[Cuda cudnn version check] #cuda #cudnn #nvidia

check_cuda_cudnn.md

To check nvidia driver

modinfo nvidia

To check cuda version

cat /usr/local/cuda/version.txt

nvcc  --version

To check cudnn version

cat /usr/local/cuda/include/cudnn.h | grep CUDNN_MAJOR -A 2

cat /usr/include/cudnn.h | grep CUDNN_MAJOR -A 2

To check GPU Card info

nvidia-smi

Python (Show what version of tensorflow in your PC.)

• for Python 2

python -c 'import tensorflow as tf; print(tf.__version__)'  

• for Python 3

python3 -c 'import tensorflow as tf; print(tf.__version__)'  

gpu check

CUDA_DEVICE_ORDER=PCI_BUS_ID CUDA_VISIBLE_DEVICES=1 python import pytorch torch.cuda.current_device()

looks like they moved the header file in 11.0
cat /path/to/cuda/include/cudnn_version.h | grep CUDNN_MAJOR -A 2

#define CUDNN_MAJOR 8
#define CUDNN_MINOR 0
#define CUDNN_PATCHLEVEL 2
--
#define CUDNN_VERSION (CUDNN_MAJOR * 1000 + CUDNN_MINOR * 100 + CUDNN_PATCHLEVEL)

Thanks I was able to verify my cudnn installation

looks like they moved the header file in 11.0
cat /path/to/cuda/include/cudnn_version.h | grep CUDNN_MAJOR -A 2

#define CUDNN_MAJOR 8
#define CUDNN_MINOR 0
#define CUDNN_PATCHLEVEL 2

#define CUDNN_VERSION (CUDNN_MAJOR * 1000 + CUDNN_MINOR * 100 + CUDNN_PATCHLEVEL)

why is your cudnn.h has head file?

dunno, ask nvidia

Thanks @bionicles Works perfectly. @Shijie97 the location of cudnn can be found using $whereis cudnn

Cuda

cat /usr/local/cuda/version.json

CUDNN

cat /usr/include/cudnn_version.h | grep CUDNN_MAJOR -A 2

Thanks @lgg @bionicles