TifaMMy Release 2.1.0:
LU Decomposition using OpenMP 3.0 task concept
TifaMMy isn't the fastest Matrix Multiplication yet! A library for cache oblivious matrix operations using the space filling Peano curve. Aim is for BLAS / LAPACK functionality (but with an intuitive C++ API).
Verison 2.0.0 of TifaMMy
inline and __forceinline and PeanoForceInline deactivated in framework
- benchmark tool uses TifaMMy Multiplication API Function
New features:
- parallel version of framework
- partial pivot available in LU Decomposition
CVS ist closed, please use SVN. Project is ported!
Version 1.2.1 is availabel
After the files where already in the CVS for a few days, they got the tag VERSION_1_0_0.
This version was bundled to a package that can also be obtained via the "Files" section.