This document
describes the new features and changes in each release of the Combinatorial
BLAS. The numbering is of the form Z.Y.X.
X denotes bug
fixes and/or minor updates for existing functionality.
Y denotes new major
functionality.
Z denotes an
overall paradigm shift in terms of abstractions.
CombBLAS
1.6.2 (April
2018)
·
Several bug fixes related to AWPM usage
and its SuperLU integration
·
Modernized cmake
usage thanks to Tristan
CombBLAS
1.6.1 (January
2018)
·
Several bug fixes
·
Modernizing some calls to ParallelReadMM as opposed to ReadDistribute
·
Introducing combblas::
namespace, thanks to Tristan.
CombBLAS
1.6 (September
2017)
·
Fully parallel text-file reader for
vectors (FullyDistSpVec::ParallelReadMM() and FullyDistVec::ParallelReadMM())
·
Fully parallel text-file writer for
vectors (FullyDistSpVec::ParallelWrite () and FullyDistVec::ParallelWrite())
·
Reverse Cuthill-McKee
(RCM) ordering implementation
-
Please cite [12] if you use this
implementation
·
Novel multithreaded SpGEMM
and SpMV (with sparse vectors) algorithms are
integrated with the rest of CombBLAS.
-
For benchmarking
multithreaded SpMV with sparse vectors, go to
Applications/SpMSpV-IPDPS2017 directory and use the code there.
-
Please cite [13] if you use the new
multithreaded SpMV with sparse vectors.
·
Extended CSC support
·
Previously deprecated SpParVec and DenseParVec (that
were distributed to diagonal processors only) classes are removed.
·
Lots of more bug
fixes
CombBLAS 1.5 (January 2016)
·
Fully parallel matrix market format
reader (SpParMat::ParallelReadMM())
·
Complete multithreading support,
including SpGEMM (previously it was solely SpMV), enabled by -DTHREADED during compilation
·
Experimental 3D SpGEMM
(the ability to switch processor grids from 2D to 3D will have to wait for
version 1.6)
-
Please cite [9] if you use this
implementation
-
cd 3DSpGEMM/, make test_mpipspgemm,
and call the executable with correct parameters
·
Maximal and Maximum cardinality
matching algorithms on bipartite graphs
-
Please cite [10] for maximal
cardinality and [11] for maximum cardinality matching
-
cd MaximumMatching,
make bpmm, and call the executable with correct parameters
·
Automated MPI_Op
creation from simple C++ function objects (simplifies semiring
descriptions and Reduce() functions)
·
FullyDistSpVec::Invert() to map from/to (integer)
values to/from indices
·
Many more helper functions
·
Experimental CSC support for low
concurrencies
·
Lots of bug fixes
CombBLAS
1.4 (January
2014)
CombBLAS
1.3 (February
2013)
Go back
to the the Combinatorial BLAS home page.