| Pete Dashwood 2006-03-08, 3:55 am |
| Thanks Charles.
Pete.
TOP POST
"charles hottel" <jghottel@yahoo.com> wrote in message
news:a1fd9$440a61da$4f9c67a$17040@DIALUP
USA.NET...
>
> "Pete Dashwood" <dashwood@enternet.co.nz> wrote in message
> news:46v3qnFckgovU1@individual.net...
>
> <snip>
>
>
> Originally IBM mainframe hardware had its own floating-point format and it
> was called hexadecimal because the digits in the mantissa were hexadecimal
> digits. At first this seems like a strange choice because a normalized
> mantissa like .1xxxxxxxxxxx would actually have 3 leftmost leading zero
> bits. This seems wasteful of the hardware bits, but there are other
> redeeming properties that make it a good format overall. Today's
> mainframe hardware continues to support the old hexadecimal format but
> also supports the IEEE format. Some compilers now generate the IEEE
> instructions but IBM's COBOL does not.
>
> See: http://www.research.ibm.com/journal/rd/435/schwarz.html and
>
> http://docs.sun.com/source/806-3568/ncg_goldberg.html for more information
> on this.
>
> If you search the page at the last URL for "hexadecimal" you will find a
> short discussion about the tradeoffs of various floating-point formats.
> The IBM hexadecimal format is mentioned in that discussion.
>
|