CPU": 1.0,
      "speedup_vs_same_host_1GPU": null,
      "process_seconds": {
        "values": [
          490.8233797189314,
          491.47775822202675,
          487.7256808039965
        ],
        "median": 490.8233797189314,
        "min": 487.7256808039965,
        "max": 491.47775822202675
      },
      "configuration": {
        "MPI_ranks": 128,
        "threads_per_rank": 1,
        "backend": "cpu"
      }
    },
    {
      "application": "LAMMPS",
      "host": "b200",
      "gpus": 1,
      "application_seconds": {
        "values": [
          136.176,
          136.125,
          136.153
        ],
        "median": 136.153,
        "min": 136.125,
        "max": 136.176
      },
      "source": "lammps-all-hosts-verified-data.json",
      "speedup_vs_same_host_CPU": 2.9678082745147005,
      "speedup_vs_same_host_1GPU": 1.0,
      "process_seconds": {
        "values": [
          143.98768367699813,
          143.99038091895636,
          143.9890152909793
        ],
        "median": 143.9890152909793,
        "min": 143.98768367699813,
        "max": 143.99038091895636
      },
      "configuration": {
        "MPI_ranks": 1,
        "threads_per_rank": 1,
        "backend": "kokkos"
      }
    },
    {
      "application": "LAMMPS",
      "host": "b200",
      "gpus": 2,
      "application_seconds": {
        "values": [
          74.3937,
          74.3986,
          74.4008
        ],
        "median": 74.3986,
        "min": 74.3937,
        "max": 74.4008
      },
      "source": "lammps-all-hosts-verified-data.json",
      "speedup_vs_same_host_CPU": 5.4312312328457795,
      "speedup_vs_same_host_1GPU": 1.8300478772450017,
      "process_seconds": {
        "values": [
          80.93654681800399,
          81.08868361299392,
          81.1396641950123
        ],
        "median": 81.08868361299392,
        "min": 80.93654681800399,
        "max": 81.1396641950123
      },
      "configuration": {
        "MPI_ranks": 2,
        "threads_per_rank": 1,
        "backend": "kokkos"
      }
    },
    {
      "application": "LAMMPS",
      "host": "b200",
      "gpus": 4,
      "application_seconds": {
        "values": [
          47.6978,
          48.2095,
          47.7829
        ],
        "median": 47.7829,
        "min": 47.6978,
        "max": 48.2095
      },
      "source": "lammps-all-hosts-verified-data.json",
      "speedup_vs_same_host_CPU": 8.456498035908243,
      "speedup_vs_same_host_1GPU": 2.8494084703942204,
      "process_seconds": {
        "values": [
          56.352892113965936,
          56.84810227993876,
          56.25123000703752
        ],
        "median": 56.352892113965936,
        "min": 56.25123000703752,
        "max": 56.84810227993876
      },
      "configuration": {
        "MPI_ranks": 4,
        "threads_per_rank": 1,
        "backend": "kokkos"
      }
    },
    {
      "application": "LAMMPS",
      "host": "b200",
      "gpus": 8,
      "application_seconds": {
        "values": [
          35.097,
          35.0559,
          35.1389
        ],
        "median": 35.097,
        "min": 35.0559,
        "max": 35.1389
      },
      "source": "lammps-all-hosts-verified-data.json",
      "speedup_vs_same_host_CPU": 11.513120779553809,
      "speedup_vs_same_host_1GPU": 3.879334416046955,
      "process_seconds": {
        "values": [
          51.09135403507389,
          50.676751674036495,
          51.02777029701974
        ],
        "median": 51.02777029701974,
        "min": 50.676751674036495,
        "max": 51.09135403507389
      },
      "configuration": {
        "MPI_ranks": 8,
        "threads_per_rank": 1,
        "backend": "kokkos"
      }
    },
    {
      "application": "GROMACS",
      "host": "a100",
      "gpus": 0,
      "application_seconds": {
        "values": [
          4317.449,
          4340.323,
          4320.487
        ],
        "median": 4320.487,
        "min": 4317.449,
        "max": 4340.323
      },
      "source": "gromacs-all-hosts-verified-data.json",
      "speedup_vs_same_host_CPU": 1.0,
      "speedup_vs_same_host_1GPU": null,
      "process_seconds": {
        "values": [
          4770.1226468089735,
          4793.670560342958,
          4771.8124862620025
        ],
        "median": 4771.8124862620025,
        "min": 4770.1226468089735,
        "max": 4793.670560342958
      },
      "configuration": {
        "precision": "double",
        "MPI_ranks": 16,
        "threads_per_rank": 4,
        "PME_ranks": -1,
        "MD_update": "cpu",
        "host_staged_GPU_communication": false
      }
    },
    {
      "application": "GROMACS",
      "host": "a100",
      "gpus": 1,
      "application_seconds": {
        "values": [
          736.01,
          736.236,
          735.603
        ],
        "median": 736.01,
        "min": 735.603,
        "max": 736.236
      },
      "source": "gromacs-all-hosts-verified-data.json",
      "speedup_vs_same_host_CPU": 5.870147144739882,
      "speedup_vs_same_host_1GPU": 1.0,
      "process_seconds": {
        "values": [
          820.623264945054,
          821.4191571060219,
          820.2592642970267
        ],
        "median": 820.623264945054,
        "min": 820.2592642970267,
        "max": 821.4191571060219
      },
      "configuration": {
        "precision": "mixed",
        "MPI_ranks": 1,
        "threads_per_rank": 8,
        "PME_ranks": 0,
        "MD_update": "gpu",
        "host_staged_GPU_communication": false
      }
    },
    {
      "application": "GROMACS",
      "host": "a100",
      "gpus": 2,
      "application_seconds": {
        "values": [
          688.636,
          691.871,
          693.109
        ],
        "median": 691.871,
        "min": 688.636,
        "max": 693.109
      },
      "source": "gromacs-all-hosts-verified-data.json",
      "speedup_vs_same_host_CPU": 6.244642426117007,
      "speedup_vs_same_host_1GPU": 1.0637965747950124,
      "process_seconds": {
        "values": [
          773.2833151110099,
          776.6603219680255,
          777.6100022689789
        ],
        "median": 776.6603219680255,
        "min": 773.2833151110099,
        "max": 777.6100022689789
      },
      "configuration": {
        "precision": "mixed",
        "MPI_ranks": 2,
        "threads_per_rank": 8,
        "PME_ranks": 1,
        "MD_update": "gpu",
        "host_staged_GPU_communication": false
      }
    },
    {
      "application": "GROMACS",
      "host": "a100",
      "gpus": 4,
      "application_seconds": {
        "values": [
          817.647,
          819.408,
          820.736
        ],
        "median": 819.408,
        "min": 817.647,
        "max": 820.736
      },
      "source": "gromacs-all-hosts-verified-data.json",
      "speedup_vs_same_host_CPU": 5.2726932126608475,
      "speedup_vs_same_host_1GPU": 0.8982216429422216,
      "process_seconds": {
        "values": [
          914.3664909409708,
          916.3400035129744,
          917.7430766180041
        ],
        "median": 916.3400035129744,
        "min": 914.3664909409708,
        "max": 917.7430766180041
      },
      "configuration": {
        "precision": "mixed",
        "MPI_ranks": 4,
        "threads_per_rank": 8,
        "PME_ranks": 1,
        "MD_update": "cpu",
        "host_staged_GPU_communication": true
      }
    },
    {
      "application": "GROMACS",
      "host": "h100",
      "gpus": 0,
      "application_seconds": {
        "values": [
          8085.651,
          8080.786,
          8083.422
        ],
        "median": 8083.422,
        "min": 8080.786,
        "max": 8085.651
      },
      "source": "gromacs-all-hosts-verified-data.json",
      "speedup_vs_same_host_CPU": 1.0,
      "speedup_vs_same_host_1GPU": null,
      "process_seconds": {
        "values": [
          8912.129527340003,
          8906.496662073012,
          8909.319356076012
        ],
        "median": 8909.319356076012,
        "min": 8906.496662073012,
        "max": 8912.129527340003
      },
      "configuration": {
        "precision": "double",
        "MPI_ranks": 16,
        "threads_per_rank": 2,
        "PME_ranks": -1,
        "MD_update": "cpu",
        "host_staged_GPU_communication": false
      }
    },
    {
      "application": "GROMACS",
      "host": "h100",
      "gpus": 1,
      "application_seconds": {
        "values": [
          523.39,
          522.866,
          522.719
        ],
        "median": 522.866,
        "min": 522.719,
        "max": 523.39
      },
      "source": "gromacs-all-hosts-verified-data.json",
      "speedup_vs_same_host_CPU": 15.459834833398997,
      "speedup_vs_same_host_1GPU": 1.0,
      "process_seconds": {
        "values": [
          586.9363163569942,
          586.3934680910315,
          586.3126844119979
        ],
        "median": 586.3934680910315,
        "min": 586.3126844119979,
        "max": 586.9363163569942
      },
      "configuration": {
        "precision": "mixed",
        "MPI_ranks": 1,
        "threads_per_rank": 8,
        "PME_ranks": 0,
        "MD_update": "gpu",
        "host_staged_GPU_communication": false
      }
    },
    {
      "application": "GROMACS",
      "host": "h100",
      "gpus": 2,
      "application_seconds": {
        "values": [
          1141.725,
          1143.447,
          1145.659
        ],
        "median": 1143.447,
        "min": 1141.725,
        "max": 1145.659
      },
      "source": "gromacs-all-hosts-verified-data.json",
      "speedup_vs_same_host_CPU": 7.069345584010453,
      "speedup_vs_same_host_1GPU": 0.45727174062287107,
      "process_seconds": {
        "values": [
          1270.9648178979987,
          1272.821257388976,
          1275.1793042360223
        ],
        "median": 1272.821257388976,
        "min": 1270.9648178979987,
        "max": 1275.1793042360223
      },
      "configuration": {
        "precision": "mixed",
        "MPI_ranks": 2,
        "threads_per_rank": 8,
        "PME_ranks": 1,
        "MD_update": "cpu",
        "host_staged_GPU_communication": true
      }
    },
    {
      "application": "GROMACS",
      "host": "b200",
      "gpus": 0,
      "application_seconds": {
        "values": [
          1339.209,
          1338.165,
          1339.56
        ],
        "median": 1339.209,
        "min": 1338.165,
        "max": 1339.56
      },
      "source": "gromacs-all-hosts-verified-data.json",
      "speedup_vs_same_host_CPU": 1.0,
      "speedup_vs_same_host_1GPU": null,
      "process_seconds": {
        "values": [
          1552.540378854028,
          1554.6627327030292,
          1554.356649287045
        ],
        "median": 1554.356649287045,
        "min": 1552.540378854028,
        "max": 1554.6627327030292
      },
      "configuration": {
        "precision": "double",
        "MPI_ranks": 128,
        "threads_per_rank": 1,
        "PME_ranks": -1,
        "MD_update": "cpu",
        "host_staged_GPU_communication": false
      }
    },
    {
      "application": "GROMACS",
      "host": "b200",
      "gpus": 1,
      "application_seconds": {
        "values": [
          333.776,
          333.767,
          333.831
        ],
        "median": 333.776,
        "min": 333.767,
        "max": 333.831
      },
      "source": "gromacs-all-hosts-verified-data.json",
      "speedup_vs_same_host_CPU": 4.012298667369733,
      "speedup_vs_same_host_1GPU": 1.0,
      "process_seconds": {
        "values": [
          375.25044887000695,
          375.29816360899713,
          375.3903588090325
        ],
        "median": 375.29816360899713,
        "min": 375.25044887000695,
        "max": 375.3903588090325
      },
      "configuration": {
        "precision": "mixed",
        "MPI_ranks": 1,
        "threads_per_rank": 8,
        "PME_ranks": 0,
        "MD_update": "gpu",
        "host_staged_GPU_communication": false
      }
    },
    {
      "application": "GROMACS",
      "host": "b200",
      "gpus": 2,
      "application_seconds": {
        "values": [
          323.072,
          323.194,
          323.157
        ],
        "median": 323.157,
        "min": 323.072,
        "max": 323.194
      },
      "source": "gromacs-all-hosts-verified-data.json",
      "speedup_vs_same_host_CPU": 4.144143558703664,
      "speedup_vs_same_host_1GPU": 1.0328601887008484,
      "process_seconds": {
        "values": [
          367.5341690700734,
          367.54226585896686,
          367.5357702510664
        ],
        "median": 367.5357702510664,
        "min": 367.5341690700734,
        "max": 367.54226585896686
      },
      "configuration": {
        "precision": "mixed",
        "MPI_ranks": 2,
        "threads_per_rank": 8,
        "PME_ranks": 1,
        "MD_update": "gpu",
        "host_staged_GPU_communication": false
      }
    },
    {
      "application": "GROMACS",
      "host": "b200",
      "gpus": 4,
      "application_seconds": {
        "values": [
          129.696,
          129.737,
          129.612
        ],
        "median": 129.696,
        "min": 129.612,
        "max": 129.737
      },
      "source": "gromacs-all-hosts-verified-data.json",
      "speedup_vs_same_host_CPU": 10.325754071058476,
      "speedup_vs_same_host_1GPU": 2.573525783370343,
      "process_seconds": {
        "values": [
          157.74705794709735,
          157.9499094269704,
          157.70245147601236
        ],
        "median": 157.74705794709735,
        "min": 157.70245147601236,
        "max": 157.9499094269704
      },
      "configuration": {
        "precision": "mixed",
        "MPI_ranks": 4,
        "threads_per_rank": 4,
        "PME_ranks": 1,
        "MD_update": "gpu",
        "host_staged_GPU_communication": false
      }
    },
    {
      "application": "GROMACS",
      "host": "b200",
      "gpus": 8,
      "application_seconds": {
        "values": [
          59.478,
          59.4,
          59.552
        ],
        "median": 59.478,
        "min": 59.4,
        "max": 59.552
      },
      "source": "gromacs-all-hosts-verified-data.json",
      "speedup_vs_same_host_CPU": 22.516039544033088,
      "speedup_vs_same_host_1GPU": 5.611755607115236,
      "process_seconds": {
        "values": [
          88.77683183597401,
          88.57246998196933,
          88.97569250699598
        ],
        "median": 88.77683183597401,
        "min": 88.57246998196933,
        "max": 88.97569250699598
      },
      "configuration": {
        "precision": "mixed",
        "MPI_ranks": 8,
        "threads_per_rank": 4,
        "PME_ranks": 1,
        "MD_update": "gpu",
        "host_staged_GPU_communication": false
      }
    },
    {
      "application": "Quantum ESPRESSO",
      "host": "a100",
      "gpus": 0,
      "application_seconds": {
        "values": [
          3642.97,
          3634.07,
          3636.75
        ],
        "median": 3636.75,
        "min": 3634.07,
        "max": 3642.97
      },
      "source": "qe-a100-verified-data.json",
      "speedup_vs_same_host_CPU": 1.0,
      "speedup_vs_same_host_1GPU": null,
      "process_seconds": {
        "values": [
          3656.477973907022,
          3647.5107133849524,
          3650.203663389024
        ],
        "median": 3650.203663389024,
        "min": 3647.5107133849524,
        "max": 3656.477973907022
      },
      "configuration": {
        "MPI_ranks": 64,
        "threads_per_rank": 1,
        "kpoint_pools": 4,
        "job_local_UCC_disabled": false
      }
    },
    {
      "application": "Quantum ESPRESSO",
      "host": "a100",
      "gpus": 1,
      "application_seconds": {
        "values": [
          1283.26,
          1286.94,
          1288.88
        ],
        "median": 1286.94,
        "min": 1283.26,
        "max": 1288.88
      },
      "source": "qe-a100-verified-data.json",
      "speedup_vs_same_host_CPU": 2.8258893188493635,
      "speedup_vs_same_host_1GPU": 1.0,
      "process_seconds": {
        "values": [
          1285.3453688619775,
          1289.091894584999,
          1290.958282472042
        ],
        "median": 1289.091894584999,
        "min": 1285.3453688619775,
        "max": 1290.958282472042
      },
      "configuration": {
        "MPI_ranks": 1,
        "threads_per_rank": 8,
        "kpoint_pools": 1,
        "job_local_UCC_disabled": false
      }
    },
    {
      "application": "Quantum ESPRESSO",
      "host": "a100",
      "gpus": 2,
      "application_seconds": {
        "values": [
          680.52,
          679.94,
          680.09
        ],
        "median": 680.09,
        "min": 679.94,
        "max": 680.52
      },
      "source": "qe-a100-verified-data.json",
      "speedup_vs_same_host_CPU": 5.347454013439397,
      "speedup_vs_same_host_1GPU": 1.8923083709509037,
      "process_seconds": {
        "values": [
          683.0085796510102,
          682.3934280449757,
          682.5677977579762
        ],
        "median": 682.5677977579762,
        "min": 682.3934280449757,
        "max": 683.0085796510102
      },
      "configuration": {
        "MPI_ranks": 2,
        "threads_per_rank": 1,
        "kpoint_pools": 2,
        "job_local_UCC_disabled": false
      }
    },
    {
      "application": "Quantum ESPRESSO",
      "host": "a100",
      "gpus": 4,
      "application_seconds": {
        "values": [
          299.69,
          299.43,
          299.43
        ],
        "median": 299.43,
        "min": 299.43,
        "max": 299.69
      },
      "source": "qe-a100-verified-data.json",
      "speedup_vs_same_host_CPU": 12.14557659553151,
      "speedup_vs_same_host_1GPU": 4.29796613565775,
      "process_seconds": {
        "values": [
          302.85891161998734,
          302.65170053899055,
          302.55273421801394
        ],
        "median": 302.65170053899055,
        "min": 302.55273421801394,
        "max": 302.85891161998734
      },
      "configuration": {
        "MPI_ranks": 4,
        "threads_per_rank": 1,
        "kpoint_pools": 4,
        "job_local_UCC_disabled": true
      }
    },
    {
      "application": "Quantum ESPRESSO",
      "host": "h100",
      "gpus": 0,
      "application_seconds": {
        "values": [
          10538.18,
          10586.08,
          10582.52
        ],
        "median": 10582.52,
        "min": 10538.18,
        "max": 10586.08
      },
      "source": "qe-h100-cpu-verified-data.json",
      "speedup_vs_same_host_CPU": 1.0,
      "speedup_vs_same_host_1GPU": null,
      "process_seconds": {
        "values": [
          10542.209861334995,
          10590.141620139999,
          10586.604125366022
        ],
        "median": 10586.604125366022,
        "min": 10542.209861334995,
        "max": 10590.141620139999
      },
      "configuration": {
        "ranks": 32,
        "threads_per_rank": 1,
        "kpoint_pools": 2,
        "ndiag": 4
      }
    },
    {
      "application": "Quantum ESPRESSO",
      "host": "h100",
      "gpus": 1,
      "application_seconds": {
        "values": [
          774.69,
          772.6,
          775.04
        ],
        "median": 774.69,
        "min": 772.6,
        "max": 775.04
      },
      "source": "qe-h100-gpu1-verified-data.json",
      "speedup_vs_same_host_CPU": 13.660328647588067,
      "speedup_vs_same_host_1GPU": 1.0,
      "process_seconds": {
        "values": [
          776.9017812979873,
          774.7222192149493,
          777.1740283140098
        ],
        "median": 776.9017812979873,
        "min": 774.7222192149493,
        "max": 777.1740283140098
      },
      "configuration": "g1-selected"
    },
    {
      "application": "Quantum ESPRESSO",
      "host": "h100",
      "gpus": 2,
      "application_seconds": {
        "values": [
          920.23,
          898.16,
          920.63
        ],
        "median": 920.23,
        "min": 898.16,
        "max": 920.63
      },
      "source": "qe-h100-gpu2-verified-data.json",
      "speedup_vs_same_host_CPU": 11.499864164393685,
      "speedup_vs_same_host_1GPU": 0.8418438868543734,
      "process_seconds": {
        "values": [
          922.8903200349887,
          900.834444764012,
          923.2825870969682
        ],
        "median": 922.8903200349887,
        "min": 900.834444764012,
        "max": 923.2825870969682
      },
      "configuration": {
        "ranks": 2,
        "threads_per_rank": 1,
        "kpoint_pools": 1,
        "GPU_aware_MPI": false,
        "UCC_enabled": false,
        "binary_sha256": "74dab1b18c007a5c62c64e3f0cb9cc8a726fd00293c95b441ea946eb08420894"
      }
    },
    {
      "application": "Quantum ESPRESSO",
      "host": "b200",
      "gpus": 0,
      "application_seconds": {
        "values": [
          1649.48,
          1641.63,
          1637.86
        ],
        "median": 1641.63,
        "min": 1637.86,
        "max": 1649.48
      },
      "source": "qe-b200-verified-data.json",
      "speedup_vs_same_host_CPU": 1.0,
      "speedup_vs_same_host_1GPU": null,
      "process_seconds": {
        "values": [
          1720.909389129025,
          1715.7627940169768,
          1711.619554085075
        ],
        "median": 1715.7627940169768,
        "min": 1711.619554085075,
        "max": 1720.909389129025
      },
      "configuration": {
        "mpi_ranks": 128,
        "omp_threads_per_rank": 1,
        "kpoint_pools": 4
      }
    },
    {
      "application": "Quantum ESPRESSO",
      "host": "b200",
      "gpus": 1,
      "application_seconds": {
        "values": [
          532.29,
          530.75,
          532.82
        ],
        "median": 532.29,
        "min": 530.75,
        "max": 532.82
      },
      "source": "qe-b200-verified-data.json",
      "speedup_vs_same_host_CPU": 3.084089500084541,
      "speedup_vs_same_host_1GPU": 1.0,
      "process_seconds": {
        "values": [
          534.4969798750244,
          532.9894314350095,
          535.0516572010238
        ],
        "median": 534.4969798750244,
        "min": 532.9894314350095,
        "max": 535.0516572010238
      },
      "configuration": {
        "mpi_ranks": 1,
        "omp_threads_per_rank": 8,
        "kpoint_pools": 1
      }
    },
    {
      "application": "Quantum ESPRESSO",
      "host": "b200",
      "gpus": 2,
      "application_seconds": {
        "values": [
          285.99,
          284.28,
          285.32
        ],
        "median": 285.32,
        "min": 284.28,
        "max": 285.99
      },
      "source": "qe-b200-verified-data.json",
      "speedup_vs_same_host_CPU": 5.753645030141596,
      "speedup_vs_same_host_1GPU": 1.8655895135286695,
      "process_seconds": {
        "values": [
          289.7554432459874,
          287.9480361859314,
          289.0477083469741
        ],
        "median": 289.0477083469741,
        "min": 287.9480361859314,
        "max": 289.7554432459874
      },
      "configuration": {
        "mpi_ranks": 2,
        "omp_threads_per_rank": 1,
        "kpoint_pools": 2
      }
    },
    {
      "application": "Quantum ESPRESSO",
      "host": "b200",
      "gpus": 4,
      "application_seconds": {
        "values": [
          135.89,
          138.91,
          136.09
        ],
        "median": 136.09,
        "min": 135.89,
        "max": 138.91
      },
      "source": "qe-b200-verified-data.json",
      "speedup_vs_same_host_CPU": 12.062826070982439,
      "speedup_vs_same_host_1GPU": 3.9113086927768386,
      "process_seconds": {
        "values": [
          142.51990720594767,
          145.67742828500923,
          142.92364127293695
        ],
        "median": 142.92364127293695,
        "min": 142.51990720594767,
        "max": 145.67742828500923
      },
      "configuration": {
        "mpi_ranks": 4,
        "omp_threads_per_rank": 1,
        "kpoint_pools": 4
      }
    },
    {
      "application": "Quantum ESPRESSO",
      "host": "b200",
      "gpus": 8,
      "application_seconds": {
        "values": [
          85.8,
          85.08,
          85.46
        ],
        "median": 85.46,
        "min": 85.08,
        "max": 85.8
      },
      "source": "qe-b200-verified-data.json",
      "speedup_vs_same_host_CPU": 19.20933770184882,
      "speedup_vs_same_host_1GPU": 6.228527966300024,
      "process_seconds": {
        "values": [
          99.91519034898374,
          99.28626293491106,
          99.8980426969938
        ],
        "median": 99.8980426969938,
        "min": 99.28626293491106,
        "max": 99.91519034898374
      },
      "configuration": {
        "mpi_ranks": 8,
        "omp_threads_per_rank": 1,
        "kpoint_pools": 4
      }
    }
  ],
  "limitations": [
    "Host CPUs, MPI layouts, interconnects and builds differ; no GPU-chip-only causality.",
    "GROMACS CPU double vs GPU mixed precision; matched numerical admission gates, not identical precision.",
    "H1002GPU QE uses host-staged MPI and isolated build;1GPU original runtime is different.",
    "Initial failures, nonconvergence, warning/monitor gaps remain in evidence. No universal optimal configuration claim."
  ],
  "includes_private_raw_logs": false
}
