Schwarz Library
Performance results
- Paper in IJHPCA; Alternative arXiv version
- Two stage update
Required components
The required components include:
- Ginkgo: The Ginkgo library is needed. It needs to be installed and preferably the installation path
provided as an environment variable in
Ginkgo_DIR
variable.
- MPI: As multiple nodes and a domain decomposition is used, an MPI implementation is necessary.
Quick Install
Building Schwarz-Lib
To build Schwarz-Lib, you can use the standard CMake procedure.
mkdir build; cd build
cmake -G "Unix Makefiles" .. && make
By default, SCHWARZ_BUILD_BENCHMARKING
is enabled. This allows you to quickly run an example with the timings if needed. For a detailed list of options available see the Benchmarking page.
For more CMake options please refer to the Installation page
Currently implemented features
-
Executor paradigm:
- [x] GPU.
- [x] OpenMP.
- [ ] Single rank per node and threading in one node.
-
Factorization paradigm:
- [x] CHOLMOD.
- [x] UMFPACK.
-
Solving paradigm:
- Direct:
- [x] Ginkgo.
- [x] CHOLMOD.
- [x] UMFPACK.
- Iterative:
- [x] Ginkgo.
- [ ] deal.ii.
-
Partitioning paradigm:
- [x] METIS.
- [x] Regular, 1D.
- [x] Regular, 2D.
- [ ] Zoltan.
-
Convergence check:
- [x] Centralized, tree based convergence (Yamazaki 2019).
- [x] Decentralized, leader election based (Bahi 2005).
-
Communication paradigm.
- [x] Onesided.
- [x] Twosided.
- [ ] Event based.
-
Communication strategies.
- [x] Remote comm strategies:
- [x] MPI_Put , gathered.
- [x] MPI_Put , one by one.
- [x] MPI_Get , gathered .
- [x] MPI_Get , one by one.
- [x] Lock strategies: MPI_Win_lock / MPI_Win_lock_all .
- [x] Lock all and unlock all.
- [x] Lock local and unlock local.
- [x] Flush strategies: MPI_Win_flush / MPI_Win_flush_local .
- [x] Flush all.
- [x] Flush local.
-
Schwarz problem type.
Any of the implemented features can be permuted and tested.
Known Issues
- On Summit, the Spectrum MPI seems to have a bug with using
MPI_Put
with GPU buffers. MPI_Get
works as expected. This bug has also been confirmed with an external micro-benchmarking library, OSU Micro-Benchmarks.
For installing and building, please check the Installation page
Credits: This code (written in C++, with additions and improvements) was inspired by the code from Ichitaro Yamazaki, ICL, UTK.