top of page
GPU Performance conversion Standard
-
To ensure fairness in inference time evluation across different GPUs, we apply a conversion process using the FLOPs reference table provided below.
-
The submitted inference time is adjusted based on the FLOPs of the GPU used and then converted to an equivalent inference time based on the NVIDIA GeForce RTX 4090 standard
-
For GPUs not listed in the table below, please reach out to us via the contact us.
-
Reference date: March 31, 2025
GPU | FLOPS | GPU ID |
|---|---|---|
NVIDIA Tesla T10 | 14.38 | 201 |
NVIDIA Tesla T40 24 GB | 14.38 | 202 |
NVIDIA RTX A4500 Embedded | 14.31 | 203 |
NVIDIA RTX A4500 Max-Q | 14.31 | 204 |
NVIDIA RTX A4000 Max-Q | 14.28 | 205 |
NVIDIA GeForce RTX 3060 3840SP | 14.22 | 206 |
NVIDIA Tesla PG500 216 | 14.13 | 207 |
NVIDIA Tesla V100 PCIe 32 GB | 14.13 | 208 |
NVIDIA Tesla V100 PCIe 16 GB | 14.13 | 209 |
AMD Radeon Pro Vega II | 14.09 | 210 |
Page 1 of 1
bottom of page