Last Updated:
Benchmark Dashboard
Nightly performance benchmarks — throughput, latency & ITL by model and batch size
Fetching benchmark data from nightly perf runs...
No data for throughput vs latency chart.
No trend data. Try a larger date range.
No overlapping models between AMD and NVIDIA for comparison.
Raw Performance Data
| Date | Model | Variant | Hardware | Deployment | GPUs | Batch / Conc | Input Len | Latency (s) | Input Tput | Output Tput | Total Tput/GPU | ITL (ms) | Link |
|---|