Beyond Affine Loop Parallelisation by Recurrence  Duplication

Computing Industry Doesn't Care about Performance: how I made things faster