[ODE] Sweep-and-Prune space for ODE

Aras Pranckevicius nearaz at interamotion.com
Thu Nov 25 16:53:47 MST 2004

> Sorry guys, but I just *have* to ask.  Has anybody done any conclusive
> benchmarks on the radix sorting code?  I remember a while back I did
> some pretty extensive tests and I found that radix sorting was 2x as
> fast as std::sort for integers, but the floating point version (that
> handled signs) was 2.5x slower!  This was pretty consistent for input
> sizes from 10,000 to 100,000,000.  Was I just using a crappy radix
> sort implementaiton?

But is really a floating point radix sorter any different from int radix 

(my answer to your question is no, I didn't do any serious benchmarking)

Aras Pranckevicius aka NeARAZ

More information about the ODE mailing list