Re: Balance
> On all existing processors interval operations are slower by magnitudes
> than the corresponding floating-point operations. This is the great
> hindrance to a wider acceptance. The main reasons why software
> simulations of interval arithmetic are slow are the case selections for
> interval multiplication (9 cases) and for division (14 cases) and the
> simulation and frequent changes of the rounding mode.
in a Dagstuhl seminar in January 2006, Branimir Lambov showed how to use
*existing* hardware (namely SSE-2) to get fast interval arithmetic. See
kathrin.dagstuhl.de/files/Materials/06/06021/06021.LambovBranimir.Slides.pdf,
especially slide 18. Does any current implementation make use of his ideas?
Paul Zimmermann