The analog floating point technique | IEEE Conference Publication | IEEE Xplore