 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
This is possibly a scoop :
Intel plans to run demos that would use POV-Ray (and possibly Megapov)
at the next LinuxWorld (in New York, Jan 29-Feb 2). The idea is to
demonstrate the speed of their Intel "Alpha" Linux Compiler on a P4 at
1.5 Gz.
In fact, they contacted me because they wanted to use my "Wet Bird" IRTC
pic for the demo, expecting it to render in 5 minutes (something that
Moore's Law don't agree with unfortunately). Very probably, they'll
contact the POV-Team and/or Nathan Kopp about compiling POV/Megapov in
an optimal way. (they also agreed for me to report this here, BTW).
G.
--
**********************
http://www.oyonale.com
**********************
Graphic experiments
Pov-ray gallery
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Gilles Tran wrote:
>
> This is possibly a scoop :
>
> Intel plans to run demos that would use POV-Ray (and possibly Megapov)
> at the next LinuxWorld (in New York, Jan 29-Feb 2). The idea is to
> demonstrate the speed of their Intel "Alpha" Linux Compiler on a P4 at
> 1.5 Gz.
>
> In fact, they contacted me because they wanted to use my "Wet Bird" IRTC
> pic for the demo, expecting it to render in 5 minutes (something that
> Moore's Law don't agree with unfortunately). Very probably, they'll
> contact the POV-Team and/or Nathan Kopp about compiling POV/Megapov in
> an optimal way. (they also agreed for me to report this here, BTW).
>
> G.
>
That's interesting, because the P4 compares quite badly to AMD's Athlon
when used with Povray. See my posting:
Date: Fri, 12 Jan 2001 20:05:26 +0100
Newsgroups: povray.general
Subject: Re: OT: cpu speed & rendering time
Christoph
--
Christoph Hormann <chr### [at] gmx de>
IsoWood include, radiosity tutorial, TransSkin and other
things on: http://www.schunter.etc.tu-bs.de/~chris/
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Christoph Hormann wrote:
>
> That's interesting, because the P4 compares quite badly to AMD's Athlon
> when used with Povray.
That's (largely) because the code is not optimized for P4; to my knowledge there
is as yet no compiler available with P4-specific optimizations. I guess that's
the point they're pushing with this new compiler demonstration. It'll be
interesting to see the difference it makes.
--
Margus Ramst
Personal e-mail: mar### [at] peakeduee
TAG (Team Assistance Group) e-mail: mar### [at] tagpovrayorg
Home page http://www.hot.ee/margusrt
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Congratulations! or as we used to say..."Way t' go!"
Gilles Tran wrote:
> In fact, they contacted me because they wanted to use my "Wet Bird" IRTC
> pic for the demo, --
>
> **********************
> http://www.oyonale.com
> **********************
> Graphic experiments
> Pov-ray gallery
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Margus Ramst wrote:
>
> That's (largely) because the code is not optimized for P4; to my knowledge there
> is as yet no compiler available with P4-specific optimizations. I guess that's
> the point they're pushing with this new compiler demonstration. It'll be
> interesting to see the difference it makes.
>
I doubt that optimisation matters that much if it's mainly about pure FPU
computation speed. I would estimate not more than about 5-10 percent.
Anyway I agree that it would be interesting to see the result.
Christoph
--
Christoph Hormann <chr### [at] gmxde>
IsoWood include, radiosity tutorial, TransSkin and other
things on: http://www.schunter.etc.tu-bs.de/~chris/
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
The interest in POVRAY is quite large for the computer industry. At the
Linuxtag Compaq and IBM had been very frequent visitors at our booth.
They asked for scenes and especially animations. More than one planned to do
realtime raytracing :-)
Marc
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> This is possibly a scoop :
It is a scoop! It's been reported to http://www.povray.co.uk/ Rick beat me
to it. :)
> Intel plans to run demos that would use POV-Ray (and possibly Megapov)
Cool.
> In fact, they contacted me because they wanted to use my "Wet Bird" IRTC
> pic for the demo, expecting it to render in 5 minutes
Ha! ... ... But what if they can do that through some strange magic? Like
when they accelerated FlaskMPEG? This would drastically alter change my
decision for buying from AMD.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Gilles Tran wrote:
> This is possibly a scoop :
>
> Intel plans to run demos that would use POV-Ray (and possibly Megapov)
> at the next LinuxWorld (in New York, Jan 29-Feb 2). The idea is to
> demonstrate the speed of their Intel "Alpha" Linux Compiler on a P4 at
> 1.5 Gz.
>
> In fact, they contacted me because they wanted to use my "Wet Bird" IRTC
> pic for the demo, expecting it to render in 5 minutes (something that
> Moore's Law don't agree with unfortunately). Very probably, they'll
> contact the POV-Team and/or Nathan Kopp about compiling POV/Megapov in
> an optimal way. (they also agreed for me to report this here, BTW).
Cool.
--
David Fontaine <dav### [at] faricynet> ICQ 55354965
My raytracing gallery: http://davidf.faricy.net/
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Well they BETTER have agreed to it, sheesh! It would go
against the spirit of POV to say no....
excited to see where this leads....
--
http://www.teleport.com/~dearmad
But why bother? I'm not interesting.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On Wed, 17 Jan 2001 20:00:07 +0100 Gilles Tran wrote:
>Intel plans to run demos that would use POV-Ray (and possibly Megapov)
>at the next LinuxWorld (in New York, Jan 29-Feb 2). The idea is to
>demonstrate the speed of their Intel "Alpha" Linux Compiler on a P4 at
>1.5 Gz.
Hi, Gilles. This wouldn't be the first time that POV-Ray has been used
to show off Intel processors. Remember this demo presented by IBM?
http://www.cnn.com/TECH/computing/9903/16/super.idg/
Obviously, when it comes to networked computers, Skyvase.pov may no
longer be a valid benchmark for processing power of this magnitude. For
single cpu home-based systems, though, it's still fun to see how close
we can get to the Cray <s>.
--
Alan - ako### [at] povrayorg - a k o n g <at> p o v r a y <dot> o r g
http://www.povray.org - Home of the Persistence of Vision Ray Tracer
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Gilles Tran wrote:
> In fact, they contacted me because they wanted to use my "Wet Bird" IRTC
> pic for the demo, expecting it to render in 5 minutes (something that
> Moore's Law don't agree with unfortunately). Very probably, they'll
> contact the POV-Team and/or Nathan Kopp about compiling POV/Megapov in
> an optimal way. (they also agreed for me to report this here, BTW).
>
It would be good, if they make this optimal way compiled program
available, so those who have money to burn and buy new processor, can
take advantage of it.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Christoph Hormann wrote:
> I doubt that optimisation matters that much if it's mainly about pure FPU
> computation speed. I would estimate not more than about 5-10 percent.
> Anyway I agree that it would be interesting to see the result.
>
The Intel person I talked to says that their latest compiler, when running on a
Pentium 4 1.5GHz, gives about 2x performance over GCC. But don't ask me what it means
since my knowledge of processors and compilers is close to nil.
G.
--
**********************
http://www.oyonale.com
**********************
Graphic experiments
Pov-ray gallery
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
J Charter wrote:
> Congratulations! or as we used to say..."Way t' go!"
In fact they won't be able to use the Wet Bird pic. It took almost a day to
render on a PII 350 so that even with their configuration it would take a few
hours. What they want is a scene able to render during the presentation, so
they're looking for something more attractive than skyvase.pov but fast enough
(like 5 minutes).
G.
--
**********************
http://www.oyonale.com
**********************
Graphic experiments
Pov-ray gallery
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
They could have my Seattle Space Needle script if they wanted it. I'd guess something
like it would render
just right. Does about an hour or two on my 500MHz PIII all depending on how large a
res., whatever AA
settings, etc.
Wonder how much faster they expect it to be?
Bob H.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <3A66B87A.899A70AA@inapg.inra.fr> , Gilles Tran
<tra### [at] inapginrafr> wrote:
> In fact they won't be able to use the Wet Bird pic. It took almost a day to
> render on a PII 350 so that even with their configuration it would take a few
> hours. What they want is a scene able to render during the presentation, so
> they're looking for something more attractive than skyvase.pov but fast enough
> (like 5 minutes).
Well, what if you reduce the output image size? Or does parsing takle very
long, too?
Thorsten
____________________________________________________
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
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Gilles Tran wrote:
>
> The Intel person I talked to says that their latest compiler, when running on a
> Pentium 4 1.5GHz, gives about 2x performance over GCC. But don't ask me what it
means
> since my knowledge of processors and compilers is close to nil.
>
That would be quite impressive although i doubt it's true for povray.
Anyway there is not much sense in speculations, so let's wait for the
result.
Christoph
--
Christoph Hormann <chr### [at] gmxde>
IsoWood include, radiosity tutorial, TransSkin and other
things on: http://www.schunter.etc.tu-bs.de/~chris/
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Thorsten Froehlich wrote:
> Well, what if you reduce the output image size? Or does parsing takle very
> long, too?
Of course, a very small size could render fast, but they want something that
impresses people, so a thumbnail is out of the question I guess. Also, this image
renders very fast for the first 600 lines and then slows down because of the big
isosurface + reflections + motion blur + media in the bottom part. Even if it could
be run in 15 minutes, the people at the show would get bored by looking at the last
200 lines being rendered pixel by pixel. People like us who are used to watching
pixels crawl on the screen would perhaps cheer up but not the general public who has
video games and movies as unique references for 3D.
G.
--
**********************
http://www.oyonale.com
**********************
Graphic experiments
Pov-ray gallery
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Gilles Tran wrote:
>
> they're looking for something more attractive than skyvase.pov but fast enough
> (like 5 minutes).
How about http://images.honesty.com/imagedata/h/506/15/25061569.jpg :)
Shiny! The way general public likes it's CG. Microsoft and nVIDIA
like it too.
______________________________________________________________________
Kari Kivisalo http://www.kivisalo.net
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
That's too bad...congratulations anyway!
Gilles Tran wrote:
> J Charter wrote:
>
> > Congratulations! or as we used to say..."Way t' go!"
>
> In fact they won't be able to use the Wet Bird pic. It took almost a day to
> render on a PII 350 so that even with their configuration it would take a few
> hours. What they want is a scene able to render during the presentation, so
> they're looking for something more attractive than skyvase.pov but fast enough
> (like 5 minutes).
>
> G.
>
> --
>
> **********************
> http://www.oyonale.com
> **********************
> Graphic experiments
> Pov-ray gallery
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Has my vote! Then again it's no R.S.O.A.C.P.
Bob H.
"Kari Kivisalo" <kar### [at] kivisalonet> wrote in message
news:3A6783F0.B18FB8C9@kivisalo.net...
> Gilles Tran wrote:
> >
> > they're looking for something more attractive than skyvase.pov but fast
enough
> > (like 5 minutes).
>
> How about http://images.honesty.com/imagedata/h/506/15/25061569.jpg :)
>
> Shiny! The way general public likes it's CG. Microsoft and nVIDIA
> like it too.
>
> ______________________________________________________________________
> Kari Kivisalo http://www.kivisalo.net
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Not to look cynic, but this just means that they have already tried the
thing and that it works.
Maybe they have even appropriately patched the compiler (did you know that
in the beginning of C compilers for PCs there were firms which embedded
handcoded asm which was emitted when the compiler did detect that it was
reading the Dhrystone benchmark, in order to look better?).
Anyway, the fact that Intel noticed your Bird means something, doesn't it? I
believe that it is the MOST realistic image ever created with POV up to now.
bye!!!
Alessandro Coppo
a.c### [at] iolit
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
This sounds interesting. Where can I see this bird?
Nekar Xenos
Alessandro Coppo <a.c### [at] iolit> wrote in message
news:3a680bbd@news.povray.org...
> Not to look cynic, but this just means that they have already tried the
> thing and that it works.
>
> Maybe they have even appropriately patched the compiler (did you know that
> in the beginning of C compilers for PCs there were firms which embedded
> handcoded asm which was emitted when the compiler did detect that it was
> reading the Dhrystone benchmark, in order to look better?).
>
> Anyway, the fact that Intel noticed your Bird means something, doesn't it?
I
> believe that it is the MOST realistic image ever created with POV up to
now.
>
> bye!!!
>
> Alessandro Coppo
> a.c### [at] iolit
>
>
>
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
J-Print News wrote:
>
> This sounds interesting. Where can I see this bird?
>
Try:
http://www.oyonale.com/ressources/english/mkof11.htm
Christoph
--
Christoph Hormann <chr### [at] gmxde>
IsoWood include, radiosity tutorial, TransSkin and other
things on: http://www.schunter.etc.tu-bs.de/~chris/
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> They asked for scenes and especially animations. More than one planned to
do
> realtime raytracing :-)
Real-time raytracing? What next - real-time raytraced games? =D
Looks like we're getting closer and closer to the Matrix...
Nekar
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> Real-time raytracing? What next - real-time raytraced games? =D
I dont think you will ever see fully real time raytraced games.
games using raytraced shadows, or raytraced refelections on certain objects
instead on an image map etc yes
for real time 3d you get so much more bang for your buck using the current
methods, and gfx accelarators are only concerned with throwing more and more
polygons around, with ever increasing textured detail
image the situation in say 5 years, a developer has a choice, hugly complex
meshes & textures, all being thrown around at blistering speed, with a huge
array of special effects at there disposal, or a vastly simpler scene, but
raytraced?
it would make more sense to raytrace where is is needed and nowhere else,
and as good rendering & lighting engine (with FSAA) will usually produce
results that are not worlds away from there raytraced counterparts.
> Looks like we're getting closer and closer to the Matrix...
and then someone says glass, and your back to square one.....
--
Rick
POV-Ray News & Resources - http://povray.co.uk
Kitty5 WebDesign - http://kitty5.com
Hi-Impact web site design & database driven e-commerce
TEL : +44 (01625) 266358 - FAX : +44 (01625) 611913 - ICQ : 15776037
PGP Public Key
http://pgpkeys.mit.edu:11371/pks/lookup?op=get&search=0x231E1CEA
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Maybe, just maybe in the future they'll have a maths processor for each
pixel on 3d cards...
Nekar
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
J-Print News wrote in message <3a6c4266@news.povray.org>...
>Maybe, just maybe in the future they'll have a maths processor for each
>pixel on 3d cards...
Not good enough. There's at least one pixel in "First Strike at Pearl
Harbor" that takes my 400Mhz AMD K6-II over half an hour to render using
media type 3.
--
Mark
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Christoph Hormann wrote:
>
> ...
>
> I doubt that optimisation matters that much if it's mainly about pure FPU
> computation speed. I would estimate not more than about 5-10 percent.
> Anyway I agree that it would be interesting to see the result.
I thought the P4 added another FPU and could simultaneously execute.
Plus can retire more instructions/clock (assuming no deleterious
branches). Since scheduling the computes requres either ASM or a good
optimizing compiler (not yet existent), I'm not suprised that running
the same old code on a P4 doesn't show improvements.
Xander
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
