CUDA on WSL2 Could not open file to read NUMA

I installed CUDA on WSL2 Ubuntu distribution. However, I was not sure if I also needed to install cudnn on WSL2. When I run Visual Studio Code from an Ubuntu distro in WSL2 and open a python tensorflow2 jupyter notebook and execute “device_name = tf.test.gpu_device_name()” I getwhat is shown below. Why is it using OneDNN instead of CuDNN and why do I get “could not open file to read NUMA node”?

2021-12-13 20:04:21.536059: I tensorflow/core/platform/cpu_feature_guard.cc:151] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations: AVX2 AVX512F FMA
To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
2021-12-13 20:04:21.571180: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:925] could not open file to read NUMA node: /sys/bus/pci/devices/0000:65:00.0/numa_node
Your kernel may have been built without NUMA support.
2021-12-13 20:04:21.604827: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:925] could not open file to read NUMA node: /sys/bus/pci/devices/0000:65:00.0/numa_node
Your kernel may have been built without NUMA support.
2021-12-13 20:04:21.605112: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:925] could not open file to read NUMA node: /sys/bus/pci/devices/0000:65:00.0/numa_node
Your kernel may have been built without NUMA support.
2021-12-13 20:04:22.369084: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:925] could not open file to read NUMA node: /sys/bus/pci/devices/0000:65:00.0/numa_node
Your kernel may have been built without NUMA support.
2021-12-13 20:04:22.369591: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:925] could not open file to read NUMA node: /sys/bus/pci/devices/0000:65:00.0/numa_node
Your kernel may have been built without NUMA support.
2021-12-13 20:04:22.369609: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1609] Could not identify NUMA node of platform GPU id 0, defaulting to 0. Your kernel may not have been built with NUMA support.
2021-12-13 20:04:22.369964: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:925] could not open file to read NUMA node: /sys/bus/pci/devices/0000:65:00.0/numa_node
Your kernel may have been built without NUMA support.
2021-12-13 20:04:22.370233: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1525] Created device /device:GPU:0 with 21629 MB memory: → device: 0, name: NVIDIA GeForce RTX 3090, pci bus id: 0000:65:00.0, compute capability: 8.6

4 Likes

same issue here, any help would be appreciated

1 Like