Version 1.1.0

This DBCSR version introduces predictive modeling for the CUDA generation of the kernels. A decision tree model is built from the data obtained by autotuning certain (m, n, k)-triplets and is used to predict optimal parameters for unseen (m, n, k)-triplets.

Developed as a part of the PASC proposal "Sparse Tensor Linear Algebra Library"

Thanks to @shoshijak for the hard work to implement it!

v1.1.0-rc.0

Version 1.1.0-rc.0

Version 1.0.0

This is the first official DBCSR release, after the extraction from CP2K repository. The main requirement for this release was to preserve the original API as it was in CP2K, while several changes were internaly made:

1. On-the-fly generation of the GPU kernels (fully revisited code for the generation of the GPU kernels)
2. Prototype of C API
3. Improved algorithm for the multiplication of rectangular matrices (when M>>N and M>>K)
4. CUBLAS support for the multiplication of dense matrices
5. CMAKE support

v1.0.0-rc.4

Version v1.0.0-rc.4

v1.0.0-rc.3

Version v1.0.0-rc.3

v1.0.0-rc.2

Version 1.0.0-rc.2

v1.0.0-rc.1

Second release candidate for 1.0.0

v1.0.0-rc.0

Tagging release candidate 1 for 1.0.0