[QE-users] QE installation error with gpu

[Pietro Davide Delugas]

Hello
Try with CC=nvc, nvcc is the cuda  compiler

Il 28 giu 2024 4:20 PM, sjang at sejong.edu ha scritto:
Hi all,
I am new to QE. I am trying to install QE with GPU.  I installed nvidia hpc_sdk v.24.5 for this on
/opt/nvidia/hpc_sdk/Linux_x86_64/24.5
My gpu is Geforce RTX 3080 (Ampere architecture I suppose).

After successful (?) configure command, I made "make all" and I got following error.
To me, it seems nvcc compiler doesn’t take '-gpu=cc80,cuda12.4' option.

In fact,
$ which nvcc
/opt/nvidia/hpc_sdk/Linux_x86_64/2024/cuda/12.4/bin/nvcc
$ nvcc -gpu=cc80,cuda12.4
nvcc fatal   : Unknown option '-gpu=cc80,cuda12.4'

I might be wrong though.

Any advice or help will be greatly apprecited.
Thank you very much in advance.
SJ

$make all
make[1]: Entering directory `/home/sjang/install/qe-7.3.1/UtilXlib'
/opt/nvidia/hpc_sdk/Linux_x86_64/24.5/comm_libs/12.4/openmpi4/openmpi-4.1.5/bin/mpif90 -fast -Mcache_align -Mpreprocess -Mlarge_arrays -D__PGI -D__CUDA -D__DFTI -D__MPI -D__SCALAPACK  -cuda -gpu=cc80,cuda12.4 -I/home/sjang/install/qe-7.3.1//external/devxlib/src -I/home/sjang/install/qe-7.3.1//external/devxlib/include -acc -I/home/sjang/install/qe-7.3.1//external/devxlib/src -I. -I/home/sjang/install/qe-7.3.1//include -I/usr/local/intel/compilers_and_libraries_2020.4.304/linux/mkl/include -I. -c parallel_include.f90
/opt/nvidia/hpc_sdk/Linux_x86_64/24.5/comm_libs/12.4/openmpi4/openmpi-4.1.5/bin/mpif90 -fast -Mcache_align -Mpreprocess -Mlarge_arrays -D__PGI -D__CUDA -D__DFTI -D__MPI -D__SCALAPACK  -cuda -gpu=cc80,cuda12.4 -I/home/sjang/install/qe-7.3.1//external/devxlib/src -I/home/sjang/install/qe-7.3.1//external/devxlib/include -acc -I/home/sjang/install/qe-7.3.1//external/devxlib/src -I. -I/home/sjang/install/qe-7.3.1//include -I/usr/local/intel/compilers_and_libraries_2020.4.304/linux/mkl/include -I. -c util_param.f90
/opt/nvidia/hpc_sdk/Linux_x86_64/24.5/comm_libs/12.4/openmpi4/openmpi-4.1.5/bin/mpif90 -fast -Mcache_align -Mpreprocess -Mlarge_arrays -D__PGI -D__CUDA -D__DFTI -D__MPI -D__SCALAPACK  -cuda -gpu=cc80,cuda12.4 -I/home/sjang/install/qe-7.3.1//external/devxlib/src -I/home/sjang/install/qe-7.3.1//external/devxlib/include -acc -I/home/sjang/install/qe-7.3.1//external/devxlib/src -I. -I/home/sjang/install/qe-7.3.1//include -I/usr/local/intel/compilers_and_libraries_2020.4.304/linux/mkl/include -I. -c clib_wrappers.f90
/opt/nvidia/hpc_sdk/Linux_x86_64/24.5/comm_libs/12.4/openmpi4/openmpi-4.1.5/bin/mpif90 -fast -Mcache_align -Mpreprocess -Mlarge_arrays -D__PGI -D__CUDA -D__DFTI -D__MPI -D__SCALAPACK  -cuda -gpu=cc80,cuda12.4 -I/home/sjang/install/qe-7.3.1//external/devxlib/src -I/home/sjang/install/qe-7.3.1//external/devxlib/include -acc -I/home/sjang/install/qe-7.3.1//external/devxlib/src -I. -I/home/sjang/install/qe-7.3.1//include -I/usr/local/intel/compilers_and_libraries_2020.4.304/linux/mkl/include -I. -c nvtx_wrapper.f90
/opt/nvidia/hpc_sdk/Linux_x86_64/24.5/comm_libs/12.4/openmpi4/openmpi-4.1.5/bin/mpif90 -fast -Mcache_align -Mpreprocess -Mlarge_arrays -D__PGI -D__CUDA -D__DFTI -D__MPI -D__SCALAPACK  -cuda -gpu=cc80,cuda12.4 -I/home/sjang/install/qe-7.3.1//external/devxlib/src -I/home/sjang/install/qe-7.3.1//external/devxlib/include -acc -I/home/sjang/install/qe-7.3.1//external/devxlib/src -I. -I/home/sjang/install/qe-7.3.1//include -I/usr/local/intel/compilers_and_libraries_2020.4.304/linux/mkl/include -I. -c clocks_handler.f90
nvcc -O3 -D__PGI -D__CUDA -D__DFTI -D__MPI -D__SCALAPACK -I. -I/home/sjang/install/qe-7.3.1//include -I/usr/local/intel/compilers_and_libraries_2020.4.304/linux/mkl/include -gpu=cc80,cuda12.4 -acc  -c cptimer.c
nvcc fatal   : Unknown option '-gpu=cc80,cuda12.4'
make[1]: *** [cptimer.o] Error 1
make[1]: Leaving directory `/home/sjang/install/qe-7.3.1/UtilXlib'
make: *** [libutil] Error 1

--------------------------------------------------------------------------------------------------
Before ‘make all’ command, this is my configure command and resulting log

$echo $NVHPC_CUDA_HOME
/opt/nvidia/hpc_sdk/Linux_x86_64/2024/cuda/12.4

$./configure F90=nvfortran MPIF90=/opt/nvidia/hpc_sdk/Linux_x86_64/24.5/comm_libs/12.4/openmpi4/openmpi-4.1.5/bin/mpif90 CC=nvcc CXX=nvc++ --with-cuda=$NVHPC_CUDA_HOME --prefix=$HOME/QE --with-cuda-cc=80 --with-cuda-runtime=12.4

….
….
all dependencies updated successfully
checking build system type... x86_64-pc-linux-gnu
checking ARCH... x86_64
checking setting AR... ... ar
checking setting ARFLAGS... ... ruv
checking whether the Fortran compiler works... yes
checking for Fortran compiler default output file name... a.out
checking for suffix of executables...
checking whether we are cross compiling... no
checking for suffix of object files... o
checking whether we are using the GNU Fortran compiler... no
checking whether /opt/nvidia/hpc_sdk/Linux_x86_64/24.5/compilers/bin/nvfortran accepts -g... yes
checking for Fortran flag to compile .f90 files... none
checking for /opt/nvidia/hpc_sdk/Linux_x86_64/24.5/comm_libs/12.4/openmpi4/openmpi-4.1.5/bin/mpif90... no
checking whether we are using the GNU Fortran compiler... no
checking whether  accepts -g... no
configure: WARNING: MPIF90 not found: using MPIF90 anyway
checking version of /opt/nvidia/hpc_sdk/Linux_x86_64/24.5/comm_libs/12.4/openmpi4/openmpi-4.1.5/bin/mpif90... nvfortran 24.5-1
configure: WARNING: Check whether user-supplied F90 is consistent with MPIF90 !!!
setting F90... nvfortran
setting MPIF90... /opt/nvidia/hpc_sdk/Linux_x86_64/24.5/comm_libs/12.4/openmpi4/openmpi-4.1.5/bin/mpif90
checking whether we are using the GNU C compiler... yes
checking whether nvcc accepts -g... yes
checking for nvcc option to accept ISO C89... none needed
setting CC... nvcc
setting CFLAGS... -O3
using F90... nvfortran
setting FFLAGS... -fast
setting F90FLAGS... -fast -Mcache_align -Mpreprocess -Mlarge_arrays
setting FFLAGS_NOOPT... -O0
setting FFLAGS_NOMAIN... -Mnomain
setting CPP... cpp
setting CPPFLAGS... -P -traditional -Uvector
setting LD... /opt/nvidia/hpc_sdk/Linux_x86_64/24.5/comm_libs/12.4/openmpi4/openmpi-4.1.5/bin/mpif90
setting LDFLAGS...
checking whether Fortran compiler accepts -cuda -gpu=cuda12.4... yes
checking whether make sets $(MAKE)... yes
checking whether Fortran files must be preprocessed... no
checking for library containing dgemm... -lmkl_intel_lp64
setting BLAS_LIBS... -lmkl_intel_lp64 -lmkl_sequential -lmkl_core
checking FFT...
checking MASS...
checking for library containing mpi_init... none required
checking for library containing pdgemr2d... no
checking for library containing pdgemr2d... -lmkl_scalapack_lp64
checking ELPA...
checking for Environ... not used
checking if legacy plugins are enabled ...  no
checking for ranlib... ranlib
checking for wget... wget -O
setting WGET... wget -O
checking for git... git
setting DFLAGS... -D__PGI -D__CUDA -D__DFTI -D__MPI -D__SCALAPACK
setting IFLAGS... -I. -I$(TOPDIR)/include -I/usr/local/intel/compilers_and_libraries_2020.4.304/linux/mkl/include
configure: creating ./config.status
config.status: creating install/make_lapack.inc
config.status: creating include/configure.h
config.status: creating make.inc
config.status: creating configure.msg
config.status: creating install/make_wannier90.inc
--------------------------------------------------------------------
ESPRESSO can take advantage of several optimized numerical libraries
(essl, fftw, mkl...).  This configure script attempts to find them,
but may fail if they have been installed in non-standard locations.
If a required library is not found, the local copy will be compiled.

The following libraries have been found:
  BLAS_LIBS=  -lmkl_intel_lp64  -lmkl_sequential -lmkl_core
  LAPACK_LIBS=
  SCALAPACK_LIBS=-lmkl_scalapack_lp64 -lmkl_blacs_openmpi_lp64
  FFT_LIBS=

Please check if this is what you expect.

If any libraries are missing, you may specify a list of directories
to search and retry, as follows:
  ./configure LIBDIRS="list of directories, separated by spaces"

Parallel environment detected successfully.\
Configured for compilation of parallel executables.

For more info, read the ESPRESSO User's Guide (Doc/users-guide.tex).
--------------------------------------------------------------------
configure: success


--------------------------------------------------------------------------------------------------
Here is my PATH info

echo $PATH
/opt/nvidia/hpc_sdk/Linux_x86_64/24.5/compilers/extras/qd/bin:/opt/nvidia/hpc_sdk/Linux_x86_64/24.5/cuda/bin:/opt/nvidia/hpc_sdk/Linux_x86_64/2024/cuda/12.4/bin:/opt/nvidia/hpc_sdk/Linux_x86_64/24.5/comm_libs/mpi/bin:/opt/nvidia/hpc_sdk/Linux_x86_64/24.5/compilers/bin:/home/sjang/software/xdrfile/bin:/home/sjang/software/plumed272_intel/bin:/usr/local/mpi/intel19/openmpi-3.1.6/bin:/home/sjang/software/GMX2021_plumed272_intel/bin:/home/sjang/vmd/bin:/home/sjang/software/packmol:/home/sjang/software/rosetta_src_2021.16.61629_bundle/main/source/bin:/home/sjang/g16/g16/bsd:/home/sjang/g16/g16:/usr/local/intel/compilers_and_libraries_2020.4.304/linux/bin/intel64:/usr/local/intel/compilers_and_libraries_2020.4.304/linux/bin:/usr/local/intel/compilers_and_libraries_2020.4.304/linux/mpi/intel64/libfabric/bin:/usr/local/intel/compilers_and_libraries_2020.4.304/linux/mpi/intel64/bin:/usr/local/intel/debugger_2020/gdb/intel64/bin:/opt/rh/devtoolset-9/root/usr/bin:/usr/local/anaconda/3/2021.05/bin:/usr/local/anaconda/3/2021.05/condabin:/opt/nvidia/hpc_sdk/Linux_x86_64/2024/cuda/12.4/bin:/usr/lib64/qt-3.3/bin:/usr/torque/sbin:/usr/torque/bin:/opt/pbs/sbin:/opt/pbs/bin:/usr/local/jspcmds:.:/home/my_uer_name:/usr/local/bin:/usr/bin:/usr/local/sbin:/usr/sbin:/opt/dell/srvadmin/bin


--------------------------------------------------------------------------------------------------
Here is my LD_LIBRARY_PATH info

echo $LD_LIBRARY_PATH
/opt/nvidia/hpc_sdk/Linux_x86_64/24.5/comm_libs/nvshmem/lib:/opt/nvidia/hpc_sdk/Linux_x86_64/24.5/comm_libs/nccl/lib:/opt/nvidia/hpc_sdk/Linux_x86_64/24.5/math_libs/lib64:/opt/nvidia/hpc_sdk/Linux_x86_64/24.5/compilers/lib:/opt/nvidia/hpc_sdk/Linux_x86_64/24.5/compilers/extras/qd/lib:/opt/nvidia/hpc_sdk/Linux_x86_64/24.5/cuda/extras/CUPTI/lib64:/opt/nvidia/hpc_sdk/Linux_x86_64/24.5/cuda/lib64:/opt/nvidia/hpc_sdk/Linux_x86_64/2024/cuda/12.4/extras/CUPTI/lib64:/opt/nvidia/hpc_sdk/Linux_x86_64/2024/cuda/12.4/lib64:/opt/nvidia/hpc_sdk/Linux_x86_64/24.5/math_libs:/usr/local/lib:/home/sjang/software/xdrfile/lib:/home/sjang/software/plumed272_intel/lib:/usr/local/mpi/intel19/openmpi-3.1.6/lib:/usr/local/intel/compilers_and_libraries_2020.4.304/linux/compiler/lib/intel64_lin:/usr/local/intel/compilers_and_libraries_2020.4.304/linux/mpi/intel64/libfabric/lib:/usr/local/intel/compilers_and_libraries_2020.4.304/linux/mpi/intel64/lib/release:/usr/local/intel/compilers_and_libraries_2020.4.304/linux/mpi/intel64/lib:/usr/local/intel/compilers_and_libraries_2020.4.304/linux/ipp/lib/intel64:/usr/local/intel/compilers_and_libraries_2020.4.304/linux/mkl/lib/intel64_lin:/usr/local/intel/compilers_and_libraries_2020.4.304/linux/tbb/lib/intel64/gcc4.8:/usr/local/intel/debugger_2020/python/intel64/lib:/usr/local/intel/debugger_2020/libipt/intel64/lib:/usr/local/intel/compilers_and_libraries_2020.4.304/linux/daal/lib/intel64_lin:/usr/local/intel/compilers_and_libraries_2020.4.304/linux/daal/../tbb/lib/intel64_lin/gcc4.4:/usr/local/intel/compilers_and_libraries_2020.4.304/linux/daal/../tbb/lib/intel64_lin/gcc4.8:/opt/rh/devtoolset-9/root/usr/lib64:/opt/rh/devtoolset-9/root/usr/lib:/opt/rh/devtoolset-9/root/usr/lib64/dyninst:/opt/rh/devtoolset-9/root/usr/lib/dyninst:/opt/nvidia/hpc_sdk/Linux_x86_64/2024/cuda/12.4/extras/CUPTI/lib64:/opt/nvidia/hpc_sdk/Linux_x86_64/2024/cuda/12.4/lib64:/home/sjang/g16/gv/lib


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20240628/a24bf558/attachment.html>