Ken <tyl### [at] pacbellnet> wrote:
: If someone would plonk down the money for a PIII optimized compiler

  You mean "a PIII optimizing compiler"?-)

-- 
#local D=array[6]{11117333955,7382340,3358,3900569407,970,4254934330}
#local I=0;#macro M()<mod(D[I],13)-6,mod(div(D[I],13),8)-3,10>#end
#while(I<6)cylinder{M()#local D[I]=div(D[I],104);M().1
pigment{rgb M()}}#local I=(D[I]>99?I:I+1);#end              /*- Warp -*/

Post a reply to this message

On Tue, 24 Apr 2001 06:15:00 -0700, Ken wrote:
>If someone would plonk down the money for a PIII optimized compiler, and
>donate it to the POV-Team, I am sure they would be happy to release an
>official version using it...

We have the Intel compiler.  We're not impressed.

Of course, my primary machine isn't even a PII, let alone a PIII, so what
do I know...

-- 
Ron Parker   http://www2.fwi.com/~parkerr/traces.html
My opinions.  Mine.  Not anyone else's.

Post a reply to this message

OK, I've done a few tests using the PIII optimised version on my Athlon 
1.33Ghz system.

Here are my results, running various versions of WinPOV with render 
priority turned up to max, and GUI priority turned right down, oh and as 
many systray programs as possible killed off.

Good old fashioned skyvase.pov rendered at 640x480 aa0.3 (standard 
quickres.ini modified to add a 500k buffer) :

Watcom 3.1g                0 minutes 19 seconds
MegaPOV 0.7                0 minutes 17 seconds
PIII optimised version     0 minutes 17 seconds

galaxy.pov also at 640x480 aa0.3 from quickres.ini :

Watcom 3.1g                1 minutes 58 seconds
MegaPOV 0.7                1 minutes 49 seconds
PIII optimised version     1 minutes 27 seconds

Pawns.pov again at 64x480 aa0.3 from quickres.ini :

Watcom 3.1g                1 minutes 47 seconds
MegaPOV 0.7                1 minutes 21 seconds
PIII optimised version     1 minutes 16 seconds

These values may be slightly skewed singe I only ran each trace once, 
rather than running multiple times each time on a fresh reboot and 
averaging results, but to be honest, that was far too much like hard 
work, so draw from these numbers what conclusions you will.

Bye for now,
     Jamie.

Post a reply to this message

> > I found it more impressive that the biggest performance boost came from a
> > new compile - think there is definitely room for an optimised pov patch
> > branch :)
> 
> If someone would plonk down the money for a PIII optimized compiler, and
> donate it to the POV-Team, I am sure they would be happy to release an
> official version using it...

I dare say that if someone were to plonk down the same cash in front of 
Nathan Kopp, he'd be happy to recompile MegaPOV with it.  Which should 
provide even more of a performance boost.

Bye for now,
     Jamie.

Post a reply to this message

In article <slr### [at] fwicom> , ron### [at] povrayorg (Ron
Parker) wrote:

> Of course, my primary machine isn't even a PII, let alone a PIII, so what
> do I know...

...that it is time to upgrade?


____________________________________________________
Thorsten Froehlich, Duisburg, Germany
e-mail: tho### [at] trfde

Visit POV-Ray on the web: http://mac.povray.org

Post a reply to this message

In article <MPG.154fe6992518dc329898ea@news.povray.org> , Jamie Davison 
<jam### [at] ntlworldcom>  wrote:

>> If someone would plonk down the money for a PIII optimized compiler, and
>> donate it to the POV-Team, I am sure they would be happy to release an
>> official version using it...
>
> I dare say that if someone were to plonk down the same cash in front of
> Nathan Kopp, he'd be happy to recompile MegaPOV with it.  Which should
> provide even more of a performance boost.

Nathan is in the POV-Team...


____________________________________________________
Thorsten Froehlich, Duisburg, Germany
e-mail: tho### [at] trfde

Visit POV-Ray on the web: http://mac.povray.org

Post a reply to this message

On Tue, 24 Apr 2001 15:06:42 -0500, Thorsten Froehlich wrote:
>In article <slr### [at] fwicom> , ron### [at] povrayorg (Ron
>Parker) wrote:
>
>> Of course, my primary machine isn't even a PII, let alone a PIII, so what
>> do I know...
>
>...that it is time to upgrade?

Well, yes, but the computer industry is conspiring against me there.
See, I have an AT power supply and all of my memory (192 MB worth!) is in 
the form of 72-pin SIMMs.  So, to upgrade from my current K6-2/366 (really
a 350, but don't tell it that) to a PIII, assuming I were to be temporarily
deprived of what sanity I have and pass up the cheaper/faster Athlon option,
would require me to replace the case, the power supply, the motherboard,
the CPU, 192 MB worth of memory, and probably an I/O card or two (since 
current motherboards are woefully deficient of ISA slots.) 

I might as well buy a new computer if I'm gonna do that.

-- 
Ron Parker   http://www2.fwi.com/~parkerr/traces.html
My opinions.  Mine.  Not anyone else's.

Post a reply to this message

What would that cost?
d.

"Ken" <tyl### [at] pacbellnet> wrote in message
news:3AE57C54.EA3F3F18@pacbell.net...
>
>
> "Rick [Kitty5]" wrote:
>
> > I found it more impressive that the biggest performance boost came from
a
> > new compile - think there is definitely room for an optimised pov patch
> > branch :)
>
> If someone would plonk down the money for a PIII optimized compiler, and
> donate it to the POV-Team, I am sure they would be happy to release an
> official version using it...
>
> --
> Ken Tyler

Post a reply to this message

Jamie Davison <jam### [at] ntlworldcom> wrote:
: These values may be slightly skewed singe I only ran each trace once, 
: rather than running multiple times each time on a fresh reboot and 
: averaging results

  The benchmarks of type "make a fresh boot, turn everything else off, give
povray the highest possible priority, don't do ANYTHING else while it's
rendering" are good for testing raw rendering speed of the processor.
  However, they are usually of no much use for the average user. Well,
they are useful when deciding which CPU to buy, but they don't give you
a too realistic idea of how fast povray is in practice with that CPU. Usually
people do other things while povray is rendering, make tens of even hundreds
of renders without a clean boot in between and so on.

-- 
#local D=array[6]{11117333955,7382340,3358,3900569407,970,4254934330}
#local I=0;#macro M()<mod(D[I],13)-6,mod(div(D[I],13),8)-3,10>#end
#while(I<6)cylinder{M()#local D[I]=div(D[I],104);M().1
pigment{rgb M()}}#local I=(D[I]>99?I:I+1);#end              /*- Warp -*/

Post a reply to this message

> : These values may be slightly skewed singe I only ran each trace once, 
> : rather than running multiple times each time on a fresh reboot and 
> : averaging results
> 
>   The benchmarks of type "make a fresh boot, turn everything else off, give
> povray the highest possible priority, don't do ANYTHING else while it's
> rendering" are good for testing raw rendering speed of the processor.
>   However, they are usually of no much use for the average user. Well,
> they are useful when deciding which CPU to buy, but they don't give you
> a too realistic idea of how fast povray is in practice with that CPU. Usually
> people do other things while povray is rendering, make tens of even hundreds
> of renders without a clean boot in between and so on.

Sorry for the delay replying.

I agree that the numbers don't mean that much in practice, but I thought 
it might demonstrate the difference the various versions of POV make on 
at least one system used by a frequenter of these groups (i.e. Me. :)

Sadly, as I don't have a P4-1.7Ghz I can't do a CPU to CPU comparison, 
but I'm not going to complain since I seem to have gotten a Pretty damn 
fast box for not a lot of money (relatively speaking)

As far as other loads on the PC go whilst rendering, well, I usually only 
have Moray running at the same time as POV, and leave my machine alone to 
render (a holdover from my DOS dats) so i suspect I'm not an average user 
<grin>

Ah well, just thought the numbers might prove useful to somebody out 
there.

Bye for now,
     Jamie.

Post a reply to this message