{
  "evidenceVersion": "local-ollama-speed-v1",
  "capturedAt": "2026-06-01T21:36:54.543Z",
  "hardwareSlug": "rtx-3080-16gb-mobile",
  "modelTag": "llama3.2:1b",
  "catalogSlug": "llama-3.2-1b-instruct",
  "quant": "Q4_K_M",
  "contextSize": 4096,
  "numPredict": 256,
  "prompt": "Write a detailed explanation of how transformer attention works.",
  "scenario": "single-stream",
  "runtime": "ollama-api",
  "command": "POST http://localhost:11434/api/generate model=llama3.2:1b temperature=0 top_p=1 seed=42 num_ctx=4096 num_predict=256",
  "env": {
    "os": "Microsoft Windows [Version 10.0.26200.8457]",
    "platform": "win32",
    "arch": "x64",
    "cpu": "AMD Ryzen 9 5900HX with Radeon Graphics        ",
    "cpuThreads": 16,
    "ramGb": 31.9,
    "gpu": {
      "vendor": "nvidia",
      "name": "NVIDIA GeForce RTX 3080 Laptop GPU",
      "driver": "571.96",
      "memoryTotalMb": 16384,
      "memoryUsedMb": 3001,
      "tempC": 68,
      "smClockMhz": 735,
      "powerDrawW": 43.4
    },
    "ollamaVersionCli": "ollama version is 0.24.0",
    "ollamaApiVersion": {
      "version": "0.24.0"
    }
  },
  "ollamaTagDetails": {
    "name": "llama3.2:1b",
    "model": "llama3.2:1b",
    "modified_at": "2026-06-02T00:30:44.0368378+03:00",
    "size": 1321098329,
    "digest": "baf6a787fdffd633537aa2eb51cfd54cb93ff08e28040095462bb63daf552878",
    "details": {
      "parent_model": "",
      "format": "gguf",
      "family": "llama",
      "families": [
        "llama"
      ],
      "parameter_size": "1.2B",
      "quantization_level": "Q8_0"
    }
  },
  "notes": null,
  "logHashSha256": "db12808c8c845efabad6d8288b60eb5c22ba67c9f91bb2b1ff621cbccd686683",
  "runsCaptured": 5,
  "decodeStatsAllRuns": {
    "min": 186.34253353492298,
    "max": 191.07600421262876,
    "median": 189.84298501990534,
    "p5": 186.92283650505107,
    "p95": 190.84128614951473,
    "mean": 189.28159701001582,
    "n": 5
  },
  "decodeStatsSteadyRuns": {
    "min": 186.34253353492298,
    "max": 191.07600421262876,
    "median": 189.5435167027344,
    "p5": 186.77776076251905,
    "p95": 190.89105133372024,
    "mean": 189.12639278825512,
    "n": 4
  },
  "runs": [
    {
      "runIndex": 1,
      "startedAt": "2026-06-01T21:36:41.100Z",
      "finishedAt": "2026-06-01T21:36:45.810Z",
      "rawPath": "public/benchmarks/evidence/local-ollama-rtx-3080-16gb-mobile-2026-06-01/llama3-2-1b/run1.ollama.json",
      "logPath": "public/benchmarks/evidence/local-ollama-rtx-3080-16gb-mobile-2026-06-01/llama3-2-1b/run1.log",
      "decodeTokPerSec": 189.90241389705864,
      "prefillTokPerSec": 3021.2262725836663,
      "totalMs": 4615.7058,
      "loadMs": 2957.1239,
      "promptEvalMs": 11.5847,
      "evalMs": 1348.0608,
      "promptTokens": 35,
      "responseTokens": 256,
      "responseHashSha256": "7bd5a08f00c007b5c44b8687fc583da1c5a9cdf08ab55dd15e985ebf63156848",
      "beforeGpu": {
        "vendor": "nvidia",
        "name": "NVIDIA GeForce RTX 3080 Laptop GPU",
        "driver": "571.96",
        "memoryTotalMb": 16384,
        "memoryUsedMb": 3001,
        "tempC": 68,
        "smClockMhz": 735,
        "powerDrawW": 39.94
      },
      "afterGpu": {
        "vendor": "nvidia",
        "name": "NVIDIA GeForce RTX 3080 Laptop GPU",
        "driver": "571.96",
        "memoryTotalMb": 16384,
        "memoryUsedMb": 4862,
        "tempC": 73,
        "smClockMhz": 1695,
        "powerDrawW": 113.69
      }
    },
    {
      "runIndex": 2,
      "startedAt": "2026-06-01T21:36:45.883Z",
      "finishedAt": "2026-06-01T21:36:47.997Z",
      "rawPath": "public/benchmarks/evidence/local-ollama-rtx-3080-16gb-mobile-2026-06-01/llama3-2-1b/run2.ollama.json",
      "logPath": "public/benchmarks/evidence/local-ollama-rtx-3080-16gb-mobile-2026-06-01/llama3-2-1b/run2.log",
      "decodeTokPerSec": 191.07600421262876,
      "prefillTokPerSec": 4477.363728236814,
      "totalMs": 2024.2819,
      "loadMs": 356.4146,
      "promptEvalMs": 7.8171,
      "evalMs": 1339.781,
      "promptTokens": 35,
      "responseTokens": 256,
      "responseHashSha256": "ea124597afb6c13e2dc13859bd5c90e29dd006373bf9604e8ef0f30b279f89d0",
      "beforeGpu": {
        "vendor": "nvidia",
        "name": "NVIDIA GeForce RTX 3080 Laptop GPU",
        "driver": "571.96",
        "memoryTotalMb": 16384,
        "memoryUsedMb": 4858,
        "tempC": 73,
        "smClockMhz": 1695,
        "powerDrawW": 107.6
      },
      "afterGpu": {
        "vendor": "nvidia",
        "name": "NVIDIA GeForce RTX 3080 Laptop GPU",
        "driver": "571.96",
        "memoryTotalMb": 16384,
        "memoryUsedMb": 4862,
        "tempC": 73,
        "smClockMhz": 1725,
        "powerDrawW": 113.25
      }
    },
    {
      "runIndex": 3,
      "startedAt": "2026-06-01T21:36:48.069Z",
      "finishedAt": "2026-06-01T21:36:50.180Z",
      "rawPath": "public/benchmarks/evidence/local-ollama-rtx-3080-16gb-mobile-2026-06-01/llama3-2-1b/run3.ollama.json",
      "logPath": "public/benchmarks/evidence/local-ollama-rtx-3080-16gb-mobile-2026-06-01/llama3-2-1b/run3.log",
      "decodeTokPerSec": 186.34253353492298,
      "prefillTokPerSec": 5320.92796983794,
      "totalMs": 2022.0854,
      "loadMs": 347.9288,
      "promptEvalMs": 6.5778,
      "evalMs": 1373.8141,
      "promptTokens": 35,
      "responseTokens": 256,
      "responseHashSha256": "ea124597afb6c13e2dc13859bd5c90e29dd006373bf9604e8ef0f30b279f89d0",
      "beforeGpu": {
        "vendor": "nvidia",
        "name": "NVIDIA GeForce RTX 3080 Laptop GPU",
        "driver": "571.96",
        "memoryTotalMb": 16384,
        "memoryUsedMb": 4862,
        "tempC": 73,
        "smClockMhz": 1725,
        "powerDrawW": 107.11
      },
      "afterGpu": {
        "vendor": "nvidia",
        "name": "NVIDIA GeForce RTX 3080 Laptop GPU",
        "driver": "571.96",
        "memoryTotalMb": 16384,
        "memoryUsedMb": 4862,
        "tempC": 74,
        "smClockMhz": 1710,
        "powerDrawW": 112.4
      }
    },
    {
      "runIndex": 4,
      "startedAt": "2026-06-01T21:36:50.255Z",
      "finishedAt": "2026-06-01T21:36:52.353Z",
      "rawPath": "public/benchmarks/evidence/local-ollama-rtx-3080-16gb-mobile-2026-06-01/llama3-2-1b/run4.ollama.json",
      "logPath": "public/benchmarks/evidence/local-ollama-rtx-3080-16gb-mobile-2026-06-01/llama3-2-1b/run4.log",
      "decodeTokPerSec": 189.84298501990534,
      "prefillTokPerSec": 4544.274214489743,
      "totalMs": 2009.8121,
      "loadMs": 345.4458,
      "promptEvalMs": 7.702,
      "evalMs": 1348.4828,
      "promptTokens": 35,
      "responseTokens": 256,
      "responseHashSha256": "ea124597afb6c13e2dc13859bd5c90e29dd006373bf9604e8ef0f30b279f89d0",
      "beforeGpu": {
        "vendor": "nvidia",
        "name": "NVIDIA GeForce RTX 3080 Laptop GPU",
        "driver": "571.96",
        "memoryTotalMb": 16384,
        "memoryUsedMb": 4862,
        "tempC": 72,
        "smClockMhz": 1710,
        "powerDrawW": 106.39
      },
      "afterGpu": {
        "vendor": "nvidia",
        "name": "NVIDIA GeForce RTX 3080 Laptop GPU",
        "driver": "571.96",
        "memoryTotalMb": 16384,
        "memoryUsedMb": 4862,
        "tempC": 74,
        "smClockMhz": 1665,
        "powerDrawW": 114.84
      }
    },
    {
      "runIndex": 5,
      "startedAt": "2026-06-01T21:36:52.444Z",
      "finishedAt": "2026-06-01T21:36:54.542Z",
      "rawPath": "public/benchmarks/evidence/local-ollama-rtx-3080-16gb-mobile-2026-06-01/llama3-2-1b/run5.ollama.json",
      "logPath": "public/benchmarks/evidence/local-ollama-rtx-3080-16gb-mobile-2026-06-01/llama3-2-1b/run5.log",
      "decodeTokPerSec": 189.24404838556345,
      "prefillTokPerSec": 4953.858347959011,
      "totalMs": 2018.5799,
      "loadMs": 357.6733,
      "promptEvalMs": 7.0652,
      "evalMs": 1352.7506,
      "promptTokens": 35,
      "responseTokens": 256,
      "responseHashSha256": "ea124597afb6c13e2dc13859bd5c90e29dd006373bf9604e8ef0f30b279f89d0",
      "beforeGpu": {
        "vendor": "nvidia",
        "name": "NVIDIA GeForce RTX 3080 Laptop GPU",
        "driver": "571.96",
        "memoryTotalMb": 16384,
        "memoryUsedMb": 4862,
        "tempC": 74,
        "smClockMhz": 1755,
        "powerDrawW": 110.66
      },
      "afterGpu": {
        "vendor": "nvidia",
        "name": "NVIDIA GeForce RTX 3080 Laptop GPU",
        "driver": "571.96",
        "memoryTotalMb": 16384,
        "memoryUsedMb": 4862,
        "tempC": 75,
        "smClockMhz": 1710,
        "powerDrawW": 114.86
      }
    }
  ],
  "errors": [],
  "uploadEligible": true
}