MLPerf v6.1 Dell XE9785L 8xMI355X llama3.1-8b Server run_1 harness summary log

Fields

id

sources/bench-mlperf-v6-1-dell-mi355x-llama3-8b-server-log

URL

https://raw.githubusercontent.com/mlcommons/inference_results_v6.1/main/closed/Dell/results/XE9785L_8xMI355X/llama3.1-8b/Server/performance/run_1/mlperf_log_summary.txt

Publisher

MLCommons

Kind

benchmark

Accessed

2026-10-04

notes

Result is : VALID. Completed samples per second 1157.99, Completed tokens per second 148563.48. Performance constraints satisfied: Yes. The full percentile block is preserved in the log: mean latency 8.146 s, p50 8.308 s, p90 8.929 s, p95 9.095 s, p97 9.207 s, p99 9.428 s - i.e. the Server scenario holds a tight tail, which is the actual reason it scores below Offline.

Notes

Result is : VALID. Completed samples per second 1157.99, Completed tokens per second 148563.48. Performance constraints satisfied: Yes. The full percentile block is preserved in the log: mean latency 8.146 s, p50 8.308 s, p90 8.929 s, p95 9.095 s, p97 9.207 s, p99 9.428 s - i.e. the Server scenario holds a tight tail, which is the actual reason it scores below Offline.

Cross-references

This record references no other records.