Practically Accurate Floating-Point Math | IEEE Journals & Magazine | IEEE Xplore