{
  "timestamp_utc": "20260921T100514Z",
  "warmup": 50,
  "repeat": 200,
  "rounds": 3,
  "dtype": "bf16",
  "batch": 4,
  "heads": 32,
  "dimension": 128,
  "sequences": [
    4096,
    8192,
    16384
  ],
  "source_sha256": {
    "kernels/v1_manual_baseline.co": "8ec3f7248ddbdd4f0be9c167990bffc745d221cc4c4ed86cb2d0f5c9064ca2de",
    "kernels/v2_manual_s2_1p2c_tma.co": "64b58ed63fa626b4df2f182a6096176d622e75c1f85964061b4b422aab0d69e9",
    "kernels/v3_fa3_overlap.co": "94e83222f8b8d6d1909efa51b6ab5de41e715a2163cfaa556ae5f55acedad060",
    "kernels/v4_nonpersistent_fa.co": "3541465fcdb0810be162750827841cd3611dc357ed47d54c0ddefad53b83ea4e",
    "generated/v1_manual_baseline.cu": "b2d56da1569d40e183de18d16a67580592e45551cdfab8847a9dd907af06f956",
    "generated/v2_manual_s2_1p2c_tma.cu": "141299f333d9fbbc4ac762b82895817408d14f986c99583bb9f7258710191a70",
    "generated/v3_fa3_overlap.cu": "ac9ec0828e6425e28066c47128d745105bf8294d2bedf554eef3c033315de495",
    "generated/v4_nonpersistent_fa.cu": "067f93dd350331fa9e8da69d9a48a4445a7517728cfe08a2d5cd82cea537ab38",
    "runtime/choreo.h": "90c4b1b2b45ba93aa47c1cad029c207cb092ab6e92451b2233f3de19de9c2484",
    "runtime/choreo_cute.h": "11e61794af34ab965e3f4afc72394ef7a9ade4f93b816c80b556a0b6577addad",
    "runtime/choreo_types.h": "2ae74d3a5dbab379e3096535912d30c5369c90c48678acb2de05519f8512e0bb",
    "runtime/choreo_types_cute.h": "d4a702971ff5ec5b9eedbe9940c9f806e826317991db4d4e67036e358d1a75f0",
    "baselines/tilelang_kernel.py": "9cba9ed14a05ac583b7c57f2e94c55279a3bf7f200df76cf41faec9c454e9b35",
    "baselines/triton_kernel.py": "515fd654338961ad4ed29f6e63018aa7e3b6f843fdd46593232d1657586c1067",
    "baselines/triton_ws_kernel.py": "ccb64d5becc654bc04871134683248682806eb0b364851b7357844f6dbd7ccea",
    "run.py": "e93317c924a5e9d5aa249984acf65ac776e75d56ce3a7ca782527f67fd483bc3",
    "baseline.py": "e002fa32e725dd4819821db25f2a122f708c7dd0df2e0b21ad3c424bd0bdc96b",
    "fa_helper.hpp": "1b762e91f32653ad42ff086fc7e0845aea84765450958fa7c84167fb11f03484",
    "bench_configs.inc": "dad6065b2e24c5ac0f430c8ced5187d96b684f20d018e59a2124e121d62cb370"
  },
  "gpu_before": "0, NVIDIA H800 PCIe, GPU-c6114605-2f2c-7a74-a505-c53021985b24, 590.48.01, 9.0, 38, 345, 1593, 50.76, 0, 14",
  "measurements": [
    {
      "backend": "fa3",
      "suite_position": 0,
      "sequence": 4096,
      "round": 0,
      "ms": 1.3179609680175781
    },
    {
      "backend": "fa3",
      "suite_position": 0,
      "sequence": 4096,
      "round": 1,
      "ms": 1.3908067321777344
    },
    {
      "backend": "fa3",
      "suite_position": 0,
      "sequence": 4096,
      "round": 2,
      "ms": 1.4447817993164063
    },
    {
      "backend": "fa3",
      "suite_position": 0,
      "sequence": 8192,
      "round": 0,
      "ms": 5.254291381835937
    },
    {
      "backend": "fa3",
      "suite_position": 0,
      "sequence": 8192,
      "round": 1,
      "ms": 5.2478369140625
    },
    {
      "backend": "fa3",
      "suite_position": 0,
      "sequence": 8192,
      "round": 2,
      "ms": 5.2577392578125
    },
    {
      "backend": "fa3",
      "suite_position": 0,
      "sequence": 16384,
      "round": 0,
      "ms": 20.44233642578125
    },
    {
      "backend": "fa3",
      "suite_position": 0,
      "sequence": 16384,
      "round": 1,
      "ms": 20.50056640625
    },
    {
      "backend": "fa3",
      "suite_position": 0,
      "sequence": 16384,
      "round": 2,
      "ms": 20.5923974609375
    },
    {
      "backend": "v1",
      "sequence": 4096,
      "round": 0,
      "suite_position": 1,
      "ms": 2.58128936768
    },
    {
      "backend": "v1",
      "sequence": 8192,
      "round": 0,
      "suite_position": 1,
      "ms": 9.25581298828
    },
    {
      "backend": "v1",
      "sequence": 16384,
      "round": 0,
      "suite_position": 1,
      "ms": 35.3456933594
    },
    {
      "backend": "v1",
      "sequence": 4096,
      "round": 1,
      "suite_position": 1,
      "ms": 2.58498596191
    },
    {
      "backend": "v1",
      "sequence": 8192,
      "round": 1,
      "suite_position": 1,
      "ms": 9.28579589844
    },
    {
      "backend": "v1",
      "sequence": 16384,
      "round": 1,
      "suite_position": 1,
      "ms": 35.3124731445
    },
    {
      "backend": "v1",
      "sequence": 4096,
      "round": 2,
      "suite_position": 1,
      "ms": 2.58051757812
    },
    {
      "backend": "v1",
      "sequence": 8192,
      "round": 2,
      "suite_position": 1,
      "ms": 9.30924865723
    },
    {
      "backend": "v1",
      "sequence": 16384,
      "round": 2,
      "suite_position": 1,
      "ms": 35.3770800781
    },
    {
      "backend": "v2",
      "sequence": 4096,
      "round": 0,
      "suite_position": 2,
      "ms": 1.66002334595
    },
    {
      "backend": "v2",
      "sequence": 8192,
      "round": 0,
      "suite_position": 2,
      "ms": 6.10557373047
    },
    {
      "backend": "v2",
      "sequence": 16384,
      "round": 0,
      "suite_position": 2,
      "ms": 23.7090405273
    },
    {
      "backend": "v2",
      "sequence": 4096,
      "round": 1,
      "suite_position": 2,
      "ms": 1.66307617188
    },
    {
      "backend": "v2",
      "sequence": 8192,
      "round": 1,
      "suite_position": 2,
      "ms": 6.13873779297
    },
    {
      "backend": "v2",
      "sequence": 16384,
      "round": 1,
      "suite_position": 2,
      "ms": 23.7252880859
    },
    {
      "backend": "v2",
      "sequence": 4096,
      "round": 2,
      "suite_position": 2,
      "ms": 1.6333770752
    },
    {
      "backend": "v2",
      "sequence": 8192,
      "round": 2,
      "suite_position": 2,
      "ms": 6.15002502441
    },
    {
      "backend": "v2",
      "sequence": 16384,
      "round": 2,
      "suite_position": 2,
      "ms": 23.5249243164
    },
    {
      "backend": "v4",
      "sequence": 4096,
      "round": 0,
      "suite_position": 3,
      "ms": 1.48040100098
    },
    {
      "backend": "v4",
      "sequence": 8192,
      "round": 0,
      "suite_position": 3,
      "ms": 5.47535095215
    },
    {
      "backend": "v4",
      "sequence": 16384,
      "round": 0,
      "suite_position": 3,
      "ms": 21.2085913086
    },
    {
      "backend": "v4",
      "sequence": 4096,
      "round": 1,
      "suite_position": 3,
      "ms": 1.46422546387
    },
    {
      "backend": "v4",
      "sequence": 8192,
      "round": 1,
      "suite_position": 3,
      "ms": 5.45627929687
    },
    {
      "backend": "v4",
      "sequence": 16384,
      "round": 1,
      "suite_position": 3,
      "ms": 21.1821435547
    },
    {
      "backend": "v4",
      "sequence": 4096,
      "round": 2,
      "suite_position": 3,
      "ms": 1.46265335083
    },
    {
      "backend": "v4",
      "sequence": 8192,
      "round": 2,
      "suite_position": 3,
      "ms": 5.4329876709
    },
    {
      "backend": "v4",
      "sequence": 16384,
      "round": 2,
      "suite_position": 3,
      "ms": 20.9237280273
    },
    {
      "backend": "v3",
      "sequence": 4096,
      "round": 0,
      "suite_position": 4,
      "ms": 1.39186309814
    },
    {
      "backend": "v3",
      "sequence": 8192,
      "round": 0,
      "suite_position": 4,
      "ms": 5.29708068848
    },
    {
      "backend": "v3",
      "sequence": 16384,
      "round": 0,
      "suite_position": 4,
      "ms": 20.7590039063
    },
    {
      "backend": "v3",
      "sequence": 4096,
      "round": 1,
      "suite_position": 4,
      "ms": 1.40790298462
    },
    {
      "backend": "v3",
      "sequence": 8192,
      "round": 1,
      "suite_position": 4,
      "ms": 5.31597290039
    },
    {
      "backend": "v3",
      "sequence": 16384,
      "round": 1,
      "suite_position": 4,
      "ms": 20.7938671875
    },
    {
      "backend": "v3",
      "sequence": 4096,
      "round": 2,
      "suite_position": 4,
      "ms": 1.39671051025
    },
    {
      "backend": "v3",
      "sequence": 8192,
      "round": 2,
      "suite_position": 4,
      "ms": 5.28245300293
    },
    {
      "backend": "v3",
      "sequence": 16384,
      "round": 2,
      "suite_position": 4,
      "ms": 20.8100439453
    },
    {
      "backend": "tilelang",
      "suite_position": 5,
      "sequence": 4096,
      "round": 0,
      "ms": 1.6484440612792968
    },
    {
      "backend": "tilelang",
      "suite_position": 5,
      "sequence": 4096,
      "round": 1,
      "ms": 1.6510588073730468
    },
    {
      "backend": "tilelang",
      "suite_position": 5,
      "sequence": 4096,
      "round": 2,
      "ms": 1.7252786254882813
    },
    {
      "backend": "tilelang",
      "suite_position": 5,
      "sequence": 8192,
      "round": 0,
      "ms": 6.211298828125
    },
    {
      "backend": "tilelang",
      "suite_position": 5,
      "sequence": 8192,
      "round": 1,
      "ms": 6.3070001220703125
    },
    {
      "backend": "tilelang",
      "suite_position": 5,
      "sequence": 8192,
      "round": 2,
      "ms": 6.2545556640625
    },
    {
      "backend": "tilelang",
      "suite_position": 5,
      "sequence": 16384,
      "round": 0,
      "ms": 23.9559765625
    },
    {
      "backend": "tilelang",
      "suite_position": 5,
      "sequence": 16384,
      "round": 1,
      "ms": 24.203779296875
    },
    {
      "backend": "tilelang",
      "suite_position": 5,
      "sequence": 16384,
      "round": 2,
      "ms": 24.26598388671875
    },
    {
      "backend": "triton",
      "suite_position": 6,
      "sequence": 4096,
      "round": 0,
      "ms": 1.6855709838867188
    },
    {
      "backend": "triton",
      "suite_position": 6,
      "sequence": 4096,
      "round": 1,
      "ms": 1.754219970703125
    },
    {
      "backend": "triton",
      "suite_position": 6,
      "sequence": 4096,
      "round": 2,
      "ms": 1.8057185363769532
    },
    {
      "backend": "triton",
      "suite_position": 6,
      "sequence": 8192,
      "round": 0,
      "ms": 6.633601684570312
    },
    {
      "backend": "triton",
      "suite_position": 6,
      "sequence": 8192,
      "round": 1,
      "ms": 6.684990234375
    },
    {
      "backend": "triton",
      "suite_position": 6,
      "sequence": 8192,
      "round": 2,
      "ms": 6.674417724609375
    },
    {
      "backend": "triton",
      "suite_position": 6,
      "sequence": 16384,
      "round": 0,
      "ms": 25.75680419921875
    },
    {
      "backend": "triton",
      "suite_position": 6,
      "sequence": 16384,
      "round": 1,
      "ms": 25.9038623046875
    },
    {
      "backend": "triton",
      "suite_position": 6,
      "sequence": 16384,
      "round": 2,
      "ms": 25.965078125
    },
    {
      "backend": "triton_ws",
      "suite_position": 7,
      "sequence": 4096,
      "round": 0,
      "ms": 1.8823226928710937
    },
    {
      "backend": "triton_ws",
      "suite_position": 7,
      "sequence": 4096,
      "round": 1,
      "ms": 1.94765869140625
    },
    {
      "backend": "triton_ws",
      "suite_position": 7,
      "sequence": 4096,
      "round": 2,
      "ms": 1.921661376953125
    },
    {
      "backend": "triton_ws",
      "suite_position": 7,
      "sequence": 8192,
      "round": 0,
      "ms": 6.856834716796875
    },
    {
      "backend": "triton_ws",
      "suite_position": 7,
      "sequence": 8192,
      "round": 1,
      "ms": 6.965403442382812
    },
    {
      "backend": "triton_ws",
      "suite_position": 7,
      "sequence": 8192,
      "round": 2,
      "ms": 6.9945703125
    },
    {
      "backend": "triton_ws",
      "suite_position": 7,
      "sequence": 16384,
      "round": 0,
      "ms": 26.4671484375
    },
    {
      "backend": "triton_ws",
      "suite_position": 7,
      "sequence": 16384,
      "round": 1,
      "ms": 26.626416015625
    },
    {
      "backend": "triton_ws",
      "suite_position": 7,
      "sequence": 16384,
      "round": 2,
      "ms": 26.63458251953125
    },
    {
      "backend": "fa3",
      "suite_position": 8,
      "sequence": 4096,
      "round": 0,
      "ms": 1.3226588439941407
    },
    {
      "backend": "fa3",
      "suite_position": 8,
      "sequence": 4096,
      "round": 1,
      "ms": 1.397989501953125
    },
    {
      "backend": "fa3",
      "suite_position": 8,
      "sequence": 4096,
      "round": 2,
      "ms": 1.5033905029296875
    },
    {
      "backend": "fa3",
      "suite_position": 8,
      "sequence": 8192,
      "round": 0,
      "ms": 5.347284545898438
    },
    {
      "backend": "fa3",
      "suite_position": 8,
      "sequence": 8192,
      "round": 1,
      "ms": 5.379610595703125
    },
    {
      "backend": "fa3",
      "suite_position": 8,
      "sequence": 8192,
      "round": 2,
      "ms": 5.387033081054687
    },
    {
      "backend": "fa3",
      "suite_position": 8,
      "sequence": 16384,
      "round": 0,
      "ms": 20.8184912109375
    },
    {
      "backend": "fa3",
      "suite_position": 8,
      "sequence": 16384,
      "round": 1,
      "ms": 20.85906005859375
    },
    {
      "backend": "fa3",
      "suite_position": 8,
      "sequence": 16384,
      "round": 2,
      "ms": 20.9091552734375
    }
  ],
  "verification": [
    {
      "backend": "fa3",
      "sequence": 4096,
      "passed": true,
      "checked": 261760,
      "max_abs_err": 0.0025254487991333008,
      "reference": "FP32 selected rows, all output dimensions"
    },
    {
      "backend": "fa3",
      "sequence": 8192,
      "passed": true,
      "checked": 262016,
      "max_abs_err": 0.0028841495513916016,
      "reference": "FP32 selected rows, all output dimensions"
    },
    {
      "backend": "fa3",
      "sequence": 16384,
      "passed": true,
      "checked": 262016,
      "max_abs_err": 0.0029709935188293457,
      "reference": "FP32 selected rows, all output dimensions"
    },
    {
      "backend": "v1",
      "sequence": 4096,
      "ms": null,
      "verification": {
        "passed": true,
        "max_abs_err": 0.00194889307022,
        "checked": 4096
      }
    },
    {
      "backend": "v1",
      "sequence": 8192,
      "ms": null,
      "verification": {
        "passed": true,
        "max_abs_err": 0.00202107429504,
        "checked": 4096
      }
    },
    {
      "backend": "v1",
      "sequence": 16384,
      "ms": null,
      "verification": {
        "passed": true,
        "max_abs_err": 0.0019359588623,
        "checked": 4096
      }
    },
    {
      "backend": "v2",
      "sequence": 4096,
      "ms": null,
      "verification": {
        "passed": true,
        "max_abs_err": 0.00194889307022,
        "checked": 4096
      }
    },
    {
      "backend": "v2",
      "sequence": 8192,
      "ms": null,
      "verification": {
        "passed": true,
        "max_abs_err": 0.00202107429504,
        "checked": 4096
      }
    },
    {
      "backend": "v2",
      "sequence": 16384,
      "ms": null,
      "verification": {
        "passed": true,
        "max_abs_err": 0.0019359588623,
        "checked": 4096
      }
    },
    {
      "backend": "v4",
      "sequence": 4096,
      "ms": null,
      "verification": {
        "passed": true,
        "max_abs_err": 0.00194889307022,
        "checked": 4096
      }
    },
    {
      "backend": "v4",
      "sequence": 8192,
      "ms": null,
      "verification": {
        "passed": true,
        "max_abs_err": 0.00202107429504,
        "checked": 4096
      }
    },
    {
      "backend": "v4",
      "sequence": 16384,
      "ms": null,
      "verification": {
        "passed": true,
        "max_abs_err": 0.0019359588623,
        "checked": 4096
      }
    },
    {
      "backend": "v3",
      "sequence": 4096,
      "ms": null,
      "verification": {
        "passed": true,
        "max_abs_err": 0.00194889307022,
        "checked": 4096
      }
    },
    {
      "backend": "v3",
      "sequence": 8192,
      "ms": null,
      "verification": {
        "passed": true,
        "max_abs_err": 0.00202107429504,
        "checked": 4096
      }
    },
    {
      "backend": "v3",
      "sequence": 16384,
      "ms": null,
      "verification": {
        "passed": true,
        "max_abs_err": 0.0019359588623,
        "checked": 4096
      }
    },
    {
      "backend": "tilelang",
      "sequence": 4096,
      "passed": true,
      "checked": 261760,
      "max_abs_err": 0.0025254487991333008,
      "reference": "FP32 selected rows, all output dimensions"
    },
    {
      "backend": "tilelang",
      "sequence": 8192,
      "passed": true,
      "checked": 262016,
      "max_abs_err": 0.0028841495513916016,
      "reference": "FP32 selected rows, all output dimensions"
    },
    {
      "backend": "tilelang",
      "sequence": 16384,
      "passed": true,
      "checked": 262016,
      "max_abs_err": 0.0029709935188293457,
      "reference": "FP32 selected rows, all output dimensions"
    },
    {
      "backend": "triton",
      "sequence": 4096,
      "passed": true,
      "checked": 261760,
      "max_abs_err": 0.0025254487991333008,
      "reference": "FP32 selected rows, all output dimensions"
    },
    {
      "backend": "triton",
      "sequence": 8192,
      "passed": true,
      "checked": 262016,
      "max_abs_err": 0.0028841495513916016,
      "reference": "FP32 selected rows, all output dimensions"
    },
    {
      "backend": "triton",
      "sequence": 16384,
      "passed": true,
      "checked": 262016,
      "max_abs_err": 0.0029709935188293457,
      "reference": "FP32 selected rows, all output dimensions"
    },
    {
      "backend": "triton_ws",
      "sequence": 4096,
      "passed": true,
      "checked": 261760,
      "max_abs_err": 0.0025254487991333008,
      "reference": "FP32 selected rows, all output dimensions"
    },
    {
      "backend": "triton_ws",
      "sequence": 8192,
      "passed": true,
      "checked": 262016,
      "max_abs_err": 0.0028841495513916016,
      "reference": "FP32 selected rows, all output dimensions"
    },
    {
      "backend": "triton_ws",
      "sequence": 16384,
      "passed": true,
      "checked": 262016,
      "max_abs_err": 0.0029709935188293457,
      "reference": "FP32 selected rows, all output dimensions"
    },
    {
      "backend": "fa3",
      "sequence": 4096,
      "passed": true,
      "checked": 261760,
      "max_abs_err": 0.0025254487991333008,
      "reference": "FP32 selected rows, all output dimensions"
    },
    {
      "backend": "fa3",
      "sequence": 8192,
      "passed": true,
      "checked": 262016,
      "max_abs_err": 0.0028841495513916016,
      "reference": "FP32 selected rows, all output dimensions"
    },
    {
      "backend": "fa3",
      "sequence": 16384,
      "passed": true,
      "checked": 262016,
      "max_abs_err": 0.0029709935188293457,
      "reference": "FP32 selected rows, all output dimensions"
    }
  ],
  "runs": [
    {
      "backend": "fa3",
      "position": 0,
      "gpu_before": "0, NVIDIA H800 PCIe, GPU-c6114605-2f2c-7a74-a505-c53021985b24, 590.48.01, 9.0, 38, 345, 1593, 50.78, 0, 14",
      "gpu_after": "0, NVIDIA H800 PCIe, GPU-c6114605-2f2c-7a74-a505-c53021985b24, 590.48.01, 9.0, 53, 1755, 1593, 170.23, 0, 14"
    },
    {
      "backend": "v1",
      "position": 1,
      "gpu_before": "0, NVIDIA H800 PCIe, GPU-c6114605-2f2c-7a74-a505-c53021985b24, 590.48.01, 9.0, 53, 1755, 1593, 154.64, 0, 14",
      "gpu_after": "0, NVIDIA H800 PCIe, GPU-c6114605-2f2c-7a74-a505-c53021985b24, 590.48.01, 9.0, 59, 1740, 1593, 287.27, 0, 14"
    },
    {
      "backend": "v2",
      "position": 2,
      "gpu_before": "0, NVIDIA H800 PCIe, GPU-c6114605-2f2c-7a74-a505-c53021985b24, 590.48.01, 9.0, 59, 1740, 1593, 271.37, 0, 14",
      "gpu_after": "0, NVIDIA H800 PCIe, GPU-c6114605-2f2c-7a74-a505-c53021985b24, 590.48.01, 9.0, 57, 1740, 1593, 285.88, 35, 14"
    },
    {
      "backend": "v4",
      "position": 3,
      "gpu_before": "0, NVIDIA H800 PCIe, GPU-c6114605-2f2c-7a74-a505-c53021985b24, 590.48.01, 9.0, 56, 1740, 1593, 271.43, 1, 14",
      "gpu_after": "0, NVIDIA H800 PCIe, GPU-c6114605-2f2c-7a74-a505-c53021985b24, 590.48.01, 9.0, 56, 1740, 1593, 283.51, 4, 14"
    },
    {
      "backend": "v3",
      "position": 4,
      "gpu_before": "0, NVIDIA H800 PCIe, GPU-c6114605-2f2c-7a74-a505-c53021985b24, 590.48.01, 9.0, 55, 1740, 1593, 268.43, 1, 14",
      "gpu_after": "0, NVIDIA H800 PCIe, GPU-c6114605-2f2c-7a74-a505-c53021985b24, 590.48.01, 9.0, 55, 1740, 1593, 269.64, 0, 14"
    },
    {
      "backend": "tilelang",
      "position": 5,
      "gpu_before": "0, NVIDIA H800 PCIe, GPU-c6114605-2f2c-7a74-a505-c53021985b24, 590.48.01, 9.0, 55, 1740, 1593, 256.05, 0, 14",
      "gpu_after": "0, NVIDIA H800 PCIe, GPU-c6114605-2f2c-7a74-a505-c53021985b24, 590.48.01, 9.0, 62, 1755, 1593, 203.05, 0, 14"
    },
    {
      "backend": "triton",
      "position": 6,
      "gpu_before": "0, NVIDIA H800 PCIe, GPU-c6114605-2f2c-7a74-a505-c53021985b24, 590.48.01, 9.0, 62, 1755, 1593, 190.30, 2, 14",
      "gpu_after": "0, NVIDIA H800 PCIe, GPU-c6114605-2f2c-7a74-a505-c53021985b24, 590.48.01, 9.0, 69, 1755, 1593, 212.77, 0, 14"
    },
    {
      "backend": "triton_ws",
      "position": 7,
      "gpu_before": "0, NVIDIA H800 PCIe, GPU-c6114605-2f2c-7a74-a505-c53021985b24, 590.48.01, 9.0, 65, 345, 1593, 63.06, 0, 14",
      "gpu_after": "0, NVIDIA H800 PCIe, GPU-c6114605-2f2c-7a74-a505-c53021985b24, 590.48.01, 9.0, 72, 1755, 1593, 225.57, 0, 14"
    },
    {
      "backend": "fa3",
      "position": 8,
      "gpu_before": "0, NVIDIA H800 PCIe, GPU-c6114605-2f2c-7a74-a505-c53021985b24, 590.48.01, 9.0, 64, 345, 1593, 61.99, 0, 14",
      "gpu_after": "0, NVIDIA H800 PCIe, GPU-c6114605-2f2c-7a74-a505-c53021985b24, 590.48.01, 9.0, 71, 1755, 1593, 171.96, 0, 14"
    }
  ],
  "complete": true,
  "versions": {
    "fa3": {
      "torch": "2.11.0+cu130",
      "cuda": "13.0",
      "triton": "3.6.0",
      "tilelang": "0.1.9",
      "flash_attn_3": "3.0.0"
    },
    "tilelang": {
      "torch": "2.11.0+cu130",
      "cuda": "13.0",
      "triton": "3.6.0",
      "tilelang": "0.1.9",
      "flash_attn_3": "3.0.0"
    },
    "triton": {
      "torch": "2.11.0+cu130",
      "cuda": "13.0",
      "triton": "3.6.0",
      "tilelang": "0.1.9",
      "flash_attn_3": "3.0.0"
    },
    "triton_ws": {
      "torch": "2.10.0+cu128",
      "cuda": "12.8",
      "triton": "3.8.0+gita59d32fd"
    }
  },
  "gpu_after": "0, NVIDIA H800 PCIe, GPU-c6114605-2f2c-7a74-a505-c53021985b24, 590.48.01, 9.0, 71, 1755, 1593, 167.23, 0, 14"
}
