Nvidia Tesla - AshokBhat/ml GitHub Wiki
About
- Products targeted at stream processing or GPGPU
- Programmable using the CUDA or OpenCL APIs.
Products
| Product |
Arch |
CUDA Cores |
Tensor Cores |
FP32 TOPS |
Power |
Segment |
| Tesla T4 |
Turing |
2560 |
320 |
8.1 TFLOPS |
70 W |
Inference |
| Tesla A100 for PCIe |
Ampere |
? |
|
? TFLOPS |
? W |
Training, Inference |
| Tesla V100 for PCIe |
Volta |
5120 |
|
14 TFLOPS |
250 W |
? |
| Tesla P100 for PCIe |
Pascal |
3584 |
|
9.3 TFLOPS |
250 W |
? |
| Tesla K80 for PCIe |
Kepler |
4992 |
? |
8.3 TFLOPS |
300 W |
? |
| Tesla M60 for PCIe |
Maxwell |
2048 |
? |
? |
300 W |
? |