Closed Krikonildo closed 6 months ago
Hi, please note the line: CUDA backend: OFF
which means that you did not build with CUDA support. Have a look at this to solve the problem.
build.sh
cmake -DCMAKE_BUILD_TYPE=$BUILD_TYPE -DCMAKE_CUDA_COMPILER=/usr/bin/nvcc -DCMAKE_CUDA_COMPILER_TOOLKIT_ROOT=/usr/lib/cuda ..
.bashrc
export LD_LIBRARY_PATH=/usr/lib/cuda/lib64\
${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}
nvcc -V
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2017 NVIDIA Corporation
Built on Fri_Nov__3_21:07:56_CDT_2017
Cuda compilation tools, release 9.1, V9.1.85
./gpufetch -v
[ERROR]: cudaErrorUnknown: unknown error
[WARNING]: Unable to find a valid device for vendor id 0x8086 using pciutils
[ERROR]: No GPU was detected! Available GPUs are:
- GPU 0:
* Vendor: NVIDIA
* PCI id: 10de:128b
Please, make sure that the appropiate backend is enabled:
- CUDA backend: ON
- Intel backend: ON
Visit https://github.com/Dr-Noob/gpufetch#2-backends for more information
I'm afraid the error cudaErrorUnknown: unknown error
is due to something unrelated to gpufetch. It is likely that the problem is in your CUDA or drivers installation.