Monte Carlo Integration on CUDA-LLM task suite
6.95Latency (ms)OptiML
Evaluation results
| Method | Links | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| OptiMLGPU=A100 80GB, Variant=G+X2026.02 | 6.95 | 1.03 | 60.24 | 13.21 | 23.61 | 488 | 1,369 | 2,974 | 84.22 | 58 | |
| HPC-Coder-V2 + OptiML-XGPU=A100 80GB2026.02 | 6.96 | 1.03 | 59.12 | 13.98 | 22.33 | 505 | 1,406 | 3,036 | 83.02 | 58 | |
| GPT 5.1 + OptiML-XGPU=A100 80GB2026.02 | 7.15 | — | 58.31 | 14.42 | 21.86 | 522 | 1,441 | 3,093 | 82.16 | 58 | |
| StarCoder2 + OptiML-XGPU=A100 80GB2026.02 | 7.28 | 0.98 | 57.41 | 14.55 | 21.58 | 529 | 1,468 | 3,122 | 81.27 | 58 | |
| Qwen2.5-Coder + OptiML-XGPU=A100 80GB2026.02 | 7.6 | 0.94 | 56.17 | 14.96 | 21.07 | 540 | 1,516 | 3,228 | 80.31 | 58 |