*** Attention: This is an external email. Use caution responding, opening attachments or clicking on links. ***
We’ve successfully compiled AmberTools/Amber without CUDA.
We are trying to compile against NVIDIA’s CUDA 11.5 library and getting the following error message:
[ 69%] Building C object AmberTools/src/fftw-3.3/mpi/CMakeFiles/fftw_mpi.dir/dft-solve.c.o
/usr/bin/ld: CMakeFiles/mdgx.cuda.dir/mdgx.cuda_generated_ArraySimulator.cu.o: in function `GetGpuSpecs':
tmpxft_0009c9a2_00000000-6_ArraySimulator.compute_80.cudafe1.cpp:(.text+0x17d): undefined reference to `nvmlDeviceGetComputeRunningProcesses_v2'
collect2: error: ld returned 1 exit status
Using cmake options:
-DCOMPILER=GNU -DMPI=TRUE -DCUDA=TRUE -DCMAKE_INSTALL_PREFIX=/usr/software -DDOWNLOAD_MINICONDA=FALSE -DPYTHON_EXECUTABLE=/usr/bin/python3 -DCUDA_TOOLKIT_ROOT_DIR=/usr/local/cuda-11.5
system:
DISTRIB_ID=Ubuntu
DISTRIB_RELEASE=20.04
DISTRIB_CODENAME=focal
DISTRIB_DESCRIPTION="Ubuntu 20.04.6 LTS"
Compiler
gcc --version
gcc (Ubuntu 9.4.0-1ubuntu1~20.04.2) 9.4.0
Copyright (C) 2019 Free Software Foundation, Inc.
This is free software; see the source for copying conditions. There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
/usr/local/cuda-11.5/version.json beginning:
"cuda" : {
"name" : "CUDA SDK",
"version" : "11.5.20211015"
},
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Mar 01 2024 - 11:00:02 PST
