LLM Inference Throughput on 4K/4K Inference Scenario
2.44Speedupgpt-oss-puzzle-88B
Evaluation results
| Method | Links | ||
|---|---|---|---|
| gpt-oss-puzzle-88BHardware Configuration=single H1002026.02 | 2.44 | — | |
| gpt-oss-puzzle-88BHardware Configuration=8×H100 node2026.02 | 1.22 | — | |
| gpt-oss-puzzle-88BHardware=8x B200, Scenario=4K/4K, Note=Max throughput at best configuration2026.02 | 1.15 | — |