MangoBoost · MangoBoost Mi300X (8x MI300X, LLMBoost)
Llama-2-70B (MLPerf reference) · llama2-70b-99 · Server · run 01a00b4a-1b1c…
valid · Reported evidence
Measurements
output_token_throughput_tps · reported25,463.1 tokens_s
Llama-2-70B (MLPerf reference) · llama2-70b-99 · Server · server · 8× AMD Instinct MI300X 192GB HBM3 · mlperf_closed_99_9 · observed 2026-03-26
Identity
modelLlama-2-70B (MLPerf reference)
stackROCm 6.10.5
workloadllama2-70b-99 · Server
topology8× AMD Instinct MI300X 192GB HBM3
harnessmlperf_loadgen v6.0
quality policymlperf_closed_99_9
comparison keysha256:89e7b1db30c59985…
sha256:0d2673b782f62b843bc3af1…
Launch configuration
dtypefp8
systemMangoBoost Mi300X (8x MI300X, LLMBoost)
softwareROCm 6.10.5
availabilityavailable
Workload
namellama2-70b-99 · Server
streamingyes
load generationopen_loop
Canonical manifest
Show manifestHide manifest
{
"kind": "ServingRun",
"spec": {
"status": "valid",
"harness": {
"name": "mlperf_loadgen",
"version": "v6.0"
},
"topology": {
"nodeCount": 1,
"acceleratorModel": "AMD Instinct MI300X 192GB HBM3",
"acceleratorVendor": "amd",
"acceleratorsPerNode": 8
},
"observedAt": "2026-03-26T19:39:54Z",
"modelDigest": "sha256:8005d6af1339caa02e60709ce2d7a6002ce9c51e436d21bb39dc5b8f93a24b0c",
"measurements": [
{
"unit": "tokens_s",
"value": 25463.1,
"metric": "output_token_throughput_tps",
"statistic": "reported"
}
],
"sourceNative": {
"source": "mlperf-inference",
"metrics": {
"performance_units": "Tokens/s",
"performance_result": 25463.1
},
"category": "datacenter",
"division": "closed",
"scenario": "server",
"benchmark": "llama2-70b-99.9",
"externalId": "v6.0/6.0-0069"
},
"qualityPolicy": "mlperf_closed_99_9",
"workloadDigest": "sha256:ee188c862b2e3bbbca29b194db4aa45d0f33b7da2bdc410aa953d53e6ad6a4ac",
"configurationDigest": "sha256:c4a90bc9ae643eaba34c8c5a1281b28f6acec2db5df028477d0e9ee43393d015"
},
"metadata": {
"name": "mlperf-v6-0-6-0-0069",
"title": "MangoBoost · llama2-70b-99.9 · Server",
"sourceRefs": [
{
"url": "https://github.com/mlcommons/submissions_inference_v6.0/tree/main/closed/MangoBoost/results/8xMI300X_2xEPYC_9534"
}
]
},
"apiVersion": "kernelindex.dev/v1alpha1"
}Cite this result (permalink, digest, access date)
Report an issue with this serving run
MLPerf Inference · closed/datacenter · entry v6.0/6.0-0069 · shown unmodified as publishedAll serving results →JSON