On Monday 19 May 2008 14:44:03 Julian Peterson wrote:
> Well, I experimented with both a useDouble=0 version and a useDouble=1
> version.  The same csd provided these two results rendering to a file:
>
> with useDouble=0 (ie, 32 bit):
> 	CPU : 167 seconds
> with useDouble=1 (64 bit):
> 	CPU : 59 seconds
>
> I suppose this makes sense, but it was contrary to my expectations.
> It's nearly 3 times as fast to use 64bit internal samples under these
> circumstances.  These experiments were done with a macbook core 2 duo
> (64bit) processor, osx 10.5 (64bit) and csound with Word64=1.

To make any sensible statistics out of it, you should try several times for 
each build. 

And BTW, the Word64 is not needed for anything (it just adds -fPIC to the 
build flags, so it's not very useful).

>
> Thanks,
> JP
>
> On May 19, 2008, at 9:57 AM, Julian Peterson wrote:
> > Hi list.
> >
> > If you have a 64bit processor, 64bit OS, and a copy of csound built
> > with Word64=1, is there any speed advantage to useDouble=0 vs
> > useDouble=1?
> >
> > Thanks,
> > JP
> >
> >
> > Send bugs reports to this list.
> > To unsubscribe, send email sympa@lists.bath.ac.uk with body
> > "unsubscribe csound"
>
> Send bugs reports to this list.
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
> csound"



-- 
Felipe Sateler