disable-runtime-threading-support-checks with-threading-support-check-during-buildĭuring configure run a test to check which level of threading your MPI library does support (default: on) In case of MPI and OPENMP builds, ELPA is allowed to limit the number of OpenMP threads, if the threading support level is not sufficient (default: on) Stores the build config in the library objectĪssumes that BLAS/LAPACK/SCALAPACK use 64bit integers (experimentatl)Īssumes that MPI uses 64bit integers (experimental)Īllows ELPA to run on clusters of nodes with different Intel CPUs (experimental) default useīuild also ScalaPack tests for performance comparison needs MPIĭisable Fortran 2008 if compiler does not support itīuild and install python wrapper, default noĮnable support for real valued skew-symmetric matrices Set the compplex kernel KERNEL as defaultīuild SCALAPACK test cases for performanceĮXPERIMENTAL FEATURE NOT FULLY SUPPORTED YET: Allows ELPA during autotuning to re-distribute the matrix to find the best (ELPA internal) block size for block-cyclic distribution (Needs Scalapack functionalityĮnables autotuning functionality, default yesĮnables the C tests for elpa, default yesĭo NOT use assumed-size Fortran arrays. Use compute capability VALUE for GPU version,Ĭompile with only a single specific real kernel.Ĭompile with only a single specific complex kernel.Ĭompile and always use the NVIDIA GPU versionĮXPERIMENTAL: Compile and always use the AMD GPU versionĮXPERIMENTAL: Compile and always use the INTEL GPU version with-NVIDIA-GPU-compute-capability=VALUE VERY EXPERIMENTAL: build INTEL GPU kernels, default: disabled If disabled some features like autotune willīuild ELPA2 with blocking in band_to_fullĭo not build GENERIC kernels, default: enabledĭo not build SPARC64 kernels, default: disabledĭo not build SSE kernels, default: enabledĭo not build SSE_ASSEMBLY kernels, default: enabledĭo not build AVX kernels, default: enabledĭo not build AVX2 kernels, default: enabledĮxperimental feature build ARM SVE128 kernels, default: disabledĮxperimental feature build ARM SVE256 kernels, default: disabledĮxperimental feature build ARM SVE512 kernels, default: disabledīuild NVIDIA GPU kernels, default: disabledĮXPERIMENTAL: build AMD GPU kernels, default: disabled Treat error arguments in C-API as optionalįor ELPA test programs, allow redirection of Please look at configure -help for all available options.Īn excerpt of the most important ( ELPA specific) options reads as follows: configure option make check | or make check CHECK_LEVEL=extended.Installing ELPA from sourceĮLPA can be installed with the build steps You can also run the autogen.sh script that does this step for you. The needed configure script! You will have to create the configure script with autoconf. If you obtained ELPA from the official git repository, you will not find
#How to install openmp mac how to#
This is the only supported way how to build and install ELPA. Please note that ELPA is shipped with a typical configure and makeĪutotools procedure. If you want to build (or have to since no packages are available) ELPA yourself, More, will hopefully follow in the future. Pre-build packages for a number of Linux distributions like Fedora,ĭebian, and OpenSuse. It is worth having a look at the ELPA webpage*Īnd/or the repositories of your Linux distribution: there exist How to install ELPAįirst of all, if you do not want to build ELPA yourself, and you run Linux, The release of ELPA 2021.11.001.rc1 does change the API and ABI compared to the release 2020.05.002. Which makes it easier to install the ELPA library.Īs anounced, with the the release 2021.11.001.rc the legacy interface has been removed. With release of version ELPA-2017.05.001 the build process has been significantly simplified,
This file provides documentation on how to build the ELPA library in version ELPA-2021.11.001. Installation guide for the ELPA library 0.
0 kommentar(er)
