2 Sep
2006
2 Sep
'06
5:34 p.m.
Chuck Guzis wrote:
On 9/2/2006 at 11:48 PM John Honniball wrote:
From the web page of William Kahan's, foremost expert in the IEEE fp
standardization effort, a paper frmo 1981 titled:
Why do we need a floating-point arithmetic standard?
http://www.cs.berkeley.edu/~wkahan/ieee754status/why-ieee.pdf
It includes some tables of the features of FP arithmetic of various
mainframe architectures, plus some handheld calculators, and the hp-85.
Interestingly, it mentions the proposed IEEE standard, as implemented
by the intel 8087 and the motorola 6839.
It runs 50 pages and I have only glossed it, but it could be a good
place to start to answer this question.
Um, it's Institute of Electrical and
Electronic
Engineers, the people who devised the standard
for floating-point representation.
One has to ask "Why?" It seems that mainframes got along just fine for
years and years without a "standard". For that matter, does any machine
actually carry out computations in IEEE format, or is the situation like
the x86 NDP--convert to a higher-precision "internal" format before
jiggling bits?