{
  "date": "2026-09-11T21:32:54.559Z",
  "rows": [
    {
      "key": "small-saxpy",
      "id": "saxpy",
      "scalars": {
        "a": 0.125,
        "n": 262144
      },
      "webgpu": {
        "key": "small-saxpy",
        "id": "saxpy",
        "scalars": {
          "a": 0.125,
          "n": 262144
        },
        "block": [
          128,
          1,
          1
        ],
        "groups": [
          2048
        ],
        "iterations": 2048,
        "samples": 9,
        "warmup": 5,
        "checks": {
          "y": {
            "pass": true,
            "failures": 0,
            "first": null,
            "maxAbs": 0,
            "maxRel": 0
          }
        },
        "gpuMedianMs": 0.001632,
        "wallMedianMs": 0.0033056640531867743,
        "timings": [
          {
            "gpuMs": 0.001632,
            "wallMs": 0.007543945306679234
          },
          {
            "gpuMs": 0.001664,
            "wallMs": 0.003881835931679234
          },
          {
            "gpuMs": 0.001664,
            "wallMs": 0.0034643554681679234
          },
          {
            "gpuMs": 0.001632,
            "wallMs": 0.0029760742181679234
          },
          {
            "gpuMs": 0.001632,
            "wallMs": 0.0033056640531867743
          },
          {
            "gpuMs": 0.001632,
            "wallMs": 0.002617187492433004
          },
          {
            "gpuMs": 0.001792,
            "wallMs": 0.0028173828177386895
          },
          {
            "gpuMs": 0.001632,
            "wallMs": 0.0035717773425858468
          },
          {
            "gpuMs": 0.001632,
            "wallMs": 0.0030419921968132257
          }
        ]
      },
      "cuda": {
        "key": "small-saxpy",
        "verified": true,
        "iterations": 2048,
        "samples": 9,
        "warmup": 5,
        "gpuMedianMs": 0.001494718716,
        "wallMedianMs": 0.001497851563,
        "directGpuMedianMs": 0.00666149985,
        "directWallMedianMs": 0.006667529297,
        "timings": [
          {
            "gpuMs": 0.00180371874,
            "wallMs": 0.001809960937,
            "directGpuMs": 0.00666149985,
            "directWallMs": 0.006667529297
          },
          {
            "gpuMs": 0.001493578078,
            "wallMs": 0.001496972656,
            "directGpuMs": 0.006717578042,
            "directWallMs": 0.006721386719
          },
          {
            "gpuMs": 0.001494046883,
            "wallMs": 0.001497314453,
            "directGpuMs": 0.00656970311,
            "directWallMs": 0.006572753906
          },
          {
            "gpuMs": 0.001492234413,
            "wallMs": 0.001495263672,
            "directGpuMs": 0.007172984537,
            "directWallMs": 0.007176611328
          },
          {
            "gpuMs": 0.001494718716,
            "wallMs": 0.001497851563,
            "directGpuMs": 0.00721826544,
            "directWallMs": 0.007221777344
          },
          {
            "gpuMs": 0.001493937452,
            "wallMs": 0.001497021484,
            "directGpuMs": 0.007405093871,
            "directWallMs": 0.007408349609
          },
          {
            "gpuMs": 0.001562421909,
            "wallMs": 0.001566503906,
            "directGpuMs": 0.006511656102,
            "directWallMs": 0.006566357422
          },
          {
            "gpuMs": 0.001702046837,
            "wallMs": 0.001705029297,
            "directGpuMs": 0.006362312473,
            "directWallMs": 0.006365625
          },
          {
            "gpuMs": 0.001563843805,
            "wallMs": 0.00156796875,
            "directGpuMs": 0.006214906462,
            "directWallMs": 0.006275683594
          }
        ]
      },
      "ratio": 1.0918442262952168
    },
    {
      "key": "small-saxpy_vec4",
      "id": "saxpy_vec4",
      "scalars": {
        "a": 0.125,
        "n4": 65536
      },
      "webgpu": {
        "key": "small-saxpy_vec4",
        "id": "saxpy_vec4",
        "scalars": {
          "a": 0.125,
          "n4": 65536
        },
        "block": [
          128,
          1,
          1
        ],
        "groups": [
          512
        ],
        "iterations": 2048,
        "samples": 9,
        "warmup": 5,
        "checks": {
          "y": {
            "pass": true,
            "failures": 0,
            "first": null,
            "maxAbs": 0,
            "maxRel": 0
          }
        },
        "gpuMedianMs": 0.00144,
        "wallMedianMs": 0.0029614257800858468,
        "timings": [
          {
            "gpuMs": 0.0016,
            "wallMs": 0.0029614257800858468
          },
          {
            "gpuMs": 0.00144,
            "wallMs": 0.00239746093575377
          },
          {
            "gpuMs": 0.00144,
            "wallMs": 0.002416992181679234
          },
          {
            "gpuMs": 0.00144,
            "wallMs": 0.0025708007742650807
          },
          {
            "gpuMs": 0.001408,
            "wallMs": 0.00335693359375
          },
          {
            "gpuMs": 0.001504,
            "wallMs": 0.0024951171944849193
          },
          {
            "gpuMs": 0.00144,
            "wallMs": 0.0030078125128056854
          },
          {
            "gpuMs": 0.001408,
            "wallMs": 0.0031030273530632257
          },
          {
            "gpuMs": 0.00144,
            "wallMs": 0.0037011718668509275
          }
        ]
      },
      "cuda": {
        "key": "small-saxpy_vec4",
        "verified": true,
        "iterations": 2048,
        "samples": 9,
        "warmup": 5,
        "gpuMedianMs": 0.00140010938,
        "wallMedianMs": 0.001403808594,
        "directGpuMedianMs": 0.006463218946,
        "directWallMedianMs": 0.006570068359,
        "timings": [
          {
            "gpuMs": 0.001720234402,
            "wallMs": 0.001735839844,
            "directGpuMs": 0.006553406361,
            "directWallMs": 0.006570068359
          },
          {
            "gpuMs": 0.001452968805,
            "wallMs": 0.001457373047,
            "directGpuMs": 0.006394437514,
            "directWallMs": 0.006450585938
          },
          {
            "gpuMs": 0.001431203098,
            "wallMs": 0.001434130859,
            "directGpuMs": 0.006342609413,
            "directWallMs": 0.006345507812
          },
          {
            "gpuMs": 0.001405062503,
            "wallMs": 0.001407958984,
            "directGpuMs": 0.006520046853,
            "directWallMs": 0.006575634766
          },
          {
            "gpuMs": 0.00140010938,
            "wallMs": 0.001403808594,
            "directGpuMs": 0.006260703318,
            "directWallMs": 0.006316796875
          },
          {
            "gpuMs": 0.001267703134,
            "wallMs": 0.001270458984,
            "directGpuMs": 0.006743531208,
            "directWallMs": 0.006746337891
          },
          {
            "gpuMs": 0.001265781233,
            "wallMs": 0.001269628906,
            "directGpuMs": 0.006463218946,
            "directWallMs": 0.006588183594
          },
          {
            "gpuMs": 0.001278656302,
            "wallMs": 0.001282080078,
            "directGpuMs": 0.00663312478,
            "directWallMs": 0.006636816406
          },
          {
            "gpuMs": 0.00126695307,
            "wallMs": 0.001269677734,
            "directGpuMs": 0.006305109244,
            "directWallMs": 0.006308251953
          }
        ]
      },
      "ratio": 1.0284910740330873
    },
    {
      "key": "small-matmul_naive",
      "id": "matmul_naive",
      "scalars": {
        "M": 128,
        "N": 128,
        "K": 128
      },
      "webgpu": {
        "key": "small-matmul_naive",
        "id": "matmul_naive",
        "scalars": {
          "M": 128,
          "N": 128,
          "K": 128
        },
        "block": [
          8,
          8,
          1
        ],
        "groups": [
          16,
          16
        ],
        "iterations": 512,
        "samples": 9,
        "warmup": 5,
        "checks": {
          "C": {
            "pass": true,
            "failures": 0,
            "first": null,
            "maxAbs": 0.00000667572021484375,
            "maxRel": 0.0014302512207944628
          }
        },
        "gpuMedianMs": 0.019968,
        "wallMedianMs": 0.024873046844732016,
        "timings": [
          {
            "gpuMs": 0.019328,
            "wallMs": 0.021298828127328306
          },
          {
            "gpuMs": 0.020096,
            "wallMs": 0.026416015636641532
          },
          {
            "gpuMs": 0.020352,
            "wallMs": 0.022919921902939677
          },
          {
            "gpuMs": 0.020224,
            "wallMs": 0.022734374972060323
          },
          {
            "gpuMs": 0.019328,
            "wallMs": 0.023046875023283064
          },
          {
            "gpuMs": 0.019968,
            "wallMs": 0.02690429682843387
          },
          {
            "gpuMs": 0.019968,
            "wallMs": 0.026621093798894435
          },
          {
            "gpuMs": 0.020096,
            "wallMs": 0.02685546875
          },
          {
            "gpuMs": 0.019712,
            "wallMs": 0.024873046844732016
          }
        ]
      },
      "cuda": {
        "key": "small-matmul_naive",
        "verified": true,
        "iterations": 512,
        "samples": 9,
        "warmup": 5,
        "gpuMedianMs": 0.006624499802,
        "wallMedianMs": 0.006635742187,
        "directGpuMedianMs": 0.01024662517,
        "directWallMedianMs": 0.01029628906,
        "timings": [
          {
            "gpuMs": 0.0067542498,
            "wallMs": 0.006815625,
            "directGpuMs": 0.01112762466,
            "directWallMs": 0.01114179688
          },
          {
            "gpuMs": 0.00694981264,
            "wallMs": 0.0069625,
            "directGpuMs": 0.009783499874,
            "directWallMs": 0.009795117188
          },
          {
            "gpuMs": 0.006624999922,
            "wallMs": 0.006635742187,
            "directGpuMs": 0.01131581236,
            "directWallMs": 0.01132792969
          },
          {
            "gpuMs": 0.006624499802,
            "wallMs": 0.006635546875,
            "directGpuMs": 0.0101393126,
            "directWallMs": 0.0101515625
          },
          {
            "gpuMs": 0.006616062485,
            "wallMs": 0.0066265625,
            "directGpuMs": 0.01127218734,
            "directWallMs": 0.01128339844
          },
          {
            "gpuMs": 0.006619624794,
            "wallMs": 0.006630273437,
            "directGpuMs": 0.01024662517,
            "directWallMs": 0.01025703125
          },
          {
            "gpuMs": 0.006616500206,
            "wallMs": 0.006627734375,
            "directGpuMs": 0.01086331252,
            "directWallMs": 0.01090429688
          },
          {
            "gpuMs": 0.006624062546,
            "wallMs": 0.006675195312,
            "directGpuMs": 0.01018243749,
            "directWallMs": 0.01019453125
          },
          {
            "gpuMs": 0.0069565624,
            "wallMs": 0.0069671875,
            "directGpuMs": 0.009704562835,
            "directWallMs": 0.01029628906
          }
        ]
      },
      "ratio": 3.0142653176578658
    },
    {
      "key": "small-matmul_tiled",
      "id": "matmul_tiled",
      "scalars": {
        "M": 128,
        "N": 128,
        "K": 128
      },
      "webgpu": {
        "key": "small-matmul_tiled",
        "id": "matmul_tiled",
        "scalars": {
          "M": 128,
          "N": 128,
          "K": 128
        },
        "block": [
          16,
          16,
          1
        ],
        "groups": [
          8,
          8
        ],
        "iterations": 512,
        "samples": 9,
        "warmup": 5,
        "checks": {
          "C": {
            "pass": true,
            "failures": 0,
            "first": null,
            "maxAbs": 0.00000667572021484375,
            "maxRel": 0.0014302512207944628
          }
        },
        "gpuMedianMs": 0.0032,
        "wallMedianMs": 0.0055175781599245965,
        "timings": [
          {
            "gpuMs": 0.003328,
            "wallMs": 0.005537109391298145
          },
          {
            "gpuMs": 0.003328,
            "wallMs": 0.005419921886641532
          },
          {
            "gpuMs": 0.0032,
            "wallMs": 0.005322265613358468
          },
          {
            "gpuMs": 0.0032,
            "wallMs": 0.0055175781599245965
          },
          {
            "gpuMs": 0.003328,
            "wallMs": 0.006582031201105565
          },
          {
            "gpuMs": 0.0032,
            "wallMs": 0.005380859365686774
          },
          {
            "gpuMs": 0.0032,
            "wallMs": 0.005458984349388629
          },
          {
            "gpuMs": 0.003328,
            "wallMs": 0.009296874981373549
          },
          {
            "gpuMs": 0.0032,
            "wallMs": 0.0074707031599245965
          }
        ]
      },
      "cuda": {
        "key": "small-matmul_tiled",
        "verified": true,
        "iterations": 512,
        "samples": 9,
        "warmup": 5,
        "gpuMedianMs": 0.002884624992,
        "wallMedianMs": 0.002895507812,
        "directGpuMedianMs": 0.006999124773,
        "directWallMedianMs": 0.00701171875,
        "timings": [
          {
            "gpuMs": 0.00300631253,
            "wallMs": 0.00306171875,
            "directGpuMs": 0.006839312613,
            "directWallMs": 0.006850976562
          },
          {
            "gpuMs": 0.002879687585,
            "wallMs": 0.002890625,
            "directGpuMs": 0.006694499869,
            "directWallMs": 0.006705078125
          },
          {
            "gpuMs": 0.002880437532,
            "wallMs": 0.002891210937,
            "directGpuMs": 0.0069085625,
            "directWallMs": 0.00692265625
          },
          {
            "gpuMs": 0.002897625091,
            "wallMs": 0.002917773438,
            "directGpuMs": 0.007057937328,
            "directWallMs": 0.00706953125
          },
          {
            "gpuMs": 0.002883937443,
            "wallMs": 0.002894921875,
            "directGpuMs": 0.007393187378,
            "directWallMs": 0.007405078125
          },
          {
            "gpuMs": 0.002889187541,
            "wallMs": 0.002905273438,
            "directGpuMs": 0.006999124773,
            "directWallMs": 0.00701171875
          },
          {
            "gpuMs": 0.002885875059,
            "wallMs": 0.002898828125,
            "directGpuMs": 0.006790500134,
            "directWallMs": 0.006801171875
          },
          {
            "gpuMs": 0.002884624992,
            "wallMs": 0.002895507812,
            "directGpuMs": 0.007254312281,
            "directWallMs": 0.00726484375
          },
          {
            "gpuMs": 0.002879937412,
            "wallMs": 0.002891210937,
            "directGpuMs": 0.007314937655,
            "directWallMs": 0.007325585938
          }
        ]
      },
      "ratio": 1.109329638644412
    },
    {
      "key": "small-matmul_register",
      "id": "matmul_register",
      "scalars": {
        "M": 128,
        "N": 128,
        "K": 128
      },
      "webgpu": {
        "key": "small-matmul_register",
        "id": "matmul_register",
        "scalars": {
          "M": 128,
          "N": 128,
          "K": 128
        },
        "block": [
          8,
          8,
          1
        ],
        "groups": [
          8,
          8
        ],
        "iterations": 512,
        "samples": 9,
        "warmup": 5,
        "checks": {
          "C": {
            "pass": true,
            "failures": 0,
            "first": null,
            "maxAbs": 0.00000667572021484375,
            "maxRel": 0.0014302512207944628
          }
        },
        "gpuMedianMs": 0.005888,
        "wallMedianMs": 0.011113281245343387,
        "timings": [
          {
            "gpuMs": 0.005888,
            "wallMs": 0.007666015648283064
          },
          {
            "gpuMs": 0.006528,
            "wallMs": 0.009306640655267984
          },
          {
            "gpuMs": 0.00576,
            "wallMs": 0.011201171902939677
          },
          {
            "gpuMs": 0.005888,
            "wallMs": 0.011240234365686774
          },
          {
            "gpuMs": 0.006528,
            "wallMs": 0.01138671871740371
          },
          {
            "gpuMs": 0.005888,
            "wallMs": 0.009179687534924597
          },
          {
            "gpuMs": 0.005888,
            "wallMs": 0.007460937544237822
          },
          {
            "gpuMs": 0.006272,
            "wallMs": 0.011113281245343387
          },
          {
            "gpuMs": 0.005888,
            "wallMs": 0.011484374990686774
          }
        ]
      },
      "cuda": {
        "key": "small-matmul_register",
        "verified": true,
        "iterations": 512,
        "samples": 9,
        "warmup": 5,
        "gpuMedianMs": 0.003062937409,
        "wallMedianMs": 0.003073632813,
        "directGpuMedianMs": 0.008001687936,
        "directWallMedianMs": 0.008012304687,
        "timings": [
          {
            "gpuMs": 0.003182874992,
            "wallMs": 0.003248242188,
            "directGpuMs": 0.009773249738,
            "directWallMs": 0.009791210937
          },
          {
            "gpuMs": 0.00307118753,
            "wallMs": 0.003083984375,
            "directGpuMs": 0.007979437709,
            "directWallMs": 0.007990625
          },
          {
            "gpuMs": 0.003059874987,
            "wallMs": 0.00307265625,
            "directGpuMs": 0.007803124841,
            "directWallMs": 0.007813476562
          },
          {
            "gpuMs": 0.003060812596,
            "wallMs": 0.003072265625,
            "directGpuMs": 0.009174937382,
            "directWallMs": 0.009185546875
          },
          {
            "gpuMs": 0.003061000025,
            "wallMs": 0.003071679687,
            "directGpuMs": 0.007516374812,
            "directWallMs": 0.007526953125
          },
          {
            "gpuMs": 0.003065312514,
            "wallMs": 0.00307578125,
            "directGpuMs": 0.008022562601,
            "directWallMs": 0.008033007812
          },
          {
            "gpuMs": 0.003062937409,
            "wallMs": 0.003073632813,
            "directGpuMs": 0.008881687187,
            "directWallMs": 0.008895507812
          },
          {
            "gpuMs": 0.003066062462,
            "wallMs": 0.003077734375,
            "directGpuMs": 0.008001687936,
            "directWallMs": 0.008012304687
          },
          {
            "gpuMs": 0.003060250077,
            "wallMs": 0.003073242187,
            "directGpuMs": 0.007349874824,
            "directWallMs": 0.007360546875
          }
        ]
      },
      "ratio": 1.9223376823499432
    },
    {
      "key": "small-reduce_sum",
      "id": "reduce_sum",
      "scalars": {
        "n": 262144
      },
      "webgpu": {
        "key": "small-reduce_sum",
        "id": "reduce_sum",
        "scalars": {
          "n": 262144
        },
        "block": [
          128,
          1,
          1
        ],
        "groups": [
          1024
        ],
        "iterations": 2048,
        "samples": 9,
        "warmup": 5,
        "checks": {
          "output": {
            "pass": true,
            "failures": 0,
            "first": null,
            "maxAbs": 0,
            "maxRel": 0
          }
        },
        "gpuMedianMs": 0.00368,
        "wallMedianMs": 0.007780761719914153,
        "timings": [
          {
            "gpuMs": 0.00368,
            "wallMs": 0.010842285162652843
          },
          {
            "gpuMs": 0.00368,
            "wallMs": 0.01041992187674623
          },
          {
            "gpuMs": 0.00368,
            "wallMs": 0.007573242182843387
          },
          {
            "gpuMs": 0.00368,
            "wallMs": 0.007973632818902843
          },
          {
            "gpuMs": 0.003712,
            "wallMs": 0.007514648430515081
          },
          {
            "gpuMs": 0.00368,
            "wallMs": 0.007561035148683004
          },
          {
            "gpuMs": 0.00368,
            "wallMs": 0.008552246101316996
          },
          {
            "gpuMs": 0.00368,
            "wallMs": 0.007780761719914153
          },
          {
            "gpuMs": 0.00368,
            "wallMs": 0.007482910150429234
          }
        ]
      },
      "cuda": {
        "key": "small-reduce_sum",
        "verified": true,
        "iterations": 2048,
        "samples": 9,
        "warmup": 5,
        "gpuMedianMs": 0.003549140645,
        "wallMedianMs": 0.003553320312,
        "directGpuMedianMs": 0.018627204,
        "directWallMedianMs": 0.01863125,
        "timings": [
          {
            "gpuMs": 0.003574390663,
            "wallMs": 0.003581982422,
            "directGpuMs": 0.01863626577,
            "directWallMs": 0.01864106445
          },
          {
            "gpuMs": 0.003576703137,
            "wallMs": 0.003581005859,
            "directGpuMs": 0.01841295324,
            "directWallMs": 0.01847241211
          },
          {
            "gpuMs": 0.003391749924,
            "wallMs": 0.003394775391,
            "directGpuMs": 0.01883826591,
            "directWallMs": 0.01884189453
          },
          {
            "gpuMs": 0.003387718694,
            "wallMs": 0.003390673828,
            "directGpuMs": 0.018627204,
            "directWallMs": 0.01863125
          },
          {
            "gpuMs": 0.003549140645,
            "wallMs": 0.003553320312,
            "directGpuMs": 0.01884410903,
            "directWallMs": 0.01884868164
          },
          {
            "gpuMs": 0.003556968644,
            "wallMs": 0.003560644531,
            "directGpuMs": 0.01837253198,
            "directWallMs": 0.01837675781
          },
          {
            "gpuMs": 0.003390109399,
            "wallMs": 0.003393847656,
            "directGpuMs": 0.01905503124,
            "directWallMs": 0.01905878906
          },
          {
            "gpuMs": 0.003387890523,
            "wallMs": 0.003391064453,
            "directGpuMs": 0.01852195337,
            "directWallMs": 0.01852495117
          },
          {
            "gpuMs": 0.003588171909,
            "wallMs": 0.003592089844,
            "directGpuMs": 0.01831506193,
            "directWallMs": 0.01837314453
          }
        ]
      },
      "ratio": 1.0368707155024566
    },
    {
      "key": "small-convolution",
      "id": "convolution",
      "scalars": {
        "n": 262144
      },
      "webgpu": {
        "key": "small-convolution",
        "id": "convolution",
        "scalars": {
          "n": 262144
        },
        "block": [
          128,
          1,
          1
        ],
        "groups": [
          2048
        ],
        "iterations": 2048,
        "samples": 9,
        "warmup": 5,
        "checks": {
          "output": {
            "pass": true,
            "failures": 0,
            "first": null,
            "maxAbs": 0,
            "maxRel": 0
          }
        },
        "gpuMedianMs": 0.002048,
        "wallMedianMs": 0.0030981445306679234,
        "timings": [
          {
            "gpuMs": 0.002048,
            "wallMs": 0.0038916015619179234
          },
          {
            "gpuMs": 0.002048,
            "wallMs": 0.0030981445306679234
          },
          {
            "gpuMs": 0.002048,
            "wallMs": 0.003588867184589617
          },
          {
            "gpuMs": 0.002048,
            "wallMs": 0.0030786132847424597
          },
          {
            "gpuMs": 0.002048,
            "wallMs": 0.003107910160906613
          },
          {
            "gpuMs": 0.002048,
            "wallMs": 0.0030615234427386895
          },
          {
            "gpuMs": 0.002112,
            "wallMs": 0.0030004882864886895
          },
          {
            "gpuMs": 0.002048,
            "wallMs": 0.0030200195324141532
          },
          {
            "gpuMs": 0.002048,
            "wallMs": 0.0030981445452198386
          }
        ]
      },
      "cuda": {
        "key": "small-convolution",
        "verified": true,
        "iterations": 2048,
        "samples": 9,
        "warmup": 5,
        "gpuMedianMs": 0.001812828123,
        "wallMedianMs": 0.001815722656,
        "directGpuMedianMs": 0.006339828018,
        "directWallMedianMs": 0.006405322266,
        "timings": [
          {
            "gpuMs": 0.001976359403,
            "wallMs": 0.001980371094,
            "directGpuMs": 0.006807984319,
            "directWallMs": 0.006811425781
          },
          {
            "gpuMs": 0.001811437542,
            "wallMs": 0.001815234375,
            "directGpuMs": 0.006339828018,
            "directWallMs": 0.006405322266
          },
          {
            "gpuMs": 0.001812828123,
            "wallMs": 0.001815722656,
            "directGpuMs": 0.006241781171,
            "directWallMs": 0.006299169922
          },
          {
            "gpuMs": 0.001934921835,
            "wallMs": 0.001939257813,
            "directGpuMs": 0.006054312456,
            "directWallMs": 0.006058056641
          },
          {
            "gpuMs": 0.001789749949,
            "wallMs": 0.001792773438,
            "directGpuMs": 0.006207265425,
            "directWallMs": 0.006267480469
          },
          {
            "gpuMs": 0.001834578114,
            "wallMs": 0.001837451172,
            "directGpuMs": 0.006085843779,
            "directWallMs": 0.006088769531
          },
          {
            "gpuMs": 0.001785625005,
            "wallMs": 0.001789453125,
            "directGpuMs": 0.006827859208,
            "directWallMs": 0.006882861328
          },
          {
            "gpuMs": 0.001876171911,
            "wallMs": 0.001879052734,
            "directGpuMs": 0.006894890685,
            "directWallMs": 0.006897851562
          },
          {
            "gpuMs": 0.00167246873,
            "wallMs": 0.001675292969,
            "directGpuMs": 0.006972078234,
            "directWallMs": 0.006975195312
          }
        ]
      },
      "ratio": 1.129726516273821
    },
    {
      "key": "small-histogram",
      "id": "histogram",
      "scalars": {
        "n": 262144
      },
      "webgpu": {
        "key": "small-histogram",
        "id": "histogram",
        "scalars": {
          "n": 262144
        },
        "block": [
          128,
          1,
          1
        ],
        "groups": [
          256
        ],
        "iterations": 2048,
        "samples": 9,
        "warmup": 5,
        "checks": {
          "bins": {
            "pass": true,
            "failures": 0,
            "first": null,
            "maxAbs": 0,
            "maxRel": 0
          }
        },
        "gpuMedianMs": 0.004096,
        "wallMedianMs": 0.007104492193320766,
        "timings": [
          {
            "gpuMs": 0.004096,
            "wallMs": 0.006555175787070766
          },
          {
            "gpuMs": 0.004096,
            "wallMs": 0.0069482421822613105
          },
          {
            "gpuMs": 0.004096,
            "wallMs": 0.007604980462929234
          },
          {
            "gpuMs": 0.004096,
            "wallMs": 0.007104492193320766
          },
          {
            "gpuMs": 0.004096,
            "wallMs": 0.0074560546927386895
          },
          {
            "gpuMs": 0.004096,
            "wallMs": 0.007851562506402843
          },
          {
            "gpuMs": 0.004096,
            "wallMs": 0.006872558587929234
          },
          {
            "gpuMs": 0.004256,
            "wallMs": 0.006628417962929234
          },
          {
            "gpuMs": 0.004288,
            "wallMs": 0.007487792972824536
          }
        ]
      },
      "cuda": {
        "key": "small-histogram",
        "verified": true,
        "iterations": 2048,
        "samples": 9,
        "warmup": 5,
        "gpuMedianMs": 0.003819062607,
        "wallMedianMs": 0.003822167969,
        "directGpuMedianMs": 0.01389850024,
        "directWallMedianMs": 0.0139019043,
        "timings": [
          {
            "gpuMs": 0.004064843524,
            "wallMs": 0.004068798828,
            "directGpuMs": 0.01370785944,
            "directWallMs": 0.01371137695
          },
          {
            "gpuMs": 0.003819062607,
            "wallMs": 0.003822167969,
            "directGpuMs": 0.01412089076,
            "directWallMs": 0.01412475586
          },
          {
            "gpuMs": 0.003818999976,
            "wallMs": 0.003821972656,
            "directGpuMs": 0.01390804723,
            "directWallMs": 0.01391152344
          },
          {
            "gpuMs": 0.003888453124,
            "wallMs": 0.003891552734,
            "directGpuMs": 0.01363385934,
            "directWallMs": 0.01363691406
          },
          {
            "gpuMs": 0.003971296828,
            "wallMs": 0.003974804687,
            "directGpuMs": 0.013308594,
            "directWallMs": 0.01331318359
          },
          {
            "gpuMs": 0.004024796654,
            "wallMs": 0.004028222656,
            "directGpuMs": 0.01394895278,
            "directWallMs": 0.0139527832
          },
          {
            "gpuMs": 0.003817390651,
            "wallMs": 0.003820751953,
            "directGpuMs": 0.01381573454,
            "directWallMs": 0.01381914063
          },
          {
            "gpuMs": 0.003818124998,
            "wallMs": 0.003821044922,
            "directGpuMs": 0.01394868735,
            "directWallMs": 0.01395219727
          },
          {
            "gpuMs": 0.003818796948,
            "wallMs": 0.003821777344,
            "directGpuMs": 0.01389850024,
            "directWallMs": 0.0139019043
          }
        ]
      },
      "ratio": 1.0725144941306797
    },
    {
      "key": "small-transpose",
      "id": "transpose",
      "scalars": {
        "width": 512,
        "height": 512
      },
      "webgpu": {
        "key": "small-transpose",
        "id": "transpose",
        "scalars": {
          "width": 512,
          "height": 512
        },
        "block": [
          32,
          8,
          1
        ],
        "groups": [
          16,
          16
        ],
        "iterations": 2048,
        "samples": 9,
        "warmup": 5,
        "checks": {
          "output": {
            "pass": true,
            "failures": 0,
            "first": null,
            "maxAbs": 0,
            "maxRel": 0
          }
        },
        "gpuMedianMs": 0.00176,
        "wallMedianMs": 0.0030664062505820766,
        "timings": [
          {
            "gpuMs": 0.00176,
            "wallMs": 0.0029199218697613105
          },
          {
            "gpuMs": 0.00192,
            "wallMs": 0.002722167977481149
          },
          {
            "gpuMs": 0.001824,
            "wallMs": 0.004074707030667923
          },
          {
            "gpuMs": 0.001952,
            "wallMs": 0.0028369140636641532
          },
          {
            "gpuMs": 0.001728,
            "wallMs": 0.0035937500069849193
          },
          {
            "gpuMs": 0.001856,
            "wallMs": 0.003928222649847157
          },
          {
            "gpuMs": 0.00176,
            "wallMs": 0.0030664062505820766
          },
          {
            "gpuMs": 0.00176,
            "wallMs": 0.0031054687569849193
          },
          {
            "gpuMs": 0.00176,
            "wallMs": 0.0026025390543509275
          }
        ]
      },
      "cuda": {
        "key": "small-transpose",
        "verified": true,
        "iterations": 2048,
        "samples": 9,
        "warmup": 5,
        "gpuMedianMs": 0.001207937486,
        "wallMedianMs": 0.001210888672,
        "directGpuMedianMs": 0.00673412485,
        "directWallMedianMs": 0.006737695312,
        "timings": [
          {
            "gpuMs": 0.001380703179,
            "wallMs": 0.001384814453,
            "directGpuMs": 0.006940734573,
            "directWallMs": 0.006944335937
          },
          {
            "gpuMs": 0.00121067185,
            "wallMs": 0.001213916016,
            "directGpuMs": 0.00673412485,
            "directWallMs": 0.006737695312
          },
          {
            "gpuMs": 0.001210937509,
            "wallMs": 0.001213916016,
            "directGpuMs": 0.006727218628,
            "directWallMs": 0.006730712891
          },
          {
            "gpuMs": 0.001207937486,
            "wallMs": 0.001210888672,
            "directGpuMs": 0.006733125076,
            "directWallMs": 0.006736572266
          },
          {
            "gpuMs": 0.001206937479,
            "wallMs": 0.001209960938,
            "directGpuMs": 0.006495609414,
            "directWallMs": 0.006498974609
          },
          {
            "gpuMs": 0.001207718742,
            "wallMs": 0.001210644531,
            "directGpuMs": 0.006499202922,
            "directWallMs": 0.006502050781
          },
          {
            "gpuMs": 0.001207578112,
            "wallMs": 0.001210546875,
            "directGpuMs": 0.006990422029,
            "directWallMs": 0.006993310547
          },
          {
            "gpuMs": 0.001206703135,
            "wallMs": 0.001209375,
            "directGpuMs": 0.007023390848,
            "directWallMs": 0.007026464844
          },
          {
            "gpuMs": 0.001208031201,
            "wallMs": 0.0012109375,
            "directGpuMs": 0.007060062606,
            "directWallMs": 0.007064453125
          }
        ]
      },
      "ratio": 1.457029043637114
    },
    {
      "key": "small-particles",
      "id": "particles",
      "scalars": {
        "n": 65536,
        "dt": 0.01600000075995922,
        "time": 2.5,
        "attraction": 1
      },
      "webgpu": {
        "key": "small-particles",
        "id": "particles",
        "scalars": {
          "n": 65536,
          "dt": 0.01600000075995922,
          "time": 2.5,
          "attraction": 1
        },
        "block": [
          128,
          1,
          1
        ],
        "groups": [
          512
        ],
        "iterations": 2048,
        "samples": 9,
        "warmup": 5,
        "checks": {
          "position": {
            "pass": true,
            "failures": 0,
            "first": null,
            "maxAbs": 1.1920928955078125e-7,
            "maxRel": 0.000014451222672921183
          },
          "velocity": {
            "pass": true,
            "failures": 0,
            "first": null,
            "maxAbs": 1.1920928955078125e-7,
            "maxRel": 0.00008962848034591471
          }
        },
        "gpuMedianMs": 0.00144,
        "wallMedianMs": 0.0033105468755820766,
        "timings": [
          {
            "gpuMs": 0.001504,
            "wallMs": 0.004060058592585847
          },
          {
            "gpuMs": 0.00144,
            "wallMs": 0.0030249023548094556
          },
          {
            "gpuMs": 0.001536,
            "wallMs": 0.003361816401593387
          },
          {
            "gpuMs": 0.00144,
            "wallMs": 0.0033349609293509275
          },
          {
            "gpuMs": 0.00144,
            "wallMs": 0.0031103515648283064
          },
          {
            "gpuMs": 0.001408,
            "wallMs": 0.0033105468755820766
          },
          {
            "gpuMs": 0.00144,
            "wallMs": 0.0031225585989886895
          },
          {
            "gpuMs": 0.00144,
            "wallMs": 0.0037841796875
          },
          {
            "gpuMs": 0.00144,
            "wallMs": 0.0023828124976716936
          }
        ]
      },
      "cuda": {
        "key": "small-particles",
        "verified": true,
        "iterations": 2048,
        "samples": 9,
        "warmup": 5,
        "gpuMedianMs": 0.001740109408,
        "wallMedianMs": 0.001743164062,
        "directGpuMedianMs": 0.006741437595,
        "directWallMedianMs": 0.006744287109,
        "timings": [
          {
            "gpuMs": 0.001816859352,
            "wallMs": 0.001862792969,
            "directGpuMs": 0.007195843849,
            "directWallMs": 0.007199658203
          },
          {
            "gpuMs": 0.001740078093,
            "wallMs": 0.001743164062,
            "directGpuMs": 0.007450171746,
            "directWallMs": 0.007453662109
          },
          {
            "gpuMs": 0.001740109408,
            "wallMs": 0.001743115234,
            "directGpuMs": 0.006813750137,
            "directWallMs": 0.006817236328
          },
          {
            "gpuMs": 0.001739515574,
            "wallMs": 0.001742480469,
            "directGpuMs": 0.007075515576,
            "directWallMs": 0.007078955078
          },
          {
            "gpuMs": 0.001740687527,
            "wallMs": 0.001743701172,
            "directGpuMs": 0.006648578215,
            "directWallMs": 0.006652148437
          },
          {
            "gpuMs": 0.001741890679,
            "wallMs": 0.001744873047,
            "directGpuMs": 0.006608500145,
            "directWallMs": 0.006611279297
          },
          {
            "gpuMs": 0.001739484374,
            "wallMs": 0.001742382813,
            "directGpuMs": 0.006660671905,
            "directWallMs": 0.006664257813
          },
          {
            "gpuMs": 0.001741953078,
            "wallMs": 0.001744873047,
            "directGpuMs": 0.006681062281,
            "directWallMs": 0.006684130859
          },
          {
            "gpuMs": 0.001739906264,
            "wallMs": 0.001742871094,
            "directGpuMs": 0.006741437595,
            "directWallMs": 0.006744287109
          }
        ]
      },
      "ratio": 0.827534173069651
    },
    {
      "key": "large-saxpy",
      "id": "saxpy",
      "scalars": {
        "a": 0.125,
        "n": 4194304
      },
      "webgpu": {
        "key": "large-saxpy",
        "id": "saxpy",
        "scalars": {
          "a": 0.125,
          "n": 4194304
        },
        "block": [
          128,
          1,
          1
        ],
        "groups": [
          32768
        ],
        "iterations": 2048,
        "samples": 9,
        "warmup": 5,
        "checks": {
          "y": {
            "pass": true,
            "failures": 0,
            "first": null,
            "maxAbs": 0,
            "maxRel": 0
          }
        },
        "gpuMedianMs": 0.016256,
        "wallMedianMs": 0.017514648439828306,
        "timings": [
          {
            "gpuMs": 0.016288,
            "wallMs": 0.017548828123835847
          },
          {
            "gpuMs": 0.016256,
            "wallMs": 0.01757568358152639
          },
          {
            "gpuMs": 0.016256,
            "wallMs": 0.017514648439828306
          },
          {
            "gpuMs": 0.01616,
            "wallMs": 0.017519531247671694
          },
          {
            "gpuMs": 0.016256,
            "wallMs": 0.017458496091421694
          },
          {
            "gpuMs": 0.01616,
            "wallMs": 0.017082519523683004
          },
          {
            "gpuMs": 0.016256,
            "wallMs": 0.017592773438082077
          },
          {
            "gpuMs": 0.01616,
            "wallMs": 0.017478027337347157
          },
          {
            "gpuMs": 0.016224,
            "wallMs": 0.01750732421351131
          }
        ]
      },
      "cuda": {
        "key": "large-saxpy",
        "verified": true,
        "iterations": 2048,
        "samples": 9,
        "warmup": 5,
        "gpuMedianMs": 0.01615923457,
        "wallMedianMs": 0.01619003906,
        "directGpuMedianMs": 0.01887839101,
        "directWallMedianMs": 0.01889189453,
        "timings": [
          {
            "gpuMs": 0.01607945375,
            "wallMs": 0.0161222168,
            "directGpuMs": 0.01882496849,
            "directWallMs": 0.01882988281
          },
          {
            "gpuMs": 0.01615923457,
            "wallMs": 0.01616376953,
            "directGpuMs": 0.01887839101,
            "directWallMs": 0.01888320313
          },
          {
            "gpuMs": 0.0161049217,
            "wallMs": 0.01614575195,
            "directGpuMs": 0.01888520271,
            "directWallMs": 0.0189440918
          },
          {
            "gpuMs": 0.01620045304,
            "wallMs": 0.01620537109,
            "directGpuMs": 0.01895223372,
            "directWallMs": 0.01895722656
          },
          {
            "gpuMs": 0.0161400307,
            "wallMs": 0.01619003906,
            "directGpuMs": 0.01883537509,
            "directWallMs": 0.01889399414
          },
          {
            "gpuMs": 0.01619631238,
            "wallMs": 0.0162003418,
            "directGpuMs": 0.01885263994,
            "directWallMs": 0.01885688477
          },
          {
            "gpuMs": 0.01632850058,
            "wallMs": 0.01641547852,
            "directGpuMs": 0.01891811006,
            "directWallMs": 0.01892285156
          },
          {
            "gpuMs": 0.01620539092,
            "wallMs": 0.01620961914,
            "directGpuMs": 0.01888771914,
            "directWallMs": 0.01889189453
          },
          {
            "gpuMs": 0.01605537534,
            "wallMs": 0.01613193359,
            "directGpuMs": 0.01881396957,
            "directWallMs": 0.01886923828
          }
        ]
      },
      "ratio": 1.0059882434146756
    },
    {
      "key": "large-saxpy_vec4",
      "id": "saxpy_vec4",
      "scalars": {
        "a": 0.125,
        "n4": 1048576
      },
      "webgpu": {
        "key": "large-saxpy_vec4",
        "id": "saxpy_vec4",
        "scalars": {
          "a": 0.125,
          "n4": 1048576
        },
        "block": [
          128,
          1,
          1
        ],
        "groups": [
          8192
        ],
        "iterations": 2048,
        "samples": 9,
        "warmup": 5,
        "checks": {
          "y": {
            "pass": true,
            "failures": 0,
            "first": null,
            "maxAbs": 0,
            "maxRel": 0
          }
        },
        "gpuMedianMs": 0.010016,
        "wallMedianMs": 0.011249999995925464,
        "timings": [
          {
            "gpuMs": 0.009984,
            "wallMs": 0.011147460929350927
          },
          {
            "gpuMs": 0.009952,
            "wallMs": 0.011249999995925464
          },
          {
            "gpuMs": 0.010016,
            "wallMs": 0.011621093755820766
          },
          {
            "gpuMs": 0.010048,
            "wallMs": 0.012238769530085847
          },
          {
            "gpuMs": 0.010016,
            "wallMs": 0.01094970702251885
          },
          {
            "gpuMs": 0.010016,
            "wallMs": 0.011640624987194315
          },
          {
            "gpuMs": 0.010048,
            "wallMs": 0.01173583984200377
          },
          {
            "gpuMs": 0.010016,
            "wallMs": 0.011164550771354698
          },
          {
            "gpuMs": 0.009984,
            "wallMs": 0.011137695313664153
          }
        ]
      },
      "cuda": {
        "key": "large-saxpy_vec4",
        "verified": true,
        "iterations": 2048,
        "samples": 9,
        "warmup": 5,
        "gpuMedianMs": 0.01007456239,
        "wallMedianMs": 0.01007963867,
        "directGpuMedianMs": 0.01267342176,
        "directWallMedianMs": 0.01270795898,
        "timings": [
          {
            "gpuMs": 0.01031648461,
            "wallMs": 0.01032163086,
            "directGpuMs": 0.01267342176,
            "directWallMs": 0.01273310547
          },
          {
            "gpuMs": 0.01023131236,
            "wallMs": 0.01023681641,
            "directGpuMs": 0.01261187531,
            "directWallMs": 0.01266845703
          },
          {
            "gpuMs": 0.0101924222,
            "wallMs": 0.0101972168,
            "directGpuMs": 0.01258434355,
            "directWallMs": 0.01264160156
          },
          {
            "gpuMs": 0.01001531258,
            "wallMs": 0.01001962891,
            "directGpuMs": 0.01247334387,
            "directWallMs": 0.01253022461
          },
          {
            "gpuMs": 0.01003831252,
            "wallMs": 0.01004287109,
            "directGpuMs": 0.01254721917,
            "directWallMs": 0.01260512695
          },
          {
            "gpuMs": 0.009966859594,
            "wallMs": 0.009971630859,
            "directGpuMs": 0.01270365622,
            "directWallMs": 0.01270795898
          },
          {
            "gpuMs": 0.01007456239,
            "wallMs": 0.01007963867,
            "directGpuMs": 0.01271198411,
            "directWallMs": 0.01271689453
          },
          {
            "gpuMs": 0.01001378149,
            "wallMs": 0.01001733398,
            "directGpuMs": 0.01278404705,
            "directWallMs": 0.01278764648
          },
          {
            "gpuMs": 0.01010475028,
            "wallMs": 0.01010859375,
            "directGpuMs": 0.01288948394,
            "directWallMs": 0.01289335937
          }
        ]
      },
      "ratio": 0.9941871033467292
    },
    {
      "key": "large-matmul_naive",
      "id": "matmul_naive",
      "scalars": {
        "M": 512,
        "N": 512,
        "K": 512
      },
      "webgpu": {
        "key": "large-matmul_naive",
        "id": "matmul_naive",
        "scalars": {
          "M": 512,
          "N": 512,
          "K": 512
        },
        "block": [
          8,
          8,
          1
        ],
        "groups": [
          64,
          64
        ],
        "iterations": 512,
        "samples": 9,
        "warmup": 5,
        "checks": {
          "C": {
            "pass": true,
            "failures": 0,
            "first": null,
            "maxAbs": 0.0000324249267578125,
            "maxRel": 0.0155089697832781
          }
        },
        "gpuMedianMs": 0.231424,
        "wallMedianMs": 0.23583007813431323,
        "timings": [
          {
            "gpuMs": 0.231552,
            "wallMs": 0.23476562497671694
          },
          {
            "gpuMs": 0.231552,
            "wallMs": 0.23629882815293968
          },
          {
            "gpuMs": 0.23168,
            "wallMs": 0.2344921875046566
          },
          {
            "gpuMs": 0.230912,
            "wallMs": 0.23434570309473202
          },
          {
            "gpuMs": 0.230528,
            "wallMs": 0.23232421872671694
          },
          {
            "gpuMs": 0.230912,
            "wallMs": 0.23583007813431323
          },
          {
            "gpuMs": 0.23104,
            "wallMs": 0.236572265625
          },
          {
            "gpuMs": 0.232064,
            "wallMs": 0.23811523435870185
          },
          {
            "gpuMs": 0.231424,
            "wallMs": 0.238037109375
          }
        ]
      },
      "cuda": {
        "key": "large-matmul_naive",
        "verified": true,
        "iterations": 512,
        "samples": 9,
        "warmup": 5,
        "gpuMedianMs": 0.1071816906,
        "wallMedianMs": 0.1071992188,
        "directGpuMedianMs": 0.1096369997,
        "directWallMedianMs": 0.1097996094,
        "timings": [
          {
            "gpuMs": 0.107058309,
            "wallMs": 0.1070792969,
            "directGpuMs": 0.1095702499,
            "directWallMs": 0.1097996094
          },
          {
            "gpuMs": 0.1070136279,
            "wallMs": 0.1070326172,
            "directGpuMs": 0.1096369997,
            "directWallMs": 0.1096566406
          },
          {
            "gpuMs": 0.1069503725,
            "wallMs": 0.1071125,
            "directGpuMs": 0.1093476266,
            "directWallMs": 0.1093636719
          },
          {
            "gpuMs": 0.1077056229,
            "wallMs": 0.1077226563,
            "directGpuMs": 0.1086299345,
            "directWallMs": 0.1086609375
          },
          {
            "gpuMs": 0.1071008742,
            "wallMs": 0.1071205078,
            "directGpuMs": 0.1099029407,
            "directWallMs": 0.1099222656
          },
          {
            "gpuMs": 0.1071816906,
            "wallMs": 0.1071992188,
            "directGpuMs": 0.1088791862,
            "directWallMs": 0.1088943359
          },
          {
            "gpuMs": 0.1079326868,
            "wallMs": 0.1079498047,
            "directGpuMs": 0.1099477485,
            "directWallMs": 0.1099640625
          },
          {
            "gpuMs": 0.1085341871,
            "wallMs": 0.1085494141,
            "directGpuMs": 0.1110989973,
            "directWallMs": 0.1111255859
          },
          {
            "gpuMs": 0.1078167483,
            "wallMs": 0.1078347656,
            "directGpuMs": 0.1106809974,
            "directWallMs": 0.1106990234
          }
        ]
      },
      "ratio": 2.1591747499455844
    },
    {
      "key": "large-matmul_tiled",
      "id": "matmul_tiled",
      "scalars": {
        "M": 512,
        "N": 512,
        "K": 512
      },
      "webgpu": {
        "key": "large-matmul_tiled",
        "id": "matmul_tiled",
        "scalars": {
          "M": 512,
          "N": 512,
          "K": 512
        },
        "block": [
          16,
          16,
          1
        ],
        "groups": [
          32,
          32
        ],
        "iterations": 512,
        "samples": 9,
        "warmup": 5,
        "checks": {
          "C": {
            "pass": true,
            "failures": 0,
            "first": null,
            "maxAbs": 0.0000324249267578125,
            "maxRel": 0.0155089697832781
          }
        },
        "gpuMedianMs": 0.064256,
        "wallMedianMs": 0.06685546872904524,
        "timings": [
          {
            "gpuMs": 0.064,
            "wallMs": 0.06685546872904524
          },
          {
            "gpuMs": 0.063616,
            "wallMs": 0.06832031247904524
          },
          {
            "gpuMs": 0.06464,
            "wallMs": 0.06649414059938863
          },
          {
            "gpuMs": 0.064768,
            "wallMs": 0.06823242187965661
          },
          {
            "gpuMs": 0.064256,
            "wallMs": 0.0667089843773283
          },
          {
            "gpuMs": 0.064384,
            "wallMs": 0.06591796875
          },
          {
            "gpuMs": 0.06464,
            "wallMs": 0.06981445313431323
          },
          {
            "gpuMs": 0.064256,
            "wallMs": 0.06580078124534339
          },
          {
            "gpuMs": 0.064128,
            "wallMs": 0.06917968747438863
          }
        ]
      },
      "cuda": {
        "key": "large-matmul_tiled",
        "verified": true,
        "iterations": 512,
        "samples": 9,
        "warmup": 5,
        "gpuMedianMs": 0.06308462471,
        "wallMedianMs": 0.06310820312,
        "directGpuMedianMs": 0.06520231068,
        "directWallMedianMs": 0.06530332031,
        "timings": [
          {
            "gpuMs": 0.06309762597,
            "wallMs": 0.06311582031,
            "directGpuMs": 0.06520231068,
            "directWallMs": 0.06522050781
          },
          {
            "gpuMs": 0.06308462471,
            "wallMs": 0.06310253906,
            "directGpuMs": 0.06489868462,
            "directWallMs": 0.06491640625
          },
          {
            "gpuMs": 0.06298668683,
            "wallMs": 0.06315800781,
            "directGpuMs": 0.06505431235,
            "directWallMs": 0.06527578125
          },
          {
            "gpuMs": 0.06333706528,
            "wallMs": 0.06351132813,
            "directGpuMs": 0.06511956453,
            "directWallMs": 0.06534238281
          },
          {
            "gpuMs": 0.06221231073,
            "wallMs": 0.06238398437,
            "directGpuMs": 0.06609806418,
            "directWallMs": 0.06611894531
          },
          {
            "gpuMs": 0.06231449917,
            "wallMs": 0.06247539063,
            "directGpuMs": 0.06614762545,
            "directWallMs": 0.06616621094
          },
          {
            "gpuMs": 0.06225531176,
            "wallMs": 0.06227558594,
            "directGpuMs": 0.06515468657,
            "directWallMs": 0.06517363281
          },
          {
            "gpuMs": 0.06309218705,
            "wallMs": 0.06310820312,
            "directGpuMs": 0.06528750062,
            "directWallMs": 0.06530332031
          },
          {
            "gpuMs": 0.06323968619,
            "wallMs": 0.06325527344,
            "directGpuMs": 0.06535500288,
            "directWallMs": 0.06537207031
          }
        ]
      },
      "ratio": 1.0185683166918216
    },
    {
      "key": "large-matmul_register",
      "id": "matmul_register",
      "scalars": {
        "M": 512,
        "N": 512,
        "K": 512
      },
      "webgpu": {
        "key": "large-matmul_register",
        "id": "matmul_register",
        "scalars": {
          "M": 512,
          "N": 512,
          "K": 512
        },
        "block": [
          8,
          8,
          1
        ],
        "groups": [
          32,
          32
        ],
        "iterations": 512,
        "samples": 9,
        "warmup": 5,
        "checks": {
          "C": {
            "pass": true,
            "failures": 0,
            "first": null,
            "maxAbs": 0.0000324249267578125,
            "maxRel": 0.0155089697832781
          }
        },
        "gpuMedianMs": 0.046208,
        "wallMedianMs": 0.04900390625698492,
        "timings": [
          {
            "gpuMs": 0.047488,
            "wallMs": 0.051025390625
          },
          {
            "gpuMs": 0.045952,
            "wallMs": 0.04889648436801508
          },
          {
            "gpuMs": 0.045952,
            "wallMs": 0.04893554688896984
          },
          {
            "gpuMs": 0.046464,
            "wallMs": 0.050751953094732016
          },
          {
            "gpuMs": 0.04608,
            "wallMs": 0.04900390625698492
          },
          {
            "gpuMs": 0.046976,
            "wallMs": 0.05038085940759629
          },
          {
            "gpuMs": 0.046336,
            "wallMs": 0.048818359384313226
          },
          {
            "gpuMs": 0.046208,
            "wallMs": 0.04896484379423782
          },
          {
            "gpuMs": 0.04608,
            "wallMs": 0.05017578124534339
          }
        ]
      },
      "cuda": {
        "key": "large-matmul_register",
        "verified": true,
        "iterations": 512,
        "samples": 9,
        "warmup": 5,
        "gpuMedianMs": 0.04308368638,
        "wallMedianMs": 0.04318808594,
        "directGpuMedianMs": 0.04582349956,
        "directWallMedianMs": 0.04589785156,
        "timings": [
          {
            "gpuMs": 0.0441048108,
            "wallMs": 0.04412304688,
            "directGpuMs": 0.04549499974,
            "directWallMs": 0.0455171875
          },
          {
            "gpuMs": 0.04363143817,
            "wallMs": 0.04365078125,
            "directGpuMs": 0.04535562545,
            "directWallMs": 0.04568632813
          },
          {
            "gpuMs": 0.04297662526,
            "wallMs": 0.04299355469,
            "directGpuMs": 0.04672574997,
            "directWallMs": 0.046746875
          },
          {
            "gpuMs": 0.04308368638,
            "wallMs": 0.04310117187,
            "directGpuMs": 0.04606231302,
            "directWallMs": 0.04607949219
          },
          {
            "gpuMs": 0.04301756248,
            "wallMs": 0.04318808594,
            "directGpuMs": 0.04575987533,
            "directWallMs": 0.04577832031
          },
          {
            "gpuMs": 0.04400868714,
            "wallMs": 0.044028125,
            "directGpuMs": 0.04587968811,
            "directWallMs": 0.04589785156
          },
          {
            "gpuMs": 0.04303700104,
            "wallMs": 0.04305527344,
            "directGpuMs": 0.04582349956,
            "directWallMs": 0.04636054687
          },
          {
            "gpuMs": 0.04317462444,
            "wallMs": 0.04318867188,
            "directGpuMs": 0.04624768719,
            "directWallMs": 0.04626503906
          },
          {
            "gpuMs": 0.04292468727,
            "wallMs": 0.04309257813,
            "directGpuMs": 0.04570674896,
            "directWallMs": 0.045725
          }
        ]
      },
      "ratio": 1.072517323435219
    },
    {
      "key": "large-reduce_sum",
      "id": "reduce_sum",
      "scalars": {
        "n": 4194304
      },
      "webgpu": {
        "key": "large-reduce_sum",
        "id": "reduce_sum",
        "scalars": {
          "n": 4194304
        },
        "block": [
          128,
          1,
          1
        ],
        "groups": [
          16384
        ],
        "iterations": 2048,
        "samples": 9,
        "warmup": 5,
        "checks": {
          "output": {
            "pass": true,
            "failures": 0,
            "first": null,
            "maxAbs": 0,
            "maxRel": 0
          }
        },
        "gpuMedianMs": 0.020608,
        "wallMedianMs": 0.025571289064828306,
        "timings": [
          {
            "gpuMs": 0.020512,
            "wallMs": 0.025595703118597157
          },
          {
            "gpuMs": 0.020704,
            "wallMs": 0.026352539061917923
          },
          {
            "gpuMs": 0.020608,
            "wallMs": 0.025405273438082077
          },
          {
            "gpuMs": 0.02048,
            "wallMs": 0.025383300773683004
          },
          {
            "gpuMs": 0.020768,
            "wallMs": 0.025883789072395302
          },
          {
            "gpuMs": 0.02048,
            "wallMs": 0.025571289064828306
          },
          {
            "gpuMs": 0.02064,
            "wallMs": 0.025957031262805685
          },
          {
            "gpuMs": 0.020672,
            "wallMs": 0.024934082030085847
          },
          {
            "gpuMs": 0.020512,
            "wallMs": 0.02541748047224246
          }
        ]
      },
      "cuda": {
        "key": "large-reduce_sum",
        "verified": true,
        "iterations": 2048,
        "samples": 9,
        "warmup": 5,
        "gpuMedianMs": 0.01871031336,
        "wallMedianMs": 0.01871420898,
        "directGpuMedianMs": 0.02536242269,
        "directWallMedianMs": 0.02536655273,
        "timings": [
          {
            "gpuMs": 0.01864484325,
            "wallMs": 0.01868706055,
            "directGpuMs": 0.02519448474,
            "directWallMs": 0.02520068359
          },
          {
            "gpuMs": 0.01865684427,
            "wallMs": 0.01866132813,
            "directGpuMs": 0.02348562516,
            "directWallMs": 0.02349023437
          },
          {
            "gpuMs": 0.01879790612,
            "wallMs": 0.01880268555,
            "directGpuMs": 0.02552993782,
            "directWallMs": 0.02558779297
          },
          {
            "gpuMs": 0.01866976544,
            "wallMs": 0.0186746582,
            "directGpuMs": 0.02522610873,
            "directWallMs": 0.02523076172
          },
          {
            "gpuMs": 0.01875364035,
            "wallMs": 0.01875830078,
            "directGpuMs": 0.02480393834,
            "directWallMs": 0.02480878906
          },
          {
            "gpuMs": 0.01871031336,
            "wallMs": 0.01871420898,
            "directGpuMs": 0.02536242269,
            "directWallMs": 0.02536655273
          },
          {
            "gpuMs": 0.01882526651,
            "wallMs": 0.01882988281,
            "directGpuMs": 0.02647646889,
            "directWallMs": 0.026534375
          },
          {
            "gpuMs": 0.01867318712,
            "wallMs": 0.01867841797,
            "directGpuMs": 0.02561918832,
            "directWallMs": 0.02562387695
          },
          {
            "gpuMs": 0.01886068657,
            "wallMs": 0.01886777344,
            "directGpuMs": 0.02590454742,
            "directWallMs": 0.025909375
          }
        ]
      },
      "ratio": 1.1014246316182488
    },
    {
      "key": "large-convolution",
      "id": "convolution",
      "scalars": {
        "n": 4194304
      },
      "webgpu": {
        "key": "large-convolution",
        "id": "convolution",
        "scalars": {
          "n": 4194304
        },
        "block": [
          128,
          1,
          1
        ],
        "groups": [
          32768
        ],
        "iterations": 2048,
        "samples": 9,
        "warmup": 5,
        "checks": {
          "output": {
            "pass": true,
            "failures": 0,
            "first": null,
            "maxAbs": 0,
            "maxRel": 0
          }
        },
        "gpuMedianMs": 0.021088,
        "wallMedianMs": 0.02223388671700377,
        "timings": [
          {
            "gpuMs": 0.021088,
            "wallMs": 0.022175292964675464
          },
          {
            "gpuMs": 0.021088,
            "wallMs": 0.022988281256402843
          },
          {
            "gpuMs": 0.020896,
            "wallMs": 0.02223388671700377
          },
          {
            "gpuMs": 0.021184,
            "wallMs": 0.022192382806679234
          },
          {
            "gpuMs": 0.020992,
            "wallMs": 0.021896972655667923
          },
          {
            "gpuMs": 0.02112,
            "wallMs": 0.022863769525429234
          },
          {
            "gpuMs": 0.021248,
            "wallMs": 0.02239257813198492
          },
          {
            "gpuMs": 0.02096,
            "wallMs": 0.02220947266323492
          },
          {
            "gpuMs": 0.02112,
            "wallMs": 0.022355957044055685
          }
        ]
      },
      "cuda": {
        "key": "large-convolution",
        "verified": true,
        "iterations": 2048,
        "samples": 9,
        "warmup": 5,
        "gpuMedianMs": 0.01638682745,
        "wallMedianMs": 0.01639956055,
        "directGpuMedianMs": 0.01899228059,
        "directWallMedianMs": 0.0189965332,
        "timings": [
          {
            "gpuMs": 0.0165521875,
            "wallMs": 0.01655795898,
            "directGpuMs": 0.01888399944,
            "directWallMs": 0.01888876953
          },
          {
            "gpuMs": 0.01638682745,
            "wallMs": 0.01639047852,
            "directGpuMs": 0.01908179745,
            "directWallMs": 0.01908598633
          },
          {
            "gpuMs": 0.01634482853,
            "wallMs": 0.01635058594,
            "directGpuMs": 0.01884492114,
            "directWallMs": 0.01884990234
          },
          {
            "gpuMs": 0.01638296805,
            "wallMs": 0.01638735352,
            "directGpuMs": 0.0191539228,
            "directWallMs": 0.01915888672
          },
          {
            "gpuMs": 0.01639353111,
            "wallMs": 0.01639956055,
            "directGpuMs": 0.01891103201,
            "directWallMs": 0.01891577148
          },
          {
            "gpuMs": 0.01634432748,
            "wallMs": 0.01634838867,
            "directGpuMs": 0.01904370263,
            "directWallMs": 0.01904819336
          },
          {
            "gpuMs": 0.01638253033,
            "wallMs": 0.01642607422,
            "directGpuMs": 0.01888701506,
            "directWallMs": 0.01889179687
          },
          {
            "gpuMs": 0.01648774929,
            "wallMs": 0.01649165039,
            "directGpuMs": 0.01909934357,
            "directWallMs": 0.01910375977
          },
          {
            "gpuMs": 0.01644218713,
            "wallMs": 0.0164465332,
            "directGpuMs": 0.01899228059,
            "directWallMs": 0.0189965332
          }
        ]
      },
      "ratio": 1.286887291902253
    },
    {
      "key": "large-histogram",
      "id": "histogram",
      "scalars": {
        "n": 4194304
      },
      "webgpu": {
        "key": "large-histogram",
        "id": "histogram",
        "scalars": {
          "n": 4194304
        },
        "block": [
          128,
          1,
          1
        ],
        "groups": [
          256
        ],
        "iterations": 2048,
        "samples": 9,
        "warmup": 5,
        "checks": {
          "bins": {
            "pass": true,
            "failures": 0,
            "first": null,
            "maxAbs": 0,
            "maxRel": 0
          }
        },
        "gpuMedianMs": 0.024704,
        "wallMedianMs": 0.02785156251047738,
        "timings": [
          {
            "gpuMs": 0.024992,
            "wallMs": 0.027839355476316996
          },
          {
            "gpuMs": 0.024768,
            "wallMs": 0.028872070310171694
          },
          {
            "gpuMs": 0.024704,
            "wallMs": 0.028374023429933004
          },
          {
            "gpuMs": 0.02464,
            "wallMs": 0.028105468751164153
          },
          {
            "gpuMs": 0.024704,
            "wallMs": 0.028332519519608468
          },
          {
            "gpuMs": 0.0248,
            "wallMs": 0.02785156251047738
          },
          {
            "gpuMs": 0.024736,
            "wallMs": 0.02739501952601131
          },
          {
            "gpuMs": 0.024672,
            "wallMs": 0.027355957034160383
          },
          {
            "gpuMs": 0.024704,
            "wallMs": 0.027785644531832077
          }
        ]
      },
      "cuda": {
        "key": "large-histogram",
        "verified": true,
        "iterations": 2048,
        "samples": 9,
        "warmup": 5,
        "gpuMedianMs": 0.01986296847,
        "wallMedianMs": 0.01986689453,
        "directGpuMedianMs": 0.02335803211,
        "directWallMedianMs": 0.0233628418,
        "timings": [
          {
            "gpuMs": 0.01973018795,
            "wallMs": 0.01973608398,
            "directGpuMs": 0.02417176589,
            "directWallMs": 0.02417641602
          },
          {
            "gpuMs": 0.01986296847,
            "wallMs": 0.01986689453,
            "directGpuMs": 0.0231561251,
            "directWallMs": 0.0231605957
          },
          {
            "gpuMs": 0.01989745349,
            "wallMs": 0.01990170898,
            "directGpuMs": 0.02450182848,
            "directWallMs": 0.02450629883
          },
          {
            "gpuMs": 0.01985967159,
            "wallMs": 0.01986416016,
            "directGpuMs": 0.02324009314,
            "directWallMs": 0.02324462891
          },
          {
            "gpuMs": 0.01990901493,
            "wallMs": 0.01991391602,
            "directGpuMs": 0.02504923381,
            "directWallMs": 0.02505385742
          },
          {
            "gpuMs": 0.01975598373,
            "wallMs": 0.01976000977,
            "directGpuMs": 0.02328406274,
            "directWallMs": 0.02328818359
          },
          {
            "gpuMs": 0.01986385882,
            "wallMs": 0.01986826172,
            "directGpuMs": 0.02472393773,
            "directWallMs": 0.0247284668
          },
          {
            "gpuMs": 0.01975418814,
            "wallMs": 0.01975805664,
            "directGpuMs": 0.02318171784,
            "directWallMs": 0.02318588867
          },
          {
            "gpuMs": 0.01994650066,
            "wallMs": 0.01995048828,
            "directGpuMs": 0.02335803211,
            "directWallMs": 0.0233628418
          }
        ]
      },
      "ratio": 1.2437214526777125
    },
    {
      "key": "large-transpose",
      "id": "transpose",
      "scalars": {
        "width": 2048,
        "height": 2048
      },
      "webgpu": {
        "key": "large-transpose",
        "id": "transpose",
        "scalars": {
          "width": 2048,
          "height": 2048
        },
        "block": [
          32,
          8,
          1
        ],
        "groups": [
          64,
          64
        ],
        "iterations": 2048,
        "samples": 9,
        "warmup": 5,
        "checks": {
          "output": {
            "pass": true,
            "failures": 0,
            "first": null,
            "maxAbs": 0,
            "maxRel": 0
          }
        },
        "gpuMedianMs": 0.013408,
        "wallMedianMs": 0.01504150390974246,
        "timings": [
          {
            "gpuMs": 0.013312,
            "wallMs": 0.014355468752910383
          },
          {
            "gpuMs": 0.013408,
            "wallMs": 0.01469970702601131
          },
          {
            "gpuMs": 0.013344,
            "wallMs": 0.01518798827601131
          },
          {
            "gpuMs": 0.013504,
            "wallMs": 0.015036621101899073
          },
          {
            "gpuMs": 0.013568,
            "wallMs": 0.01569580078648869
          },
          {
            "gpuMs": 0.013536,
            "wallMs": 0.015849609379074536
          },
          {
            "gpuMs": 0.013408,
            "wallMs": 0.015534667967585847
          },
          {
            "gpuMs": 0.0136,
            "wallMs": 0.014619140623835847
          },
          {
            "gpuMs": 0.013344,
            "wallMs": 0.01504150390974246
          }
        ]
      },
      "cuda": {
        "key": "large-transpose",
        "verified": true,
        "iterations": 2048,
        "samples": 9,
        "warmup": 5,
        "gpuMedianMs": 0.009157343768,
        "wallMedianMs": 0.009173876953,
        "directGpuMedianMs": 0.01115904655,
        "directWallMedianMs": 0.01116298828,
        "timings": [
          {
            "gpuMs": 0.009198062122,
            "wallMs": 0.009203076172,
            "directGpuMs": 0.01090398431,
            "directWallMs": 0.0109090332
          },
          {
            "gpuMs": 0.0090964064,
            "wallMs": 0.009101708984,
            "directGpuMs": 0.01072853152,
            "directWallMs": 0.01073330078
          },
          {
            "gpuMs": 0.009157343768,
            "wallMs": 0.009162255859,
            "directGpuMs": 0.01061890647,
            "directWallMs": 0.01067446289
          },
          {
            "gpuMs": 0.009200000204,
            "wallMs": 0.009204785156,
            "directGpuMs": 0.01283151563,
            "directWallMs": 0.01283637695
          },
          {
            "gpuMs": 0.009120780975,
            "wallMs": 0.009124755859,
            "directGpuMs": 0.01229884382,
            "directWallMs": 0.01230317383
          },
          {
            "gpuMs": 0.00917406287,
            "wallMs": 0.00917734375,
            "directGpuMs": 0.0132477656,
            "directWallMs": 0.01325498047
          },
          {
            "gpuMs": 0.009143234231,
            "wallMs": 0.009185498047,
            "directGpuMs": 0.01063535921,
            "directWallMs": 0.01069428711
          },
          {
            "gpuMs": 0.00917006284,
            "wallMs": 0.009173876953,
            "directGpuMs": 0.01115904655,
            "directWallMs": 0.01116298828
          },
          {
            "gpuMs": 0.009065062739,
            "wallMs": 0.009069189453,
            "directGpuMs": 0.01262364071,
            "directWallMs": 0.01262792969
          }
        ]
      },
      "ratio": 1.46418004387405
    },
    {
      "key": "large-particles",
      "id": "particles",
      "scalars": {
        "n": 262144,
        "dt": 0.01600000075995922,
        "time": 2.5,
        "attraction": 1
      },
      "webgpu": {
        "key": "large-particles",
        "id": "particles",
        "scalars": {
          "n": 262144,
          "dt": 0.01600000075995922,
          "time": 2.5,
          "attraction": 1
        },
        "block": [
          128,
          1,
          1
        ],
        "groups": [
          2048
        ],
        "iterations": 2048,
        "samples": 9,
        "warmup": 5,
        "checks": {
          "position": {
            "pass": true,
            "failures": 0,
            "first": null,
            "maxAbs": 1.1920928955078125e-7,
            "maxRel": 0.00007735700506432476
          },
          "velocity": {
            "pass": true,
            "failures": 0,
            "first": null,
            "maxAbs": 1.1920928955078125e-7,
            "maxRel": 0.000103556390861479
          }
        },
        "gpuMedianMs": 0.004512,
        "wallMedianMs": 0.006220703115104698,
        "timings": [
          {
            "gpuMs": 0.004512,
            "wallMs": 0.006369628899847157
          },
          {
            "gpuMs": 0.004512,
            "wallMs": 0.006633300785324536
          },
          {
            "gpuMs": 0.004512,
            "wallMs": 0.006232910149265081
          },
          {
            "gpuMs": 0.004704,
            "wallMs": 0.005700683599570766
          },
          {
            "gpuMs": 0.004736,
            "wallMs": 0.005827636719914153
          },
          {
            "gpuMs": 0.00464,
            "wallMs": 0.006188964849570766
          },
          {
            "gpuMs": 0.004512,
            "wallMs": 0.00630371093575377
          },
          {
            "gpuMs": 0.004512,
            "wallMs": 0.006220703115104698
          },
          {
            "gpuMs": 0.004672,
            "wallMs": 0.005908203122089617
          }
        ]
      },
      "cuda": {
        "key": "large-particles",
        "verified": true,
        "iterations": 2048,
        "samples": 9,
        "warmup": 5,
        "gpuMedianMs": 0.004932296928,
        "wallMedianMs": 0.004936328125,
        "directGpuMedianMs": 0.007679250091,
        "directWallMedianMs": 0.007684863281,
        "timings": [
          {
            "gpuMs": 0.005028468557,
            "wallMs": 0.005033007813,
            "directGpuMs": 0.007743296679,
            "directWallMs": 0.007747216797
          },
          {
            "gpuMs": 0.004994640592,
            "wallMs": 0.004998242187,
            "directGpuMs": 0.007768999785,
            "directWallMs": 0.007772900391
          },
          {
            "gpuMs": 0.0047246092,
            "wallMs": 0.004728173828,
            "directGpuMs": 0.007681765594,
            "directWallMs": 0.007684863281
          },
          {
            "gpuMs": 0.004932296928,
            "wallMs": 0.004936328125,
            "directGpuMs": 0.007632109337,
            "directWallMs": 0.007636328125
          },
          {
            "gpuMs": 0.004910390824,
            "wallMs": 0.004914453125,
            "directGpuMs": 0.00764790643,
            "directWallMs": 0.0077046875
          },
          {
            "gpuMs": 0.004726406187,
            "wallMs": 0.004729638672,
            "directGpuMs": 0.007745906245,
            "directWallMs": 0.007749169922
          },
          {
            "gpuMs": 0.004973093979,
            "wallMs": 0.004977099609,
            "directGpuMs": 0.007679250091,
            "directWallMs": 0.007683056641
          },
          {
            "gpuMs": 0.004725125153,
            "wallMs": 0.004727880859,
            "directGpuMs": 0.00763392169,
            "directWallMs": 0.007637109375
          },
          {
            "gpuMs": 0.004959781189,
            "wallMs": 0.00496328125,
            "directGpuMs": 0.007568031084,
            "directWallMs": 0.007572021484
          }
        ]
      },
      "ratio": 0.9147867749781995
    }
  ]
}