[Shootout-list] Ray tracer

[John Skaller]

On Mon, 2005-06-13 at 17:11 +1000, John Skaller wrote:
> On Mon, 2005-06-13 at 00:15 +0100, Jon Harrop wrote:

> ~/flx> time speed/exes/gccopt_3_4/ack 12
> Ack(3,12): 32765
> 
> real    0m3.038s
> user    0m3.031s
> sys     0m0.002s
> ~/flx> time speed/exes/gccopt_4_0/ack 12
> Ack(3,12): 32765
> 
> real    0m2.758s
> user    0m2.757s
> sys     0m0.000s


but I should add: these figures -- and those on the Shootout --
are in fact meaningless, as the following shows: this is the very
same binary on an unloaded AMD64 giving wildly different times
on each run: from 2.784 to 3.577 in just 4 samples.

~/flx> time speed/exes/felix/ack 12
Ack(3,12): 32765

real    0m3.049s
user    0m3.042s
sys     0m0.002s
~/flx> time speed/exes/felix/ack 12
Ack(3,12): 32765

real    0m3.412s
user    0m3.403s
sys     0m0.001s
~/flx> time speed/exes/felix/ack 12
Ack(3,12): 32765

real    0m3.577s
user    0m3.568s
sys     0m0.003s
~/flx> time speed/exes/felix/ack 12
Ack(3,12): 32765

real    0m2.784s
user    0m2.782s
sys     0m0.002s

This shows that there is absolutely no such thing as
an 'unloaded' machine. The differences here are
entirely due to cache effects from random kernel
activity.


-- 
John Skaller, skaller at users.sf.net
PO Box 401 Glebe, NSW 2037, Australia Ph:61-2-96600850 
Download Felix here: http://felix.sf.net