The SCOREC Core is a set of C/C++ libraries for unstructured mesh
simulations on supercomputers.
For more information, start at our
What is in this repository?
- PUMI: parallel unstructured mesh infrastructure API
- PCU: Communication and file IO built on MPI
- APF: Abstract definition of meshes, fields, and related operations
- GMI: Common interface for geometric modeling kernels
- MDS: Compact but flexible array-based mesh data structure
- PARMA: Scalable partitioning and load balancing procedures
- SPR: Superconvergent Patch Recovery error estimator
- MA: Anisotropic mixed mesh adaptation and solution transfer
- SAM: Sizing anisotropic meshes
- STK: Conversion from APF meshes to Sandia's STK meshes
- ZOLTAN: Interface to run Sandia's Zoltan code on APF meshes
- PHASTA: Tools and file formats related to the PHASTA fluid solver
- MTH: Math containers and routines
- CRV: Support for curved meshes with Bezier Shapes
- PYCORE: Python Wrappers (see python_wrappers/README.md for build instructions)
- REE: Residual based implicit error estimator
How do I get set up?
- Dependencies: CMake for compiling and MPI for running
- Configuration: Typical CMake configure and build.
example_config.sh shows common options to select,
use a front-end like
ccmake to see a full list of options
- Tests: the test/ subdirectory has tests and standalone
tools that can be compiled by explicitly listing them as targets
make install places libraries and headers in
a specified prefix, application code can use these
in their own compilation process.
We also install pkg-config files for all libraries.
- Don't break the build
- See the
- If in doubt, make a branch
- Run the ctest suite
- Don't try to force push to
develop; it is disabled
Who do I talk to?
If you use these tools, please cite the following paper:
Daniel A. Ibanez, E. Seegyoung Seol, Cameron W. Smith, and Mark S. Shephard. 2016. PUMI: Parallel Unstructured Mesh Infrastructure. ACM Trans. Math. Softw. 42, 3, Article 17 (May 2016), 28 pages. DOI: https://doi.org/10.1145/2814935
We would be happy to provide feedback on journal submissions using PUMI prior to publication.