Install the Pardiso MKL module

This is an optional module that enables Chrono to use the Intel MKL Pardiso solver.

Read the introduction to modules for a technical background on the modularity of the Chrono project.

Chrono::Engine usually relies on its built-in solvers, whose good perfomance are guaranteed by leveraging the internal data structure. In fact, for a wide range of applications, these suffice.
However, for higher accuracy results, a direct solver could still be needed.

This module provides access to the third-party Intel MKL Pardiso solver. The interface to Pardiso is provided by Eigen.

Features

Two Chrono-specific features are implemented:

  • sparsity pattern lock
    In many cases, the internal data structures undergo very little changes between different timesteps. In these cases it is useful to inform the solver that the sparsity pattern does not change consistently in order to speed up the matrix assembly.
  • sparsity pattern learner
    The sparsity pattern learning feature acquires the sparsity pattern in advance, in order to speed up matrix assembly. Enabled by default, the sparsity matrix learner identifies the exact matrix sparsity pattern (without actually setting any nonzeros) Look at the API section of this module for a more details.

Requirements

  • To build applications based on this unit:
  • To run applications based on this unit:
The Intel MKL Redistributables are included into the Intel MKL Library
The Intel MKL Library is now distributed for free!

Building instructions

  1. Install the Intel MKL Library following the Get Started Guide.
    • if you are not installing to the default directory, then you have to set the environment variable MKLROOT in order to point to the MKL folder, that should be something like:
      <install_folder>/Intel/oneAPI/mkl (Windows + MKL>=2020)
      <install_folder>/IntelSWTools/compilers_and_libraries/windows/mkl (Windows + MKL=2016..2019)
  2. Follow the guide for the full installation of Chrono, but when running CMake make sure that also the option ENABLE_MODULE_PARDISO_MKL is set to ON.
    The CMake output window, on Windows OS, should return the following:
    (Please mind that the MATH library is not required.)
    MKL include dirs: C:/Program Files (x86)/Intel/oneAPI/mkl/latest/include
    MKL libraries: C:/Program Files (x86)/Intel/oneAPI/mkl/latest/lib/intel64/mkl_rt.lib
    IOMP5 library: C:/Program Files (x86)/Intel/oneAPI/compiler/latest/windows/compiler/lib/intel64_win/libiomp5md.lib
    MATH library: C:/Program Files (x86)/Intel/oneAPI/compiler/latest/windows/compiler/lib/intel64_win/libmmd.lib
    MKL library dirs: C:/Program Files (x86)/Intel/oneAPI/mkl/latest/lib/intel64;C:/Program Files (x86)/Intel/oneAPI/compiler/latest/windows/compiler/lib/intel64_win
  3. Press 'Generate' and build the project.

Building this unit will produce an additional shared library, called ChronoEngine_pardisomkl, that can be linked to your application if you want to use it.
The file extension will be .dll for Win and .so on Linux.

How to use it

Set up the environment

Before creating executables that can actually leverage the Intel MKL library, you have to make sure that the Intel run-time libraries are available to the executable itself. This is achieved by setting the PATH environmental variable in a proper way. To do so, two options are available:

  • the official method proposed by Intel; this method temporarly sets the PATH variable only from the current command prompt session; this means that, if you run it from Visual Studio or from a new command prompt, this method is difficult/cumbersome for you;
  • the unofficial method, for which the PATH variable is set manually, but once and for all; description follows...

The following unofficial method needs that you set the environmental variable of your system.

  1. Add the following MKL directories to your system PATH (Windows) or LD_LIBRARY_PATH (Linux or MacOS) environment variable, adapting them to your OS and architecture:
    <install_folder>/Intel/oneAPI/mkl/latest/redist/intel64
    <install_folder>/Intel/oneAPI/compiler/latest/windows/redist/intel64_win/compiler
    or, for older installations:
    <install_folder>/IntelSWTools/compilers_and_libraries/windows/redist/intel64_win/mkl
    <install_folder>/IntelSWTools/compilers_and_libraries/windows/redist/intel64_win/compiler.
  1. Add these system environment variables
    MKL_INTERFACE_LAYER = LP64
    MKL_THREADING_LAYER = INTEL
    or, more in general, you can have different options, depending on your Architecture and the desired Threading Layer.
  2. reboot your IDE, close any open CMake

By doing this, you will be able to start the MKL-based demos and programs directly from your IDE or just by double-clicking on them.

Set up the code

  • Simply add this snippet anywhere in your code, before running the main simulation loop.
    This will inform Chrono to use the Eigen interface to the Intel MKL Pardiso solver.
    auto mkl_solver = std::make_shared<ChSolverPardisoMKL>();
    my_system.SetSolver(mkl_solver);
  • (Optional) Turn on the sparsity pattern lock (see chrono::ChSolverPardisoMKL and chrono::ChDirectSolverLS for further details)
    auto mkl_solver = std::make_shared<ChSolverPardisoMKL>();
    mkl_solver->SetSparsityPatternLock(true);
    my_system.SetSolver(mkl_solver);
  • By default, this solver uses the sparsity pattern learner (see chrono::ChDirectSolverLS) to infer the sparsity pattern before actually loading the non-zero elements. To disable the use of the sparsity pattern learner, call
    mkl_solver->UseSparsityPatternLearner(false);
  • Look at the API section of this module for documentation about classes and functions.