目录
目录README

=================== MAGMA README FILE ===================


  • Quick start (make)

    Create a make.inc file to indicate your C/C++ compiler, Fortran compiler, and where CUDA, HIP, CPU BLAS, and LAPACK are installed on your system. Examples are given in the make.inc-examples directory for various libraries and operating systems. The examples rely on paths such as $CUDADIR, $ROCM_PATH, and $MKLROOT being set in your environment. There are some examples in make.inc-examples

    To compile shared and static libraries in lib, and testers in testing and sparse/testing:

      make

    or:

      make lib
      make testing
      make sparse-lib
      make sparse-testing
      make install prefix=/usr/local/magma

    MAGMA uses HIP code generation to compile for AMD GPUs. MAGMA HIP sources will be generated from interface_cuda, magmablas, and sparse into interface_hip, magmablas_hip, and sparse_hip, respectively.

  • Quick start (CMake)

    There is also a CMake option to configure and build MAGMA. For more Windows-specific instructions, see README-Windows. On Unix:

    Step 0: setup If you downloaded an official release (e.g., magma-2.6.0.tar.gz), you can skip this step. If you checked out MAGMA from bitbucket, you first need to generate all the precisions. Currently this is done only by the Makefile, not by CMake. Using a minimal make.inc configuration (with BACKEND=hip or cuda), e.g.,

      echo -e 'BACKEND = hip\nFORT = true' > make.inc
      make generate

    That should run python tools/codegen.py on all the src files and create CMake.src.{hip|cuda}

    Step 1: compile

      mkdir build
      cd build
      rm -rf *            # to clear any cached CMake configuration
      cmake [options] ..  # or ccmake ..

    then:

      make

    or:

      make lib
      make testing
      make sparse-lib
      make sparse-testing
      make install

    Options include:

      -DMAGMA_ENABLE_CUDA=ON to install MAGMA for CUDA. This is default.
      -DMAGMA_ENABLE_HIP=ON to install MAGMA for HIP. This option requires
          to specify the hipcc compiler, e.g.,
          cmake -DMAGMA_ENABLE_HIP=ON -DCMAKE_CXX_COMPILER=hipcc ..
    
      -DCMAKE_INSTALL_PREFIX=/path/to/magma, directory to install MAGMA's
          headers and libraries, default /usr/local/magma.
    
      -DGPU_TARGET='target', where target includes one or more of:
          Kepler, Maxwell, Pascal, Volta, Turing, Ampere
          or valid sm_[0-9][0-9] for NVIDIA GPUs.
          For AMD GPUs include one or more valid GPU gfx numbers
          (https://llvm.org/docs/AMDGPUUsage.html#target-triples).
    
      -DBLA_VENDOR=vendor, where vendor is one of:
          Intel10_64lp, Intel10_64lp_seq, Intel10_64ilp, Intel10_64ilp_seq,
          Intel10_32, OpenBLAS, FLAME, Generic, etc.
          See https://cmake.org/cmake/help/latest/module/FindLAPACK.html
              https://cmake.org/cmake/help/latest/module/FindBLAS.html
    
      -DLAPACK_LIBRARIES='libs', where libs is the libraries to link with for
          BLAS and LAPACK, e.g., -DLAPACK_LIBRARIES='-llapack -lblas'.
          This overrides CMake's BLAS/LAPACK search.
    
      -DBUILD_SHARED_LIBS=[on|off], to turn on/off shared libraries.
    
      -DUSE_FORTRAN=[on|off], to turn on/off Fortran.
    
      -DFORTRAN_CONVENTION=flag, when USE_FORTRAN=off, where flag is one of:
          -DADD_, -DNOCHANGE, -DUPCASE
          for whether a Fortran routine such as "dgemm" is called
          dgemm_, dgemm, or DGEMM, respectively, in the BLAS/LAPACK library.
          ADD_ is most common.
  • Detailed installation instructions and further documentation is provided in docs/html/index.html or http://icl.utk.edu/projectsfiles/magma/doxygen/


  • Forum

    For more information, please refer to the MAGMA homepage and user forum:

      http://icl.utk.edu/magma/
      https://groups.google.com/a/icl.utk.edu/g/magma-user

    The MAGMA project supports the package in the sense that reports of errors or poor performance will gain immediate attention from the developers. Such reports, descriptions of interesting applications, and other comments should be posted on the MAGMA user forum.

关于

Magma 是一个大型的、支持良好的软件包,专为代数、数论、代数几何和代数组合学中的计算而设计。

35.0 MB
邀请码