On Fri, 14 Mar 2003 11:04:40 -0500, "Greg M. Johnson" <gregj:-)565### [at] aolcom>
wrote:
> "Thorsten Froehlich" <tho### [at] trfde> wrote in message
> news:3e71f4ef$1@news.povray.org...
> > Something is misconfigured.  That's all ;-)
>
> Do you suspect it's an outright  warranty issue or is this just a humorous
> comment?

knowledge?

Did you used WinGUI installed over previous version ? Did you used benchmark
from menu of from file ? Did you had the same priority settings verified ? Both
had energy saving off ? Non other applications ? Any other benchmark application
tested ?

ABX

Post a reply to this message

In article <3e71e290$1@news.povray.org>, "Greg M. Johnson" <gregj:-
)565### [at] aolcom> says...
> I got a new WinXP PC with a 2.8 GHz chip.  I ran the benchmark.pov scene
> this PC and an old one with 1 GHz.  Both boxes are WinXP.
> 
> Results:
>   2.8 GHz  13361  secs
>   1.0 GHz    5315 secs.
> 
> Comments:
> 
> 1) This *does* play in to my cynicism that advances in hardware are going to
> be taken up by the demands of the operating system for ads,  checks on the
> internet for Windows bugifixes every other minute, etc...
> 
Well, I seem to remember someone once saying that, "if you idiot proof 
things enough, only an idiot will want to use them." I think in some 
respects Windows is starting to approach that level. The blasted OS 
already liked to do incomprehensibly useless things in Win98 (like 
letting some programs spawn 20 dial-up connection wizards if they 
couldn't access the internet), now it does even more stuff behind your 
back in XP. lol

-- 
void main () {

    call functional_code()
  else
    call crash_windows();
}

Post a reply to this message

Was that an Intel or AMD CPU?

- Nekar Xenos

Post a reply to this message

3e71e290$1@news.povray.org...
> I got a new WinXP PC with a 2.8 GHz chip.  I ran the benchmark.pov scene
> this PC and an old one with 1 GHz.  Both boxes are WinXP.
>
> Results:
>   2.8 GHz  13361  secs
>   1.0 GHz    5315 secs.

Sorry, it doesn't make any sense. You sure you didn't forget the
antialiasing on the 1GHz machine ? This alone would make these figures
consistent.
The only other reason I can think of is a RAM issue (not enough RAM on the
2.8 GHz machine) but I'm not sure the benchmark is such a RAM hog.

G.


--
**********************
http://www.oyonale.com
**********************
- Graphic experiments
- POV-Ray and Poser computer images
- Posters

Post a reply to this message

I've now been able to try again.
Second try on the 2.8GHz machine:   2446 secs (vs. 13361 on first try).

Last night, I hit the Render button and went to sleep.  Next morning PC was
in sleep mode (a deeper one than I usually see on the 1.0 machine). and the
next morning the  Message board said 13361 secs.

This morning I hit Render again and went to work.  When I was able to check
the PC again, it said 2446 secs.


So two wacky things going on:
1) WinXP's overagressive sleep mode.
2) WinXP's extreme neurosis about the state of its connectivity to the
internet (I've seen renders hang due to a warning message from XP).

Both cases  512 x 384 AA 0.3 of benchmark.

Will try new experiments.


"Gilles Tran" <git### [at] wanadoofr> wrote in message
news:3e726dfb$1@news.povray.org...

>
> Sorry, it doesn't make any sense. You sure you didn't forget the
> antialiasing on the 1GHz machine ? This alone would make these figures
> consistent.

Post a reply to this message

Greg M. Johnson's furious key-hammering produced this:
> So two wacky things going on:
> 1) WinXP's overagressive sleep mode.

As far as I know, Windows XP does little but tell the computer's BIOS when to
go into sleep mode.  Also, if going into sleep mode were based on processor
activity, the computer would never go into sleep mode.

> 2) WinXP's extreme neurosis about the state of its connectivity to the
> internet

Make sure the auto-update features are off and MSN Messenger is shut down.

> (I've seen renders hang due to a warning message from XP)

Shouldn't happen in Windows.  Even if a window popped up in front of POV-Ray it
should keep tracing in the background.

-- 
/*^*/light_source{100*<-5,2,-5>2}#macro I(i,n)#while(strlen(i)>=n)#local A=asc(
substr(i,n,1));#local a=asc(substr(i,n+1,1));cylinder{<div(A,8)-12,mod(A,8)-4,4
><div(a,8)-12,mod(a,8)-4,4>,0.1pigment{rgb z}}#local n=n+2;#end#end I("ScUe[]"1
/*<*/)I("mkmtlttk"1)//@_$#!,:<"Thhis polysig brought to you by Ian Burgmyer :)"

Post a reply to this message

Greg M. Johnson's furious key-hammering produced this:
> 3) The new PC went into a sleep mode by the next morning, a harder, deeper
> sleep, and I know it does so after a much shorter time than the old one.

That's the cause of the inconsistancies.

> Perhaps the folks who design sleep modes  never consider you might need to
> do some CPU-intensive stuff while you're away: they probably figure that if
> you're not typing at the word processor, you have absolutely no need for the
> CPU at all.

If the computer's sleep state were based on CPU usage the system would never
enter sleep mode.  Any OS you use has to constantly run through instructions in
order to keep the system running smoothly.  Take a quick look in your Windows
task manager (go to the performance tab and don't move the mouse at all).
You'll notice that the processor utilization hangs between 0% and 2%.  Remember
that, even though you aren't going anything, background processes and services
still need a small chunk of CPU time to do their thing.

Knowing that, you could argue that CPU utilization can be checked before the
computer drops into sleep mode.  I don't believe this is easily possible.  CPU
utilization is (I believe) an abstraction made by software based on how it
divides CPU time between the various tasks that are running.  Implementing
something like this in a computer's BIOS would probably be impossible without
changing the entire architecture to the point of having the BIOS handle multi-
tasking, a choice that would be very expensive and not allow for the same degree
of freedom that OS developers have.

This is getting a bit off-topic.  Followups set.

-- 
/*^*/light_source{100*<-5,2,-5>2}#macro I(i,n)#while(strlen(i)>=n)#local A=asc(
substr(i,n,1));#local a=asc(substr(i,n+1,1));cylinder{<div(A,8)-12,mod(A,8)-4,4
><div(a,8)-12,mod(a,8)-4,4>,0.1pigment{rgb z}}#local n=n+2;#end#end I("ScUe[]"1
/*<*/)I("mkmtlttk"1)//@_$#!,:<"Thhis polysig brought to you by Ian Burgmyer :)"

Post a reply to this message

"Ian J. Burgmyer" <the### [at] maccom> wrote in message
news:3e72b1ca$1@news.povray.org...
> > (I've seen renders hang due to a warning message from XP)
>
> Shouldn't happen in Windows.  Even if a window popped up in front of
POV-Ray it
> should keep tracing in the background.
>

This is definitely happening.

Post a reply to this message

Greg M. Johnson <gregj:-)565### [at] aolcom> wrote:
> I got a new WinXP PC with a 2.8 GHz chip.

  Which chip? That would be nice to know.

-- 
#macro N(D)#if(D>99)cylinder{M()#local D=div(D,104);M().5,2pigment{rgb M()}}
N(D)#end#end#macro M()<mod(D,13)-6mod(div(D,13)8)-3,10>#end blob{
N(11117333955)N(4254934330)N(3900569407)N(7382340)N(3358)N(970)}//  - Warp -

Post a reply to this message

Warp's furious key-hammering produced this:
>> I got a new WinXP PC with a 2.8 GHz chip.
> 
>   Which chip? That would be nice to know.

Most likely a Pentium 4.  Most people refer to Athlons with the AMD name
(Athlon 2800+).

-- 
/*^*/light_source{100*<-5,2,-5>2}#macro I(i,n)#while(strlen(i)>=n)#local A=asc(
substr(i,n,1));#local a=asc(substr(i,n+1,1));cylinder{<div(A,8)-12,mod(A,8)-4,4
><div(a,8)-12,mod(a,8)-4,4>,0.1pigment{rgb z}}#local n=n+2;#end#end I("ScUe[]"1
/*<*/)I("mkmtlttk"1)//@_$#!,:<"Thhis polysig brought to you by Ian Burgmyer :)"

Post a reply to this message