 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Hello,
there is a new (second), unofficial Version of POVRAY at
http://students.washington.edu/sschmitt/pov/ for download. This Version is
optimized (with support from Intel) for PIII/Athlon and P4. And I'm deeply
impressed how immense the difference of rendertime between the non-optimized
and the optimized code is.
I use the POVRAY-File chess2.pov to compare CPU's at http://www.tabsnet.com/
since the Beginning of 2000. I hope, someone of you send me an Result of an
P4 running the new, optimized POVRAY-Version.
Here are some results of my Computer.
Rendertime in second with this Parameter :
-d -q9 +w800 +h600 +b1000 +r3 +a0.300 +v1 +ft -ichess2.pov
Non-optimized 1'st optim. 2'nd optim.
Athlon 1200 MHz 542 sec. 436 sec. 300 sec.
Pentium III 667 MHz 1056 sec. ---- 689 sec.
Tilly
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On Thu, 30 Aug 2001 20:13:57 +0200, Tilly wrote...
> Hello,
>
> there is a new (second), unofficial Version of POVRAY at
> http://students.washington.edu/sschmitt/pov/ for download. This Version is
> optimized (with support from Intel) for PIII/Athlon and P4. And I'm deeply
> impressed how immense the difference of rendertime between the non-optimized
> and the optimized code is.
IIRC, this was discussed in April in a thread in this group that I
started, but the optimised version definitely makes a difference to the
performance on my system.
Bye for now,
Jamie.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On Thu, 30 Aug 2001 20:13:57 +0200, Tilly wrote...
> Hello,
>
> there is a new (second), unofficial Version of POVRAY at
> http://students.washington.edu/sschmitt/pov/ for download. This Version is
> optimized (with support from Intel) for PIII/Athlon and P4. And I'm deeply
> impressed how immense the difference of rendertime between the non-optimized
> and the optimized code is.
>
> I use the POVRAY-File chess2.pov to compare CPU's at http://www.tabsnet.com/
> since the Beginning of 2000. I hope, someone of you send me an Result of an
> P4 running the new, optimized POVRAY-Version.
>
> Here are some results of my Computer.
> Rendertime in second with this Parameter :
> -d -q9 +w800 +h600 +b1000 +r3 +a0.300 +v1 +ft -ichess2.pov
>
> Non-optimized 1'st optim. 2'nd optim.
> Athlon 1200 MHz 542 sec. 436 sec. 300 sec.
> Pentium III 667 MHz 1056 sec. ---- 689 sec.
<sigh>
That'll teach me to read articles more thoroughly before I reply to
them...
I managed to miss the important concepts of _new_ and _second_
versions...
<goes away to attempt to reboot brain with aid of brick wall and
concussive impacts...>
Bye for now,
Jamie.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Woha, what a difference! My Duron 700 got 7m10s with those settings,
as opposed to 12m52s with the official version. Any chance MegaPov or
v3.5 will be compiled with this?
sig
--
ICQ 74734588
Post a reply to this message
|
|
| |
| |
|
|
From: Thorsten Froehlich
Subject: Re: PIII/Athlon and P4 optimized Binaries
Date: 30 Aug 2001 16:46:17
Message: <3b8ea619$1@news.povray.org>
|
|
|
| |
| |
|
|
In article <3b8e8393$1@news.povray.org> , "Tilly" <web### [at] tabsnet de>
wrote:
> I use the POVRAY-File chess2.pov to compare CPU's at http://www.tabsnet.com/
> since the Beginning of 2000. I hope, someone of you send me an Result of an
> P4 running the new, optimized POVRAY-Version.
Which "Non-optimized" version of the official 3.1g for Windows are you
using? I guess you are _not_ using the Visual C compile available as an
optional download on ftp.povray.org?
Thorsten
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
The "not-optimized" Version I mean is the Official Version that we can
download from www.povray.org . The first PIII and P4 optimized Versions from
http://students.washington.edu/sschmitt/pov/ are from April, 2001 I guess.
The second, more optimized (PIII/Athlon and P4) Versions are just a few days
on the Server.
But I used the MS-VC-Version too. But the Intel C 5.0 and 5.1 compiled
Version is much faster.
Tilly
"Thorsten Froehlich" <tho### [at] trfde> wrote in message
news:3b8ea619$1@news.povray.org...
> Which "Non-optimized" version of the official 3.1g for Windows are you
> using? I guess you are _not_ using the Visual C compile available as an
> optional download on ftp.povray.org?
>
> Thorsten
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
I'll echo the request for a MegaPOV version ....
== John ==
"Tilly" <web### [at] tabsnetde> wrote in message
news:3b8eaaac@news.povray.org...
> The "not-optimized" Version I mean is the Official Version that we can
> download from www.povray.org . The first PIII and P4 optimized Versions
from
> http://students.washington.edu/sschmitt/pov/ are from April, 2001 I guess.
> The second, more optimized (PIII/Athlon and P4) Versions are just a few
days
> on the Server.
>
> But I used the MS-VC-Version too. But the Intel C 5.0 and 5.1 compiled
> Version is much faster.
>
> Tilly
>
> "Thorsten Froehlich" <tho### [at] trfde> wrote in message
> news:3b8ea619$1@news.povray.org...
> > Which "Non-optimized" version of the official 3.1g for Windows are you
> > using? I guess you are _not_ using the Visual C compile available as an
> > optional download on ftp.povray.org?
> >
> > Thorsten
>
>
>
Post a reply to this message
|
|
| |
| |
|
|
From: Andy Cocker
Subject: Re: PIII/Athlon and P4 optimized Binaries
Date: 30 Aug 2001 19:28:36
Message: <3b8ecc24@news.povray.org>
|
|
|
| |
| |
|
|
OMG.. I hadn't realised until now just how slow my P200MMX is. I gave up at
line 209 as the render time was already 36 minutes!!
I NEED to upgrade.. with 3.5 just around the corner, that's a good enough
excuse, right? :-)
Andy Cocker
"Sigmund Kyrre Aas" <as### [at] studntnuno> wrote in message
news:fu8tot0p3r7nqqm90ikbvakej30pnptl83@4ax.com...
> Woha, what a difference! My Duron 700 got 7m10s with those settings,
> as opposed to 12m52s with the official version. Any chance MegaPov or
> v3.5 will be compiled with this?
>
> sig
> --
> ICQ 74734588
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Andy Cocker" <big### [at] mariner9fsnetcouk> wrote in message
news:3b8ecc24@news.povray.org...
> OMG.. I hadn't realised until now just how slow my P200MMX is. I gave up
at
> line 209 as the render time was already 36 minutes!!
>
> I NEED to upgrade.. with 3.5 just around the corner, that's a good enough
> excuse, right? :-)
You'd need an excuse? j/k! ha ha ha
Got to 800MHz finally here and you know I want to upgrade already!
Bob H.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Tilly wrote:
>
> Hello,
>
> there is a new (second), unofficial Version of POVRAY at
> http://students.washington.edu/sschmitt/pov/ for download. This Version is
> optimized (with support from Intel) for PIII/Athlon and P4. And I'm deeply
> impressed how immense the difference of rendertime between the non-optimized
> and the optimized code is.
The numbers for my 1333 MHz Athlon are:
Off. compile 482 s
1. optimized 391 s
2. optimized 267 s
This is just as good as getting a new 2.4 GHz processor! And a lot
cheaper!
/Ib
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
I wonder if it would be possible to get an official Athlon/P4 optimized
binary of POV-Ray 3.5. I'll ask...
--
#macro N(D,I)#if(I<6)cylinder{M()#local D[I]=div(D[I],104);M().5,2pigment{
rgb M()}}N(D,(D[I]>99?I:I+1))#end#end#macro M()<mod(D[I],13)-6,mod(div(D[I
],13),8)-3,10>#end blob{N(array[6]{11117333955,
7382340,3358,3900569407,970,4254934330},0)}// - Warp -
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Warp wrote:
>
> I wonder if it would be possible to get an official Athlon/P4 optimized
> binary of POV-Ray 3.5. I'll ask...
Bring one for me too!
_____________
Kari Kivisalo
Post a reply to this message
|
|
| |
| |
|
|
From: Andy Cocker
Subject: Re: PIII/Athlon and P4 optimized Binaries
Date: 31 Aug 2001 13:25:06
Message: <3b8fc872@news.povray.org>
|
|
|
| |
| |
|
|
"Bob H." <omn### [at] msncom> wrote in message
news:3b8f1844@news.povray.org...
> You'd need an excuse? j/k! ha ha ha
:-).. I would actually, as I can't work due to ill health, so money is in
real short supply. I'm tempted though, as I can't believe how cheap RAM is
right now! There's no reason not to have 512Mb, which compared to my
48Mb(!!!) would itself make a huge difference.
>
> Got to 800MHz finally here and you know I want to upgrade already!
I've been looking at 850MHz Celeron of 900MHz Duron, around that level, but
don't want the 810 chipset, as I'm worried that the shared memory
architecture and the modem and sound will be a noticeable hit on
performance. I can get the above for £400, with 64Mb RAM, which I'd need to
upgrade immediately.
What did you upgrade FROM to 800MHz Bob, and was performance enhanced
noticeably, and was it worth doing?
All the best,
Andy Cocker
Post a reply to this message
|
|
| |
| |
|
|
From: Andy Cocker
Subject: Re: PIII/Athlon and P4 optimized Binaries
Date: 31 Aug 2001 13:26:10
Message: <3b8fc8b2@news.povray.org>
|
|
|
| |
| |
|
|
"Kari Kivisalo" <ray### [at] engineercom> wrote in message
news:3B8F8A39.FE643FA4@engineer.com...
> Warp wrote:
> >
> > I wonder if it would be possible to get an official Athlon/P4
optimized
> > binary of POV-Ray 3.5. I'll ask...
>
> Bring one for me too!
And me! And can I have an ATHLON too?
Andy Cocker
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Andy Cocker wrote:
> What did you upgrade FROM to 800MHz Bob, and was performance enhanced
> noticeably, and was it worth doing?
I went from an Intel PI 200mmx to an Athlon 950. The difference is like
an old '62 VW bug vs. a modern race car :)
Skyvase on the 200 was 4m36s on the 950 it's 20sec..
--
Ken Tyler
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Andy Cocker" <big### [at] mariner9fsnetcouk> wrote in message
news:3b8fc872@news.povray.org...
>
> >
> > Got to 800MHz finally here and you know I want to upgrade already!
>
> What did you upgrade FROM to 800MHz Bob, and was performance enhanced
> noticeably, and was it worth doing?
Well, I wasn't exactly hurting before except that I felt confined to a
desktop system far too long and got myself a DELL notebook.
It was a 500MHz PII before (talk like it's gone but it's still here).
I have the 815E(M) chipset in this notebook, almost without a doubt not a
great system, just affordable luxury having the 3-spindle setup (HD, CD-RW,
DVD).
I don't ever seem to be able to sort out the best hardware even if I knew
what I was doing because I always end up buying what I can manage to pay for
instead.
You're lucky the RAM prices have dropped. If I had bought the two 256MB
modules back in February when I got this thing it'd have cost me around $600
USD. Got them for only $86 a couple weeks ago.
Wish you good luck with your soon to be (hopefully) new system.
Bob H.
Post a reply to this message
|
|
| |
| |
|
|
From: Andy Cocker
Subject: Re: PIII/Athlon and P4 optimized Binaries
Date: 31 Aug 2001 23:31:21
Message: <3b905689@news.povray.org>
|
|
|
| |
| |
|
|
"Ken" <tyl### [at] pacbellnet> wrote in message
news:3B9041EA.336F9E66@pacbell.net...
>
>
> Andy Cocker wrote:
>
> > What did you upgrade FROM to 800MHz Bob, and was performance enhanced
> > noticeably, and was it worth doing?
>
> I went from an Intel PI 200mmx to an Athlon 950. The difference is like
> an old '62 VW bug vs. a modern race car :)
>
> Skyvase on the 200 was 4m36s on the 950 it's 20sec..
Wow! That's some speed increase! Thanks Ken.
Andy Cocker
Post a reply to this message
|
|
| |
| |
|
|
From: Andy Cocker
Subject: Re: PIII/Athlon and P4 optimized Binaries
Date: 31 Aug 2001 23:34:14
Message: <3b905736@news.povray.org>
|
|
|
| |
| |
|
|
"Bob H." <omn### [at] msncom> wrote in message
news:3b905183@news.povray.org...
> I don't ever seem to be able to sort out the best hardware even if I knew
> what I was doing because I always end up buying what I can manage to pay
for
> instead.
Yeah, me too. That's how I ended up with my P200 just a year ago. I recently
helped my brother choose an Athlon 1.2GHz system for his family, so they
could surf a little and do some word processing!! Aaargh.. :-) Green with
envy!
> Wish you good luck with your soon to be (hopefully) new system.
Thankyou.
Andy Cocker
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
>
> I went from an Intel PI 200mmx to an Athlon 950. The difference
is like
> an old '62 VW bug vs. a modern race car :)
>
> Skyvase on the 200 was 4m36s on the 950 it's 20sec..
Huh?
What have I done wrong?
I installed a Athlon 1.4GHz (512MB DDR RAM) last month and it's
"still" 20 seconds.
I thought, it would be fine - 6 times faster than my ald machine
(PII 300), but now: no difference between your 950 and my 1400?
20 sec with Win98
22 sec with Dos
cukk
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> > Skyvase on the 200 was 4m36s on the 950 it's 20sec..
>
> Huh?
>
> What have I done wrong?
> I installed a Athlon 1.4GHz (512MB DDR RAM) last month and it's
> "still" 20 seconds.
Are you running the same compile?
Try the benchmark at www.tabsnet.de, and see how your Athlon compares to
the other in the list.
/Ib
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> Are you running the same compile?
it's 3.1g.msvc.win32 [Pentium II optimized]
which one do you recommend?
> Try the benchmark at www.tabsnet.de, and see how your Athlon
compares to
> the other in the list.
there's the 1.4 with somewhat 4:31 - i get 6:55
Is there a difference between Athlon and Athlon TBird? I thought
they were the same - if not, that may explain something.
cukk
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
KalleK <kal### [at] gmxde> wrote:
: I installed a Athlon 1.4GHz (512MB DDR RAM) last month and it's
: "still" 20 seconds.
Check that your processor is really running at 1.4GHz and not at 900MHz.
You can do this for example with the DirectX diagnostic tool. If you have
it installed properly, it sould be found under
Start/Settings/Control Panel/DirectX/DxDiag... It opens a dialog where,
among other things, it tells your processor type and its clock speed.
Check that it says "~1400MHz".
There's a problem with some bioses which detect wrongly the clock speed
of the Athlon processor. If this is the case, you have to set the right
speed by hand from bios.
I know this because it happened to me.
--
#macro N(D,I)#if(I<6)cylinder{M()#local D[I]=div(D[I],104);M().5,2pigment{
rgb M()}}N(D,(D[I]>99?I:I+1))#end#end#macro M()<mod(D[I],13)-6,mod(div(D[I
],13),8)-3,10>#end blob{N(array[6]{11117333955,
7382340,3358,3900569407,970,4254934330},0)}// - Warp -
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> Is there a difference between Athlon and Athlon TBird? I thought
> they were the same - if not, that may explain something.
What OS are you using, and what other things have you got running in
the background?
Bye for now,
Jamie.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
I tried the P4 optimization with my P4 1.4 ghz and it gave some really weird
results on skyvase.pov. Did that happen to anyone else? The P3 optimized
version worked fine.
"Tilly" <web### [at] tabsnetde> wrote in message
news:3b8e8393$1@news.povray.org...
> Hello,
>
> there is a new (second), unofficial Version of POVRAY at
> http://students.washington.edu/sschmitt/pov/ for download. This Version is
> optimized (with support from Intel) for PIII/Athlon and P4. And I'm deeply
> impressed how immense the difference of rendertime between the
non-optimized
> and the optimized code is.
>
> I use the POVRAY-File chess2.pov to compare CPU's at
http://www.tabsnet.com/
> since the Beginning of 2000. I hope, someone of you send me an Result of
an
> P4 running the new, optimized POVRAY-Version.
>
> Here are some results of my Computer.
> Rendertime in second with this Parameter :
> -d -q9 +w800 +h600 +b1000 +r3 +a0.300 +v1 +ft -ichess2.pov
>
> Non-optimized 1'st optim. 2'nd optim.
> Athlon 1200 MHz 542 sec. 436 sec. 300 sec.
> Pentium III 667 MHz 1056 sec. ---- 689 sec.
>
> Tilly
>
>
>
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> Check that your processor is really running at 1.4GHz and not
> at 900MHz.
> You can do this for example with the DirectX diagnostic tool.
> If you have it installed properly, it sould be found under
> Start/Settings/Control Panel/DirectX/DxDiag... It opens a
> dialog where, among other things, it tells your processor type
> and its clock speed.
> Check that it says "~1400MHz".
>
> There's a problem with some bioses which detect wrongly the
> clock speed of the Athlon processor. If this is the case, you
> have to set the right speed by hand from bios.
> I know this because it happened to me.
Well thank you, but it says:
AMD Athlon(tm) Processor, MMX, 3DNOW, ~1400MHz
:-/
The BIOS says 1400, too.
I'm down to 17sec, now - don't know why, but the msvc-compile
does 17~18sec now.
With the P3/Athlon optimized (from
http://students.washington.edu/sschmitt/pov/), I get 12 seconds,
but I thought Ken did not use this one with his Athlon 950, but
an official compile.
cukk
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> What OS are you using, and what other things have you got
> running in the background?
Well I tried with DOS 6.22 - I got 22 seconds
I tried Win98 - no other tasks were running, but a bunch of
processes (seen with Ctrl-Alt-Entf) - I killed nearly all of
them - it did not matter: all the time 20 seconds.
Somehow I get 17sec-18sec now, maybe because I updated my bios -
is 17sec normal for 1400?
(with Dos (watcom.3.1a) it's still 22 seconds.)
cukk
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
KalleK <kal### [at] gmxde> wrote:
: I'm down to 17sec, now
My 1.2GHz Athlon renders it in 18sec (didn't bother closing other programs
while rendering).
I think 1.4GHz with DDR could render it faster (mine has only SDRAM; or
whatever it was called... I don't remember anymore).
--
#macro N(D,I)#if(I<6)cylinder{M()#local D[I]=div(D[I],104);M().5,2pigment{
rgb M()}}N(D,(D[I]>99?I:I+1))#end#end#macro M()<mod(D[I],13)-6,mod(div(D[I
],13),8)-3,10>#end blob{N(array[6]{11117333955,
7382340,3358,3900569407,970,4254934330},0)}// - Warp -
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> it's 3.1g.msvc.win32 [Pentium II optimized]
> which one do you recommend?
Get the new Pentium III optimized compile from
http://students.washington.edu/sschmitt/pov/ (the announcement of which,
started this thread).
> there's the 1.4 with somewhat 4:31 - i get 6:55
Thats sounds about right for the difference between the standard compile
and the PIII optimized.
> Is there a difference between Athlon and Athlon TBird? I thought
> they were the same - if not, that may explain something.
I really don't know. I call my Athlon a TBird because everybody else
seems to do it :-)
Anyway, get Steve Schmitt's compile asap - you'll be amazed at the
difference it does - and start praying with the rest of us, that Steve
will also do his magic to ver 3.5, as soon as the source is available.
:)
/Ib
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> Get the new Pentium III optimized compile from
> http://students.washington.edu/sschmitt/pov/ (the announcement
> of which, started this thread).
ok. its 4:21 now.
> Anyway, get Steve Schmitt's compile asap - you'll be amazed at
> the difference it does - and start praying with the rest of us,
> that Steve will also do his magic to ver 3.5, as soon as the
> source is available.
I AM praying.
thank you
cukk
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> My 1.2GHz Athlon renders it in 18sec (didn't bother closing
> other programs while rendering).
> I think 1.4GHz with DDR could render it faster (mine has only
> SDRAM; or whatever it was called... I don't remember anymore).
At the store, they told me, that the support of DDR is not the
best yet, maybe with a new board, i could get better results.
At least I'm within this 18sec. The measurement by "int(seconds)"
isn't that exact anyway... (sigh)
thank you
cukk
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Works great - Skyvase went from 24 seconds on my 1G Athlon to 13. Problem
is, all my stuff is done in MegaPOV. Anyone seen an optimized version of
0.7?
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On 4 Sep 2001 03:29:07 -0400, Warp wrote:
>KalleK <kal### [at] gmxde> wrote:
>: I'm down to 17sec, now
>
> My 1.2GHz Athlon renders it in 18sec (didn't bother closing other programs
>while rendering).
> I think 1.4GHz with DDR could render it faster (mine has only SDRAM; or
>whatever it was called... I don't remember anymore).
My 1.4 with 256M of DDR does skyvase in 16 seconds using the 3.5 pre-beta.
I don't have 3.1 installed on this machine. I wondered how much of that time
was spent waiting for disk I/O, so I also rendered with -f instead of +ft,
but it was still about sixteen seconds.
--
#macro R(L P)sphere{L F}cylinder{L P F}#end#macro P(V)merge{R(z+a z)R(-z a-z)R(a
-z-z-z a+z)torus{1F clipped_by{plane{a 0}}}translate V}#end#macro Z(a F T)merge{
P(z+a)P(z-a)R(-z-z-x a)pigment{rgbt 1}hollow interior{media{emission T}}finish{
reflection.1}}#end Z(-x-x.2y)Z(-x-x.4x)camera{location z*-10rotate x*90}
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Ron Parker <ron### [at] povrayorg> wrote:
: I wondered how much of that time
: was spent waiting for disk I/O, so I also rendered with -f instead of +ft,
: but it was still about sixteen seconds.
I think that even windows caches most disk writes and then flushes the
caches using DMA (ie. no CPU intervention needed) so that it doesn't affect
too much the performance of programs. Of course this AFAIK.
--
#macro N(D,I)#if(I<6)cylinder{M()#local D[I]=div(D[I],104);M().5,2pigment{
rgb M()}}N(D,(D[I]>99?I:I+1))#end#end#macro M()<mod(D[I],13)-6,mod(div(D[I
],13),8)-3,10>#end blob{N(array[6]{11117333955,
7382340,3358,3900569407,970,4254934330},0)}// - Warp -
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> My 1.4 with 256M of DDR does skyvase in 16 seconds using the
3.5 pre-beta.
> I don't have 3.1 installed on this machine. I wondered how
much of that time
> was spent waiting for disk I/O, so I also rendered with -f
instead of +ft,
> but it was still about sixteen seconds.
Thank you very much for those results to compare - now I just
need 3.5 beta...
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> there is a new (second), unofficial Version of POVRAY at
> http://students.washington.edu/sschmitt/pov/ for download. This Version is
> optimized (with support from Intel) for PIII/Athlon and P4.
In case anybody wonders, it is also faster on Pentium II. My trusty old
300MHz PII takes 47 min to run the Chess2 benchmark with the standard
compile, but less than 31 min with Steve Schmitts PIII optimized
compile.
/Ib
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Ib Rasmussen <ib### [at] ibrasdk> wrote:
: In case anybody wonders, it is also faster on Pentium II. My trusty old
: 300MHz PII takes 47 min to run the Chess2 benchmark with the standard
: compile, but less than 31 min with Steve Schmitts PIII optimized
: compile.
P-III is practically a P-II with more MHz. There are no drastical additions.
This means that optimizing for P-III is practically the same as optimizing
for P-II.
--
#macro N(D,I)#if(I<6)cylinder{M()#local D[I]=div(D[I],104);M().5,2pigment{
rgb M()}}N(D,(D[I]>99?I:I+1))#end#end#macro M()<mod(D[I],13)-6,mod(div(D[I
],13),8)-3,10>#end blob{N(array[6]{11117333955,
7382340,3358,3900569407,970,4254934330},0)}// - Warp -
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
[Warp <war### [at] tagpovrayorg>]
| P-III is practically a P-II with more MHz. There are no drastical additions.
Aren't you forgetting something? The Katmai instruction set for instance?
--
This message has not been scanned for viruses.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Bjørn Vaggen Konestabo <bjo### [at] infostreamno> wrote:
: Aren't you forgetting something? The Katmai instruction set for instance?
Never heard of it.
And if there are new instructions in P-III, the optimized version mentioned
earlier certainly doesn't use them because it works fine in a P-II.
--
#macro N(D,I)#if(I<6)cylinder{M()#local D[I]=div(D[I],104);M().5,2pigment{
rgb M()}}N(D,(D[I]>99?I:I+1))#end#end#macro M()<mod(D[I],13)-6,mod(div(D[I
],13),8)-3,10>#end blob{N(array[6]{11117333955,
7382340,3358,3900569407,970,4254934330},0)}// - Warp -
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
[Warp <war### [at] tagpovrayorg>]
| Never heard of it.
Surprising.
| And if there are new instructions in P-III, the optimized version mentioned
| earlier certainly doesn't use them because it works fine in a P-II.
There are about 70 new instructions to the P-III compared to P-II,
including a new processor state which allows MMX operations to be performed
at the same time as normal floating points operations, which isn't possible
on the P-II.
I say "Normal floating points" operations because the P3 have a new
type in the so called KNI (Katmai New Instructions also known as MMX2), which
is a type of SIMD instructions. (Single Instruction Multiple Data)
The 70 new instructions are connected to eight new 128-bit 4*32-bit wite
single precision CPU registers that are tailor made for 3d vector operations.
I doubt they are used by the P3-optimized compile since 32-bit isn't wide
enough for POV-Ray, but a similar instruction set on the P4 have 64-bit
registers and could give POV-Ray quite a boost, if the compiler is willing.
For maximum speed one would have to rewrite the source code to take full
advantage of this.
I don't know you assume that asking a compiler to optimize a program for
a certain CPU will break all backwards support. It is quite possible to
include two bits of code for critical areas where new instructions are used
and have the runtime environment chose the correct one just prior to runtime.
--
This message has not been scanned for viruses.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
