See paper: https://pdfs.semanticscholar.org/84af/ba2f267459041cf0721e213d5f68f02475f5.pdf
To build: 'clone', 'cmake .' and 'make'.
Ro run a timing test for {npoints} points to interpolate and evaluate over a field of bit length {flen} run:
./test_fastpoly {npoints} {flen}