On 07/19/2017 09:49 AM, Paul Koning via cctalk wrote: > An example where you'd be way off is the CDC 6000 series, which has > integer mantissas and doesn't necessarily normalize the answers. So > even apart from the encoding conversion, the answer won't match. Or > the Electrologica X8, which has an exponent range substantially > larger than IEEE, uses integer mantissa, and has a normalization rule > entirely different from IEEE. ...or the IBM S/360, with uses a strange floating point normalized to 4 bit boundaries and a biased exponent that's a power of 16, not 2. 1 bit for sign, 7 bits for exponent (biased by 64) and 24 bits for mantissa.
Double-precision (64 bits) extended the mantissa to 56 bits; given the 4 bit normalization, not much more that CDC 6000 single-precision (48 bits). --Chuck
