"jhu" <nomail@nomail> wrote:
> "Anthony D. Baye" <Sha### [at] spamnomorehotmailcom> wrote:
> > clipka <ano### [at] anonymousorg> wrote:
> > > Am 04.07.2014 23:17, schrieb clipka:
> > >
> > > > Jerome and I have spent the last days wrestling with Linux build
> > > > problems in the current master; and although the most up-to-date version
> > > > seems to build now with icpc, g++ and clang, the g++ build is
> > > > dysfunctional and doesn't render the benchmark scene properly (and also
> > > > too fast).
> > >
> > > That problem should be solved now in the current master; Anthony, would
> > > you mind giving it another try on the G5?
> >
> > Seems to be fixed.
> >
> > Render Time:
> >   Photon Time:      0 hours  0 minutes  6 seconds (6.402 seconds)
> >               using 7 thread(s) with 7.346 CPU-seconds total
> >   Radiosity Time:   No radiosity
> >   Trace Time:       0 hours 12 minutes 12 seconds (732.817 seconds)
> >               using 4 thread(s) with 2889.894 CPU-seconds total
> > POV-Ray finished
> >
> > Image looks good too.
> >
> > Regards,
> > A.D.B.
>
> That's odd. Clock for clock, it's not much faster than a PowerPC 74xx.
>
> PowerPC 7400 @ 466 MHz
>   Photon Time:      0 hours  0 minutes 37 seconds (37.148 seconds)
>               using 7 thread(s) with 36.298 CPU-seconds total
>   Radiosity Time:   No radiosity
>   Trace Time:       4 hours 29 minutes 44 seconds (16184.849 seconds)
>               using 4 thread(s) with 16095.266 CPU-seconds total
>
> That's 34.71 pps/GHz for the 7400 and 35.77 pps/core/GHz for your 970MP.


I can't remember much from the last time I took Computer Organization and
Architecture... It wasn't my strongest subject by far... but the speedup gained
from parallel processing is a function of how much of the application in
question is parallel vs how much is serial.

Have you run the numbers?

Regards,
A.D.B.

Post a reply to this message