Sparse MAtrix mpi+X Kernels is a lightweight, portable C++ library providing high-performance implementations of popular sparse matrix kernels of the form C = A op B.
| Sparse Kernel | OpenMP | CUDA | MPI | Matrix Formats |
|---|---|---|---|---|
| Matrix-Vector Multiplication -- SpMV | ✅ | ✅ | ❌ | CRS, Sell-C-sigma |
| Matrix-Multiple Vector Multiplication -- SpMM | ✅ | ❌ | ❌ | CRS |
| Matrix-Sparse Matrix Multiplication -- SpGEMM | ✅ | ❌ | ❌ | CRS |
| Triangular Solve -- SpTRSV | ✅ | ❌ | ❌ | CRS |
| Batched Triangular Solve -- SpTRSM | ✅ | ❌ | ❌ | CRS |
- Clean, minimalist library interface
- Stacked timers around key regions
- Supports signed and unsigned 16-, 32-, and 64-bit integers, as well as 32- and 64-bit floating-point types
git clone https://github.com/DanecLacey/SmaxKernels.git
cd SmaxKernels
mkdir build && cd build
cmake .. -DCMAKE_INSTALL_PREFIX=<INSTALL_PATH>
make install -jBasic usage examples are provided in the /examples directory.
cd ../examples
mkdir build && cd build
cmake .. -DCMAKE_PREFIX_PATH=<INSTALL_PATH>
make -j- Very basic demonstrations of the API are provided in
/examples/demos. - More realistic examples can be found in
/examples/applications. - There are pre-made benchmarks of the kernels provided by SmaxKernels, as well as optional third party [Eigen, MKL, PETSc] kernels to compare against, which are all found in
/examples/benchmarks. - Validation of Smax kernels against MKL kernels are found in
/examples/validation(requires MKL). - Unit tests are found in
/examples/tests
This project is very much still in development, and many features may be unfinished, broken, or subject to change.
- This project requires C++17 features
- It is assumed that all optional third party libraries are installed in
$INSTALL_PATH - The PETSc library is found via. the PkgConfig module. So if benchmarking PETSc kernels, you should configure PETSc with
--with-pkg-config=1when building.
Pull requests and issues are welcome. Please open an issue first to propose changes or feature additions.