MLPerf Inference v5.1 results — RedHat/IBM Cloud 1xL40S vLLM llama3.1-8b submission (raw logs)

Fields

id

sources/bench-mlperf-v5-1-redhat-ibm-l40s-vllm-llama3-8b

URL

https://github.com/mlcommons/inference_results_v5.1/tree/main/closed/RedHat/results/IBM_Cloud_gx3-48x240x2l40s_vLLM

Publisher

MLCommons / RedHat

Kind

benchmark

Accessed

2026-10-14

notes

MLPerf Inference v5.1 closed submission. IBM Cloud gx3-48x240x2l40s instance, ONE NVIDIA L40S 48GB used (system has 2, only 1 in test), vLLM 0.10.0, RHEL 9.6, CUDA 12.9. llama3.1-8b Offline + Server. Single-GPU vLLM datapoint.

Notes

MLPerf Inference v5.1 closed submission. IBM Cloud gx3-48x240x2l40s instance, ONE NVIDIA L40S 48GB used (system has 2, only 1 in test), vLLM 0.10.0, RHEL 9.6, CUDA 12.9. llama3.1-8b Offline + Server. Single-GPU vLLM datapoint.

Cross-references

This record references no other records.