CuDNN error on Jetson Nano :: "reason: CUDNN_STATUS_NOT_SUPPORTED"

I am using Dlib 19.21 on Jetson Nano on a project. But running a program causes a strange CUDNN error wheres the program was running fine on previous L4T 32.4 version. My Jetson Nano software stacks are like this:

NVIDIA Jetson Nano
 L4T 32.5.0 [ JetPack UNKNOWN ]
   Ubuntu 18.04.5 LTS
   Kernel Version: 4.9.201-tegra
 CUDA 10.2.89
   CUDA Architecture: 5.3
 OpenCV version: 4.1.1
   OpenCV Cuda: NO
 CUDNN: 8.0.0.180
 TensorRT: 7.1.3.0
 Vision Works: 1.6.0.501
 VPI: ii libnvvpi1 1.0.12 arm64 NVIDIA Vision Programming Interface library

But its causing error like this:

terminate called after throwing an instance of 'dlib::cudnn_error'
  what():  Error while calling cudnnGetConvolutionBackwardFilterWorkspaceSize( context(), descriptor(data), descriptor(dest_desc), (const cudnnConvolutionDescriptor_t)conv_handle, (const cudnnFilterDescriptor_t)filter_handle, (cudnnConvolutionBwdFilterAlgo_t)backward_filters_algo, &backward_filters_workspace_size_in_bytes) in file /home/nvidia/dlib/dlib/cuda/cudnn_dlibapi.cpp:1046. code: 9, reason: CUDNN_STATUS_NOT_SUPPORTED

Your OpenCV version is compiled without CUDA support and release 4.1.1 is quite old, try a newer version, e.g. 4.4.
For compiling and installing OpenCV with CUDA support you can use follwing:

Yes, thanks for the suggestion. I have tried to install OCV 4.5 with CUDA but faced some error. Anyways, that shouldn’t affect the problem mentioned here, should it?

Hi,

Based on the log, the error might be caused by the large workspace size.
May I know which Nano do you use? 4GB or 2GB?

Since Nano has limited resource, it’s possible that the workspace is too large for cuDNN to allocate.
Could you update to a smaller size to see if it works?

Thanks.

It’s a 4GB version. No such error in previous L4T version on the same device.

There is no update from you for a period, assuming this is not an issue any more.
Hence we are closing this topic. If need further support, please open a new one.
Thanks

Hi,

We try to reproduce this issue in our environment.
Could you share the detailed steps and the working/non-working JetPack version?

Thanks.