A100
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| GPU Inference Throughput | A100-SXM4-40GB | Throughput (tok/s)5,081 | 8 | |
| LLM Generation | A100 80GB (inference) | Maximum Batch Size128 | 6 | |
| Inference Efficiency | A100 80 GB GPU | Latency (s)0.019 | 5 | |
| Model Discovery | A100 | vit_tiny Discovery Rate5 | 4 |
Showing 4 of 4 rows