High-radix implementation of IEEE floating-point addition | IEEE Conference Publication | IEEE Xplore