Extended Precision
Robert P. Munafo, 2023 Jun 21.
This term refers to any number format for calculation with greater precision than the largest that has commonly available "native" hardware support. Currently this means that anything more than double precision counts as extended precision.
There is an 80-bit "extended precision" format that had hardware support on the Intel 8087 coprocessor and subsequent CPUs such as 486 and Pentium, with IEEE 754 compliance. It is most notable for having a 16-bit exponent and supports numbers as small as 10-4951 making it useful as a floatexp solution for zooms deeper than 10308×.
See double-double and arbitrary-precision arithmetic.
revisions: 20230619 first version; 20230621 80-bit 8087 format
From the Mandelbrot Set Glossary and Encyclopedia, by Robert Munafo, (c) 1987-2024.
Mu-ency main page — index — recent changes — DEMZ
This page was written in the "embarrassingly readable" markup language RHTF, and was last updated on 2023 Jul 01.
s.27