I have built Cp2k with the following toolchain command:" <div>./install_cp2k_toolchain.sh --install-all --with-tblite=no --with-ace=no --with-gcc=system --math-mode=mkl --with-mkl=/opt/intel/oneapi/mkl/2025.2 --mpi-mode=openmpi --enable-cuda=yes --gpu-ver=A100 --target-cpu=znver2 -j 32 " </div><div><br /></div><div>And built Gromacs with the following cmake command:" </div><div><br /></div><div>cmake .. \<br /> -DGMX_BUILD_OWN_FFTW=ON \<br /> -DREGRESSIONTEST_DOWNLOAD=OFF \<br /> -DCMAKE_INSTALL_PREFIX=~/gromacs \<br /> -DBUILD_SHARED_LIBS=OFF \<br /> -DGMX_PREFER_STATIC_LIBS=ON \<br /> -DGMXAPI=OFF \<br /> -DGMX_INSTALL_NBLIB_API=OFF \<br /> -DGMX_DOUBLE=OFF \<br /> -DGMX_SIMD=AVX2_256 \<br /> -DGMX_MPI=ON \<br /> -DGMX_GPU=CUDA \ <br /> -DGMX_CUDA_TARGET_SM="80" \<br /> -DGMX_CP2K=ON \<br /> -DCMAKE_CUDA_COMPILER=/usr/local/cuda-12.6//bin/nvcc \<br /> -DCMAKE_C_COMPILER=/home/users/jonasbaa/WaterPhase/cp2k/tools/toolchain/install/openmpi-5.0.8/bin/mpicc \<br /> -DCMAKE_CXX_COMPILER=/home/users/jonasbaa/WaterPhase/cp2k/tools/toolchain/install/openmpi-5.0.8/bin/mpicxx \<br /> -DCMAKE_Fortran_COMPILER=/home/users/jonasbaa/WaterPhase/cp2k/tools/toolchain/install/openmpi-5.0.8/bin/mpifort \<br /> -DCP2K_DIR=/home/users/jonasbaa/WaterPhase/cp2k/lib/Linux-gnu-x86_64/psmp \<br /> -DCP2K_LINKER_FLAGS=' To many flags to mention here '.</div><div><br /></div><div>But when I run my energy minimization/simulation with the simple PBE functional:" </div>; CP2K section<br />qmmm-cp2k-active = true ; Activate QMMM MdModule<br />qmmm-cp2k-qmgroup = QMatoms ; Index group of QM atoms <br />qmmm-cp2k-qmmethod = PBE ; Can be amended to INPUT and then use the -qmi flag for a custom cp2k.inp file in the gmx_mpi_d grompp command.<br />qmmm-cp2k-qmcharge = 0 ; Charge of QM system<br /><div>qmmm-cp2k-qmmultiplicity = 1 ; Multiplicity of QM system",</div><div><br /></div><div>I get the following warning:" *** WARNING in fm/cp_fm_elpa.F:522 :: Setting real_kernel for ELPA failed *** ", and as a result my GPU are not active during the run. I would appreciate all help I can get on how to make sure that the QM part of my QM/MM simulation can be done on the GPU. I assume it can be done by solving this warning.</div><div><br /></div><div>Best regards,</div><div>Jonas</div><div><br /></div>
<p></p>
-- <br />
You received this message because you are subscribed to the Google Groups "cp2k" group.<br />
To unsubscribe from this group and stop receiving emails from it, send an email to <a href="mailto:cp2k+unsubscribe@googlegroups.com">cp2k+unsubscribe@googlegroups.com</a>.<br />
To view this discussion visit <a href="https://groups.google.com/d/msgid/cp2k/4d6023d1-25e1-4792-8957-1d449437a6c8n%40googlegroups.com?utm_medium=email&utm_source=footer">https://groups.google.com/d/msgid/cp2k/4d6023d1-25e1-4792-8957-1d449437a6c8n%40googlegroups.com</a>.<br />