Intel · 1-node-4x-BMG-B50
Llama-3.1-8B (MLPerf reference) · llama3.1-8b · Offline · run 01a00b4a-0cfc…
valid · Reported evidence
Measurements
output_token_throughput_tps · reported1,721.84 tokens_s
Llama-3.1-8B (MLPerf reference) · llama3.1-8b · Offline · offline · 4× Intel(R) Arc Pro(R) B50 · mlperf_closed_99 · observed 2026-03-26
Identity
modelLlama-3.1-8B (MLPerf reference)
stackPyTorch vLLM
workloadllama3.1-8b · Offline
topology4× Intel(R) Arc Pro(R) B50
harnessmlperf_loadgen v6.0
quality policymlperf_closed_99
comparison keysha256:1bc93489362c31c2…
sha256:4473cfc64ca6cbe7c98d77e…
Launch configuration
dtypeint8
notesN/A. N/A
system1-node-4x-BMG-B50
softwarePyTorch vLLM
availabilityavailable
Workload
namellama3.1-8b · Offline
streamingno
load generationoffline
Canonical manifest
Show manifestHide manifest
{
"kind": "ServingRun",
"spec": {
"status": "valid",
"harness": {
"name": "mlperf_loadgen",
"version": "v6.0"
},
"topology": {
"nodeCount": 1,
"acceleratorModel": "Intel(R) Arc Pro(R) B50",
"acceleratorsPerNode": 4
},
"observedAt": "2026-03-26T19:39:54Z",
"modelDigest": "sha256:dafdb8bb411a12b698d36856eed69f16242fb467999fb68680ca09a5601b6bd4",
"measurements": [
{
"unit": "tokens_s",
"value": 1721.84,
"metric": "output_token_throughput_tps",
"statistic": "reported"
}
],
"sourceNative": {
"source": "mlperf-inference",
"metrics": {
"performance_units": "Tokens/s",
"performance_result": 1721.84
},
"category": "datacenter",
"division": "closed",
"scenario": "offline",
"benchmark": "llama3.1-8b",
"externalId": "v6.0/6.0-0058"
},
"qualityPolicy": "mlperf_closed_99",
"workloadDigest": "sha256:a49a37592d011507287e85cda6488466913158a2cd8c7ecb16cf07f92327bf15",
"configurationDigest": "sha256:1318c235ff7ffeb6aa76f714fe869ff33d362164bc0849b24255f6fcf82eaa4a"
},
"metadata": {
"name": "mlperf-v6-0-6-0-0058",
"title": "Intel · llama3.1-8b · Offline",
"sourceRefs": [
{
"url": "https://github.com/mlcommons/submissions_inference_v6.0/tree/main/closed/Intel/results/1-node-4x-BMG-B50"
}
]
},
"apiVersion": "kernelindex.dev/v1alpha1"
}Cite this result (permalink, digest, access date)
Report an issue with this serving run
MLPerf Inference · closed/datacenter · entry v6.0/6.0-0058 · shown unmodified as publishedAll serving results →JSON