Skip to content
KernelIndex
Search⌘K
Sign in

Cisco · Cisco UCS X210 M8 - X580P PCIe Node (4x H200-NVL-141GB, TensorRT)

Llama-3.1-8B (MLPerf reference) · llama3.1-8b · Server · run 01a00b49-bf67…
valid · Reported evidence
Measurements
output_token_throughput_tps · reported25,731.6 tokens_s

Llama-3.1-8B (MLPerf reference) · llama3.1-8b · Server · server · 4× NVIDIA H200-NVL-141GB · mlperf_closed_99 · observed 2026-03-26

Identity
modelLlama-3.1-8B (MLPerf reference)
stackTensorRT 10.11, CUDA 12.9
workloadllama3.1-8b · Server
topology4× NVIDIA H200-NVL-141GB
harnessmlperf_loadgen v6.0
quality policymlperf_closed_99
comparison keysha256:3d08393138e3f2e9…
sha256:264168da4e0486cb6481796…

Launch configuration

dtypefp8
notesH200-NVL TGP 600W
systemCisco UCS X210 M8 - X580P PCIe Node (4x H200-NVL-141GB, TensorRT)
softwareTensorRT 10.11, CUDA 12.9
availabilityavailable

Workload

namellama3.1-8b · Server
streamingyes
load generationopen_loop

Canonical manifest

Show manifest
{
  "kind": "ServingRun",
  "spec": {
    "status": "valid",
    "harness": {
      "name": "mlperf_loadgen",
      "version": "v6.0"
    },
    "topology": {
      "nodeCount": 1,
      "acceleratorModel": "NVIDIA H200-NVL-141GB",
      "acceleratorVendor": "nvidia",
      "acceleratorsPerNode": 4
    },
    "observedAt": "2026-03-26T19:39:54Z",
    "modelDigest": "sha256:dafdb8bb411a12b698d36856eed69f16242fb467999fb68680ca09a5601b6bd4",
    "measurements": [
      {
        "unit": "tokens_s",
        "value": 25731.6,
        "metric": "output_token_throughput_tps",
        "statistic": "reported"
      }
    ],
    "sourceNative": {
      "source": "mlperf-inference",
      "metrics": {
        "performance_units": "Tokens/s",
        "performance_result": 25731.6
      },
      "category": "datacenter",
      "division": "closed",
      "scenario": "server",
      "benchmark": "llama3.1-8b",
      "externalId": "v6.0/6.0-0013"
    },
    "qualityPolicy": "mlperf_closed_99",
    "workloadDigest": "sha256:209ab3ba41ef397171c081377f84bcb2512e3d47cd6334278a712a3858d8eedd",
    "configurationDigest": "sha256:7c9eb7a04d8523c53ffe19662694e63d9508d0a4cdbb11492d311612eaad9421"
  },
  "metadata": {
    "name": "mlperf-v6-0-6-0-0013",
    "title": "Cisco · llama3.1-8b · Server",
    "sourceRefs": [
      {
        "url": "https://github.com/mlcommons/submissions_inference_v6.0/tree/main/closed/Cisco/results/X210M8_X580P_H200NVL_141GBx4_TRT"
      }
    ]
  },
  "apiVersion": "kernelindex.dev/v1alpha1"
}
Cite this result (permalink, digest, access date)
Report an issue with this serving run

A maintainer reviews every report. If accepted, the record is retracted or superseded; its history stays visible.

MLPerf Inference · closed/datacenter · entry v6.0/6.0-0013 · shown unmodified as publishedAll serving results →JSON