This library is part of my lecture "Software Basics for High Performance Computing" (MATH9367) at Ulm University:

http://www.mathematik.uni-ulm.de/~lehn/ulmBLAS

http://www.mathematik.uni-ulm.de/~lehn/sghpc

And yes, I am particularly proud of the section demonstrating how to achieve peak performance for the matrix-matrix product:

http://www.mathematik.uni-ulm.de/~lehn/sghpc/gemm/index.html

Note: Further development will take place in ulmBLAS-core