List your company

ResearchFormerly WizWand

A100

Benchmarks

TaskDatasetSOTA resultTrendResults

GPU Inference Throughput

A100-SXM4-40GB

5,081Throughput (tok/s)

8

A100 80GB (inference)

128Maximum Batch Size

6

Inference Efficiency

A100 80 GB GPU

0.019Latency (s)

5

Model Discovery

A100

5vit_tiny Discovery Rate

4