OT : Approximate / fast math libraries ?

[Bruno Wolff III]

On Tue, Sep 04, 2007 at 17:32:45 -0700,
  Les <hlhowell at pacbell.net> wrote:
> 
> 	The same is true of Floating point operations vs integers.  When floats
> had to be calculated by loops with an integer processor, they were
> expensive and integers were faster.  Now with high speed floating point
> units, simple float operations are quite fast if done in line.  Ditto

On some models of crays the floating point units were faster than the
integer units. I don't remember if that was because interger operations
may not have been vectorizable or if there was some other reason for it.