News for hackers. Stuff that matters.

A comment on Practically Accurate Floating-Point Math. Parent comment

Re:Practically Accurate Floating-Point Math by acqq on Monday October 27, 2014 @03:05PM (#8517439)
The expression won't turn to be good conditioned only by using fma. You actually missed the point of conistonwater's post.
[ Parent | on HN ]
Re:Practically Accurate Floating-Point Math by stephencanon on Monday October 27, 2014 @03:34PM (#8517666)
Condition number only matters if you assume that the inputs are perturbed by some error. Under the assumption that inputs are exact (which is appropriate to make for some uses), it does not enter into the analysis.
[ Parent | on HN ]