Tesla v100 CUDA Support


I’m trying to utilize Tensorflow with CUDA for accelerated training.
GPU Model: Tesla V100 FHHL 16GB
I have installed CUDA and cuDNN using the run file and the copying the necessary libraries.

CUDA 10.1, cuDNN 7.6
I get this error whenever I try to run the tensorflow CNN model:
Attempting to fetch value instead of handling error Internal: failed to get device attribute 13 for device 0: CUDA_ERROR_UNKNOWN: unknown error.

NVIDIA-SMI give the following output:| NVIDIA-SMI 450.51.05 Driver Version: 450.51.05 CUDA Version: 11.0

nvcc-V the following output:
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2019 NVIDIA Corporation
Built on Fri_Feb__8_19:08:17_PST_2019
Cuda compilation tools, release 10.1, V10.1.105

Pls let tme know if this is a CUDA version mismatch or any other things need to be taken care of.

Hi @anandshubham1998 ,
Can you please validate the CUDA installation steps from here and let us know.


So Tesla V100 do they have limitations on the CUDA, cuDNN compatibility? Can I install the latest CUDA and cuDNN test configurations for tensorflow?

cuDNN8.1 CUDA11.2