Cuda suddenly stopped working after some Windows update?


I have the following specs on my Windows laptop with an NVidia graphics card, RTX 3060:
Cuda version 11.2.2_461.33 with the Nvidia driver 11.2.109, cudnn version cudnn-11.2-windows-x64-v8.1.1.33 for Windows 10.
I am running tensorflow version 2.8.0 in Jupyter notebook with Python 3.9 through anaconda. I have GPU enabled successfully.

It has been working fine until this morning. When I tried to run my lstm model in keras-tensorflow, it only runs one epoch and stops, e.g.

44/44 - 4s - loss: 0.7971 - binary_accuracy: 0.5086 - val_loss: 0.7673 - val_binary_accuracy: 0.5550 - 4s/epoch - 102ms/step

I have observed the following changes.

I’m running my jupyter notebooks through launching a virtual environment in Anaconda Powershell - it used to open up in the Google Chrome browser and it now opens up in Microsoft Edge, so I had to change it back to Chrome.

I have checked the latest Windows Update history. It shows there were some updates in Windows Defender on May 23rd and May 24th.

With other models, it still refuses to train giving me a Cuda error:

UnknownError: Graph execution error:

Fail to find the dnn implementation.
	 [[{{node CudnnRNN}}]]
	 [[sequential/bidirectional/forward_gru/PartitionedCall]] [Op:__inference_train_function_5869]

Also, my system prompted me to update the bios two days ago.

Which one is likely to be the cause of my problem and what I need to do to fix it?

thank you.