C++ parallel blocked matrix library PRE-RELEASE VERSION Author:  Risi Kondor Features:     Flexibility:  a range of different dense and sparse matrix classes with a uniform interface     Blocking:  support for blocked and structured vectors/matrices     Abstraction:  matrix views, operators, active matrices, expression templates     Parallelism:  multithreaded matrix operations, atomic objects, custom schedulers     Design:  fully object oriented design with a transparent API Requires:     C++11 compatible compiler Optional:     Eigen for some linear algebra tasks Code on Github:  [github] Documentation:  [pdf] BibTeX entry:  [bib] Mondrian is distributed under the terms of the GNU Public License version 3.0.