Hi,
I have an issue regarding the installation with GPU.
I successfully built amber (Ambertools22 and Amber20) from source using
cmake. (results from make test.serial looked good)
Then I tried to compile the GPU version as suggested in Amber20 manual
(topic 20.6.5), then things failed.
I ran these commands
export CUDA_HOME=/usr/local/cuda-11.2
cd /home/xxx/amber22_scr/build
# I sudo nano run_cmake to edit -DCUDA=TRUE
./run_cmake
(I choose to use cuda 11.2 with amber20 because I tried using cuda 11.7 and
it said cuda later than 11.2 not supported)
It returned this error
--
**************************************************************************
-- Starting configuration of Amber version 22.0.0...
-- CMake Version: 3.23.2
-- For how to use this build system, please read this wiki:
-- http://ambermd.org/pmwiki/pmwiki.php/Main/CMake
-- For a list of important CMake variables, check here:
-- http://ambermd.org/pmwiki/pmwiki.php/Main/CMake-Common-Options
--
**************************************************************************
-- Amber source found, building AmberTools and Amber
-- CUDA version 11.7 detected
-- Configuring for SM3.5, SM5.0, SM5.2, SM5.3, SM6.0, SM6.1, SM7.0, SM7.5
and SM8.0
--
************************************************************
*Error: Incompatible CUDA and GNU versions 11.2.0*
See
https://stackoverflow.com/questions/6622454/cuda-incompatible-with-my-gcc-version
************************************************************
--
CMake Error at cmake/CudaConfig.cmake:124 (message):
Call Stack (most recent call first):
CMakeLists.txt:120 (include)
-- Configuring incomplete, errors occurred!
See also "/home/matthew/Amber/amber22_src/build/CMakeFiles/CMakeOutput.log".
See also "/home/matthew/Amber/amber22_src/build/CMakeFiles/CMakeError.log".
My system has an RTX A5000 GPU with nvidia driver 515.48.07, cuda 11.7.
However, in my ubuntu I have multiple versions of cuda, 11.2 and 11.7.
Initially, I thought CUDA and GNU version incompatibility means cuda could
not locate the correct version of gcc and g++ (as I have gcc-9, gcc-10,
gcc-11 in the system). So I created symlink:
gcc-10 and g++-10 to cuda 11.2 as suggested by the forum in stackoverflow
but it still does not fix the issue.
How could I solve this issue?
Best regards,
Matthew
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Aug 04 2022 - 13:35:04 PDT