About tensorflow error : failed call to cuInit: CUDA_ERROR_NO_DEVICE

Autonomous Machines Jetson & Embedded Systems Jetson Nano
1

My Env :

Jetson Nano 4GB.
R32 (release), REVISION: 4.4, GCID: 23942405, BOARD: t210ref, EABI: aarch64

my tensorflow version is : 1.15.4.

When my tensorflow program is executed, the GPU rate is 0% (see by jtop). The log is as follows:

I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcuda.so.1
E tensorflow/stream_executor/cuda/cuda_driver.cc:339] failed call to cuInit: CUDA_ERROR_NO_DEVICE: no CUDA-capable device is detected
I tensorflow/stream_executor/cuda/cuda_diagnostics.cc:156] kernel driver does not appear to be running on this host (hre499-nano1): /proc/driver/nvidia/version does not exist
W tensorflow/core/platform/profile_utils/cpu_utils.cc:98] Failed to find bogomips in /proc/cpuinfo; cannot determine CPU frequency
I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x368e68f0 initialized for platform Host (this does not guarantee that XLA will be used). Devices:
I tensorflow/compiler/xla/service/service.cc:176]   StreamExecutor device (0): Host, Default Version

I’ve tried to add code :

os.environ['CUDA_VISIBLE_DEVICES'] ="0"

But, still error : failed call to cuInit: CUDA_ERROR_NO_DEVICE: no CUDA-capable device is detected

How can I fixed it ?
Thanks a lot!

3

Hi,

Which TensorFlow package do you install?
Please noted that you would need to install a package built with the same CUDA environment to ensure compatibility.

For R32.4.4, you can try the below command:

$ sudo pip3 install --pre --extra-index-url https://developer.download.nvidia.com/compute/redist/jp/v44 tensorflow==1.15.4+nv20.12

Thanks.