nvcc fatal : Value 'sm_20' is not defined for option 'gpu-architecture'
I've looked at many pages and either could not follow what they were saying because they were unclear and/or my knowledge is just not sufficient enough.
I am trying to run:
luarocks install https://raw.githubusercontent.com/qassemoquab/stnbhwd/master/stnbhwd-scm-1.rockspec
So that I may run DenseCap over some images using GPU Acceleration. When I run it, I get this error:
$ luarocks install https://raw.githubusercontent.com/qassemoquab/stnbhwd/master/stnbhwd-scm-1.rockspec
Using https://raw.githubusercontent.com/qassemoquab/stnbhwd/master/stnbhwd-scm-1.rockspec... switching to 'build' mode
Cloning into 'stnbhwd'...
remote: Counting objects: 24, done.
remote: Compressing objects: 100% (23/23), done.
remote: Total 24 (delta 0), reused 14 (delta 0), pack-reused 0
Receiving objects: 100% (24/24), 19.42 KiB | 0 bytes/s, done.
Checking connectivity... done.
cmake -E make_directory build && cd build && cmake .. -DCMAKE_BUILD_TYPE=Release -DCMAKE_PREFIX_PATH="/home/tex/torch/install/bin/.." -DCMAKE_INSTALL_PREFIX="/home/tex/torch/install/lib/luarocks/rocks/stnbhwd/scm-1" && make
-- The C compiler identification is GNU 5.4.0
-- The CXX compiler identification is GNU 5.4.0
-- Check for working C compiler: /usr/bin/cc
-- Check for working C compiler: /usr/bin/cc -- works
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Detecting C compile features
-- Detecting C compile features - done
-- Check for working CXX compiler: /usr/bin/c++
-- Check for working CXX compiler: /usr/bin/c++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Found Torch7 in /home/tex/torch/install
-- Try OpenMP C flag = [-fopenmp]
-- Performing Test OpenMP_FLAG_DETECTED
-- Performing Test OpenMP_FLAG_DETECTED - Success
-- Try OpenMP CXX flag = [-fopenmp]
-- Performing Test OpenMP_FLAG_DETECTED
-- Performing Test OpenMP_FLAG_DETECTED - Success
-- Found OpenMP: -fopenmp
-- Compiling with OpenMP support
-- Looking for pthread.h
-- Looking for pthread.h - found
-- Looking for pthread_create
-- Looking for pthread_create - not found
-- Looking for pthread_create in pthreads
-- Looking for pthread_create in pthreads - not found
-- Looking for pthread_create in pthread
-- Looking for pthread_create in pthread - found
-- Found Threads: TRUE
-- Found CUDA: /usr/local/cuda (found suitable version "9.0", minimum required is "5.5")
-- Configuring done
-- Generating done
-- Build files have been written to: /tmp/luarocks_stnbhwd-scm-1-4197/stnbhwd/build
Scanning dependencies of target stn
[ 25%] Building C object CMakeFiles/stn.dir/init.c.o
[ 50%] Linking C shared module libstn.so
[ 50%] Built target stn
[ 75%] Building NVCC (Device) object CMakeFiles/custn.dir/custn_generated_init.cu.o
nvcc fatal : Value 'sm_20' is not defined for option 'gpu-architecture'
CMake Error at custn_generated_init.cu.o.cmake:207 (message):
Error generating
/tmp/luarocks_stnbhwd-scm-1-4197/stnbhwd/build/CMakeFiles/custn.dir//./custn_generated_init.cu.o
CMakeFiles/custn.dir/build.make:63: recipe for target 'CMakeFiles/custn.dir/custn_generated_init.cu.o' failed
make[2]: *** [CMakeFiles/custn.dir/custn_generated_init.cu.o] Error 1
CMakeFiles/Makefile2:104: recipe for target 'CMakeFiles/custn.dir/all' failed
make[1]: *** [CMakeFiles/custn.dir/all] Error 2
Makefile:127: recipe for target 'all' failed
make: *** [all] Error 2
Error: Build error: Failed building.
I was able to luarocks install cutorch, luarocks install cunn, and luarocks install cudnn just fine.
I am running Ubuntu 16.04 on a GTX 1080ti.
$ nvidia-smi
Tue Dec 5 16:25:42 2017
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 384.90 Driver Version: 384.90 |
|-------------------------------+----------------------+----------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
|===============================+======================+======================|
| 0 GeForce GTX 108... Off | 00000000:29:00.0 On | N/A |
| 0% 47C P8 16W / 250W | 716MiB / 11169MiB | 1% Default |
+-------------------------------+----------------------+----------------------+
+-----------------------------------------------------------------------------+
| Processes: GPU Memory |
| GPU PID Type Process name Usage |
|=============================================================================|
| 0 1128 G /usr/lib/xorg/Xorg 479MiB |
| 0 1782 G compiz 234MiB |
+-----------------------------------------------------------------------------+
For some reason, when I run $ nvcc -V, I get:
The program 'nvcc' is currently not installed. You can install it by typing:
sudo apt install nvidia-cuda-toolkit
..and this is another big problem I am facing. When I install nvcc, it installs the toolkit for Cuda 7.5, but I have Cuda 9.0. I installed it using the .deb file that comes from Cuda's website.
$ sudo apt-get install cuda
Reading package lists... Done
Building dependency tree
Reading state information... Done
cuda is already the newest version (9.0.176-1).
The following packages were automatically installed and are no longer required:
libcublas7.5 libcudart7.5 libcufft7.5 libcufftw7.5 libcuinj64-7.5
libcurand7.5 libcusolver7.5 libcusparse7.5 libnppc7.5 libnppi7.5 libnpps7.5
libnvblas7.5 libnvrtc7.5 libnvtoolsext1 libnvvm3 libthrust-dev libvdpau-dev
nvidia-cuda-dev nvidia-cuda-doc nvidia-cuda-gdb nvidia-opencl-dev
nvidia-profiler nvidia-visual-profiler opencl-headers
Use 'sudo apt autoremove' to remove them.
0 upgraded, 0 newly installed, 0 to remove and 222 not upgraded.
I honestly expect that I need to install nvcc, but like I said, it installs for the wrong version, and I cannot for the life of me figure out how to install it for the right version, and Im really confused as to why I am getting this error now even though I already installed cutorch, cudnn, and cunn just fine. Any help is appreciated...
Thanks
Answer
Try to change the code architecture (such as sm_20) to some higher version in CMakeLists.txt of stnbhwd that you are trying to install.
From:
IF (CUDA_FOUND)
LIST(APPEND CUDA_NVCC_FLAGS "-arch=sm_20")
to:
IF (CUDA_FOUND)
LIST(APPEND CUDA_NVCC_FLAGS "-arch=sm_30")