A tool to determine at what point it is worthwhile offloading BLAS computations to the GPU on heterogeneous systems. - View it on GitHub