I was just wondering.....what would it be like to have a realtime,
doom-style game with a POVray based rendering engine....?  Obviously,
this wouldn't be very possible (speed-wise) on today's machines, but at
the rate technology is advancing.....
	How fast do you think computers would have to be in order to run a game
like this at reasonable speed...say 15-20 fps?  Would it even be
possible?
	Kyle

Post a reply to this message

Depends on the effects you'd use. Antialiasing? Atmosphere? Halos?
Radiosity? Recommended system: 500GHz Mercred V, 100 GB holographic memory
etc... Oh, and a fast bus, too.
But seriously, raytracing was never intended for such an application. Other
methods give results much more suitable for games. And anyway, why
speculate? Just wait and see.

Margus

Kyle wrote in message <36B754E9.5C926C7D@geocities.com>...
>I was just wondering.....what would it be like to have a realtime,
>doom-style game with a POVray based rendering engine....?  Obviously,
>this wouldn't be very possible (speed-wise) on today's machines, but at
>the rate technology is advancing.....
> How fast do you think computers would have to be in order to run a game
>like this at reasonable speed...say 15-20 fps?  Would it even be
>possible?
> Kyle

Post a reply to this message

Oh, I know raytracing wasn't intended for that, but if we had the
computing power for it to be possible, why not do it?  It'd be much more
realistic looking than anything else that I can imagine.  I was just
wondering how much computing power that would be.  
	For Effects, I'd say... Antialiasing .3 and Atmosphere,Radiosity,media
= yes
Anyone have a clue?
		Kyle

Margus Ramst wrote:
> 
> Depends on the effects you'd use. Antialiasing? Atmosphere? Halos?
> Radiosity? Recommended system: 500GHz Mercred V, 100 GB holographic memory
> etc... Oh, and a fast bus, too.
> But seriously, raytracing was never intended for such an application. Other
> methods give results much more suitable for games. And anyway, why
> speculate? Just wait and see.
> 
> Margus
> 
> Kyle wrote in message <36B754E9.5C926C7D@geocities.com>...
> >I was just wondering.....what would it be like to have a realtime,
> >doom-style game with a POVray based rendering engine....?  Obviously,
> >this wouldn't be very possible (speed-wise) on today's machines, but at
> >the rate technology is advancing.....
> > How fast do you think computers would have to be in order to run a game
> >like this at reasonable speed...say 15-20 fps?  Would it even be
> >possible?
> > Kyle

Post a reply to this message

Well, do a simple test. Render a scene with all the effects you want,
representing a typical frame of the game you have in mind.
Now do some math to figure out, how many times faster your system would have
to be to render it, say, 30fps.
Believe me, the multiplier will be quite staggering.
I figure a relatively simple scene with 640x480 AA, radiosity, atmosphere,
reflections etc. would take _at least_ 10 minutes, that's 600 seconds. To
render it 30 fps, your system would need to be... computing... 18,000 times
faster. That's a LOT.

Margus

Kyle wrote in message <36B76321.FE8B62EB@geocities.com>...
>Oh, I know raytracing wasn't intended for that, but if we had the
>computing power for it to be possible, why not do it?  It'd be much more
>realistic looking than anything else that I can imagine.  I was just
>wondering how much computing power that would be.
> For Effects, I'd say... Antialiasing .3 and Atmosphere,Radiosity,media
>= yes
>Anyone have a clue?
> Kyle
>

Post a reply to this message

WOW, that is alot.  That's much much faster than I thought it'd have to
be.  Well, we all might have to wait a few months. :-(
		Kyle


Margus Ramst wrote:
> 
> Well, do a simple test. Render a scene with all the effects you want,
> representing a typical frame of the game you have in mind.
> Now do some math to figure out, how many times faster your system would have
> to be to render it, say, 30fps.
> Believe me, the multiplier will be quite staggering.
> I figure a relatively simple scene with 640x480 AA, radiosity, atmosphere,
> reflections etc. would take _at least_ 10 minutes, that's 600 seconds. To
> render it 30 fps, your system would need to be... computing... 18,000 times
> faster. That's a LOT.
> 
> Margus
> 
> Kyle wrote in message <36B76321.FE8B62EB@geocities.com>...
> >Oh, I know raytracing wasn't intended for that, but if we had the
> >computing power for it to be possible, why not do it?  It'd be much more
> >realistic looking than anything else that I can imagine.  I was just
> >wondering how much computing power that would be.
> > For Effects, I'd say... Antialiasing .3 and Atmosphere,Radiosity,media
> >= yes
> >Anyone have a clue?
> > Kyle
> >

Post a reply to this message

Yeah, It'd be fantastic. But even now, a scanline realtime render is
possible... Quake/Quake2 to name examples. Using a VooDoo2 card... 
Admittedly, the Anti aliasing is nonexistant, everything is polygon
meshes and so on, butr still...


//Spider

Post a reply to this message

What do you mean? If I'm not using Voodoo2, QII uses a different rendering
method? I'm running it in software and the only difference I see is no tex
smoothing no 16K color.

Margus

Spider wrote in message <36B768AA.50932D56@bahnhof.se>...
>Yeah, It'd be fantastic. But even now, a scanline realtime render is
>possible... Quake/Quake2 to name examples. Using a VooDoo2 card...
>Admittedly, the Anti aliasing is nonexistant, everything is polygon
>meshes and so on, butr still...
>
>
>//Spider

Post a reply to this message

Really?  I didn't know that.  That's one thing I've never understood. 
What's the difference between a scanline and a raytracing engine?  All I
know is that 3DS has a scanline renderer and it's considerably faster
than POVray.  Not as god quality though I don't think.
	Kyle



Spider wrote:
> 
> Yeah, It'd be fantastic. But even now, a scanline realtime render is
> possible... Quake/Quake2 to name examples. Using a VooDoo2 card...
> Admittedly, the Anti aliasing is nonexistant, everything is polygon
> meshes and so on, butr still...
> 
> //Spider

Post a reply to this message

There was a post here somewhere by somebody (sorry... memory failing... need
more coffee...) that gave a good overview of the existing rendering
techniques.
Basically, scanline uses Z-buffering to sort polygons by their distance from
the viewpoint. Rendering is fast, but reflections/refractions and many other
things can't be done directly; instead they have to be "mapped" onto objects
and this is not very precise. The two methods can be used together (3DSMax,
for example).

Margus

Kyle wrote in message <36B76C24.ADD718EF@geocities.com>...
>Really?  I didn't know that.  That's one thing I've never understood.
>What's the difference between a scanline and a raytracing engine?  All I
>know is that 3DS has a scanline renderer and it's considerably faster
>than POVray.  Not as god quality though I don't think.
> Kyle

Post a reply to this message

What I meant is taht the effects he asked for can't be created without a
card capable of hardware acceleration. 
You don't get coloured light, fog/mist reflective surfaces or such in a
softeware mode, Adn the render engine is actually different.

If you study the Q2 archive, you'll find teh glide.dll/software.dll(or
something) files. in theese the actual render engiunes are. So yes, they
differ in the software/hardware modes.

//Spider

Margus Ramst wrote:
> 
> What do you mean? If I'm not using Voodoo2, QII uses a different rendering
> method? I'm running it in software and the only difference I see is no tex
> smoothing no 16K color.
> 
> Margus
> 
> Spider wrote in message <36B768AA.50932D56@bahnhof.se>...
> >Yeah, It'd be fantastic. But even now, a scanline realtime render is
> >possible... Quake/Quake2 to name examples. Using a VooDoo2 card...
> >Admittedly, the Anti aliasing is nonexistant, everything is polygon
> >meshes and so on, butr still...
> >
> >
> >//Spider

Post a reply to this message

POVraytraced images are overkill for actionbased games, I guess. You'd only
notice in the screenshots. Realism is one thing FX is another. That's what you
want in action-games and -movies and such.
My 2p

Margus Ramst wrote:
> 
> Depends on the effects you'd use. Antialiasing? Atmosphere? Halos?
> Radiosity? Recommended system: 500GHz Mercred V, 100 GB holographic memory
> etc... Oh, and a fast bus, too.
> But seriously, raytracing was never intended for such an application. Other
> methods give results much more suitable for games. And anyway, why
> speculate? Just wait and see.
> 
> Margus
> 
> Kyle wrote in message <36B754E9.5C926C7D@geocities.com>...
> >I was just wondering.....what would it be like to have a realtime,
> >doom-style game with a POVray based rendering engine....?  Obviously,
> >this wouldn't be very possible (speed-wise) on today's machines, but at
> >the rate technology is advancing.....
> > How fast do you think computers would have to be in order to run a game
> >like this at reasonable speed...say 15-20 fps?  Would it even be
> >possible?
> > Kyle

Post a reply to this message

Memory would not be a big factor I think.  You would need a math
processor or some other specialized piece of hardware, like 3d boards
today.  Something maybe like a quantum computer.  

Margus Ramst wrote:
> 
> Depends on the effects you'd use. Antialiasing? Atmosphere? Halos?
> Radiosity? Recommended system: 500GHz Mercred V, 100 GB holographic memory
> etc... Oh, and a fast bus, too.
> But seriously, raytracing was never intended for such an application. Other
> methods give results much more suitable for games. And anyway, why
> speculate? Just wait and see.
> 
> Margus
> 
> Kyle wrote in message <36B754E9.5C926C7D@geocities.com>...
> >I was just wondering.....what would it be like to have a realtime,
> >doom-style game with a POVray based rendering engine....?  Obviously,
> >this wouldn't be very possible (speed-wise) on today's machines, but at
> >the rate technology is advancing.....
> > How fast do you think computers would have to be in order to run a game
> >like this at reasonable speed...say 15-20 fps?  Would it even be
> >possible?
> > Kyle

Post a reply to this message

If it is realism you want you could consider stepping outside ;-)
Oh, and please forget to take the gun.

In fact I don't think it is realism you're after, because if it were what would
happen if you went on shooting everything in the vicinity (to quote an average
game)? Not to mention the enormous complexity of the algorithms that are use to
calculate the splatter and blood. Somewhere up there someone must have one
laaaarge computer (bet it's not Intel inside).

Remco

Kyle wrote:
> 
> Oh, I know raytracing wasn't intended for that, but if we had the
> computing power for it to be possible, why not do it?  It'd be much more
> realistic looking than anything else that I can imagine.  I was just
> wondering how much computing power that would be.
>         For Effects, I'd say... Antialiasing .3 and Atmosphere,Radiosity,media
> = yes
> Anyone have a clue?
>                 Kyle
> 
> Margus Ramst wrote:
> >
> > Depends on the effects you'd use. Antialiasing? Atmosphere? Halos?
> > Radiosity? Recommended system: 500GHz Mercred V, 100 GB holographic memory
> > etc... Oh, and a fast bus, too.
> > But seriously, raytracing was never intended for such an application. Other
> > methods give results much more suitable for games. And anyway, why
> > speculate? Just wait and see.
> >
> > Margus
> >
> > Kyle wrote in message <36B754E9.5C926C7D@geocities.com>...
> > >I was just wondering.....what would it be like to have a realtime,
> > >doom-style game with a POVray based rendering engine....?  Obviously,
> > >this wouldn't be very possible (speed-wise) on today's machines, but at
> > >the rate technology is advancing.....
> > > How fast do you think computers would have to be in order to run a game
> > >like this at reasonable speed...say 15-20 fps?  Would it even be
> > >possible?
> > > Kyle

Post a reply to this message

Remco de Korte wrote:
> 
> POVraytraced images are overkill for actionbased games, I guess. You'd only
> notice in the screenshots. Realism is one thing FX is another. That's what you
> want in action-games and -movies and such.
> My 2p
> 
This is exactly what John C. (hallowed be his name) states in his .plan
files, and therefore has abandoned the 1st player single games that only
are played three times(Novice -> getting used to it; hard -> making it;
hard->finding the secrets)
Soo, back for the multiplayer, where the art will be viewed several
times, and he can do the thing he does best. A great game that people
will play _many_ times, and therefore get their value back, in time
wasted ;-)

//Spider

Post a reply to this message

Actually, you're right... the leap to quantum computers in the distant
but forseeable future should actually make Raytracing a nearly realtime
even.  (They should be able to easily break current encryption
standards)  I am personally extrememly excited about this avenue of
advancement.

Steve


portelli wrote:
> 
> Memory would not be a big factor I think.  You would need a math
> processor or some other specialized piece of hardware, like 3d boards
> today.  Something maybe like a quantum computer.
> 
> Margus Ramst wrote:
> >
> > Depends on the effects you'd use. Antialiasing? Atmosphere? Halos?
> > Radiosity? Recommended system: 500GHz Mercred V, 100 GB holographic memory
> > etc... Oh, and a fast bus, too.
> > But seriously, raytracing was never intended for such an application. Other
> > methods give results much more suitable for games. And anyway, why
> > speculate? Just wait and see.
> >
> > Margus
> >
> > Kyle wrote in message <36B754E9.5C926C7D@geocities.com>...
> > >I was just wondering.....what would it be like to have a realtime,
> > >doom-style game with a POVray based rendering engine....?  Obviously,
> > >this wouldn't be very possible (speed-wise) on today's machines, but at
> > >the rate technology is advancing.....
> > > How fast do you think computers would have to be in order to run a game
> > >like this at reasonable speed...say 15-20 fps?  Would it even be
> > >possible?
> > > Kyle

Post a reply to this message

The leap to a biological, quad computer will be far better at destroying
encryptions, but too slow for games.

just a note.

//Spider

Post a reply to this message

I'm not at all sure that is true... since each op on a biological
computer takes so much longer, I believe the numbers I've heard are that
quantum computers should be able to break 128bit encryption in 20
minutes or so... If I recall, DNA computers require like an hour for a
basic math operation, so while the difference may be insignificant, at
least gaming on a Quantum computer will rock.

Steve


Spider wrote:
> 
> The leap to a biological, quad computer will be far better at destroying
> encryptions, but too slow for games.
> 
> just a note.
> 
> //Spider

Post a reply to this message

portelli wrote:
> 
> Memory would not be a big factor I think.  You would need a math
> processor or some other specialized piece of hardware, like 3d boards
> today.  Something maybe like a quantum computer.
> 
Yeah! Yeah! Drool! Drool!
A quantum computer!

(...and then after hours, no, days (seven or so) of intensive coding you start
the realtime renderer and with a Big Bang you step into your newly created
universe...)

Post a reply to this message

Remco de Korte wrote:
> 
> portelli wrote:
> >
> > Memory would not be a big factor I think.  You would need a math
> > processor or some other specialized piece of hardware, like 3d boards
> > today.  Something maybe like a quantum computer.
> >
> Yeah! Yeah! Drool! Drool!
> A quantum computer!
> 
> (...and then after hours, no, days (seven or so) of intensive coding you start
> the realtime renderer and with a Big Bang you step into your newly created
> universe...)

Only to have it collapse from a quantum singularity... fizzle

-- 
Ken Tyler

tyl### [at] pacbellnet

Post a reply to this message

Ya a simple quantum computer today could crack the best encryption in
about 11 seconds

Stephen Lavedas wrote:
> 
> Actually, you're right... the leap to quantum computers in the distant
> but forseeable future should actually make Raytracing a nearly realtime
> even.  (They should be able to easily break current encryption
> standards)  I am personally extrememly excited about this avenue of
> advancement.
> 
> Steve
> 
> portelli wrote:
> >
> > Memory would not be a big factor I think.  You would need a math
> > processor or some other specialized piece of hardware, like 3d boards
> > today.  Something maybe like a quantum computer.
> >
> > Margus Ramst wrote:
> > >
> > > Depends on the effects you'd use. Antialiasing? Atmosphere? Halos?
> > > Radiosity? Recommended system: 500GHz Mercred V, 100 GB holographic memory
> > > etc... Oh, and a fast bus, too.
> > > But seriously, raytracing was never intended for such an application. Other
> > > methods give results much more suitable for games. And anyway, why
> > > speculate? Just wait and see.
> > >
> > > Margus
> > >
> > > Kyle wrote in message <36B754E9.5C926C7D@geocities.com>...
> > > >I was just wondering.....what would it be like to have a realtime,
> > > >doom-style game with a POVray based rendering engine....?  Obviously,
> > > >this wouldn't be very possible (speed-wise) on today's machines, but at
> > > >the rate technology is advancing.....
> > > > How fast do you think computers would have to be in order to run a game
> > > >like this at reasonable speed...say 15-20 fps?  Would it even be
> > > >possible?
> > > > Kyle

Post a reply to this message

Ken wrote:
> 
> Remco de Korte wrote:
> >
> > portelli wrote:
> > >
> > > Memory would not be a big factor I think.  You would need a math
> > > processor or some other specialized piece of hardware, like 3d boards
> > > today.  Something maybe like a quantum computer.
> > >
> > Yeah! Yeah! Drool! Drool!
> > A quantum computer!
> >
> > (...and then after hours, no, days (seven or so) of intensive coding you start
> > the realtime renderer and with a Big Bang you step into your newly created
> > universe...)
> 
> Only to have it collapse from a quantum singularity... fizzle
> 
> --
> Ken Tyler
> 
> tyl### [at] pacbellnet

Hee hee!
But isn't that actually a passage to another (liquid) universum?
That'd be neat, especially with media.

Remco

Post a reply to this message

What I've read points at the fact that a biological computer would
increase its efficiency by each "generation" and thus, in a short range
thing(addition, and so on) would loose, but in a longer term equation
would gain on the quantum computer. (count all decimals to pi, the
gemoetrical sum of 1 to sqrt(-1) and so on... (yes, I know that neither
of them are possible, but, whatta heck :-) )

//Spider

Stephen Lavedas wrote:
> 
> I'm not at all sure that is true... since each op on a biological
> computer takes so much longer, I believe the numbers I've heard are that
> quantum computers should be able to break 128bit encryption in 20
> minutes or so... If I recall, DNA computers require like an hour for a
> basic math operation, so while the difference may be insignificant, at
> least gaming on a Quantum computer will rock.
> 
> Steve
> 
> Spider wrote:
> >
> > The leap to a biological, quad computer will be far better at destroying
> > encryptions, but too slow for games.
> >
> > just a note.
> >
> > //Spider

Post a reply to this message

I think very simple quantum computers have been made.  With only a few
'bits'.  Maybe I'm mistaken but I remember reading about them in
Scientific American.

Spider wrote:
> 
> What I've read points at the fact that a biological computer would
> increase its efficiency by each "generation" and thus, in a short range
> thing(addition, and so on) would loose, but in a longer term equation
> would gain on the quantum computer. (count all decimals to pi, the
> gemoetrical sum of 1 to sqrt(-1) and so on... (yes, I know that neither
> of them are possible, but, whatta heck :-) )
> 
> //Spider
> 
> Stephen Lavedas wrote:
> >
> > I'm not at all sure that is true... since each op on a biological
> > computer takes so much longer, I believe the numbers I've heard are that
> > quantum computers should be able to break 128bit encryption in 20
> > minutes or so... If I recall, DNA computers require like an hour for a
> > basic math operation, so while the difference may be insignificant, at
> > least gaming on a Quantum computer will rock.
> >
> > Steve
> >
> > Spider wrote:
> > >
> > > The leap to a biological, quad computer will be far better at destroying
> > > encryptions, but too slow for games.
> > >
> > > just a note.
> > >
> > > //Spider

Post a reply to this message

On Tue, 02 Feb 1999 21:01:35 -0500, Stephen Lavedas
<swl### [at] virginiaedu> wrote:

>Actually, you're right... the leap to quantum computers in the distant
>but forseeable future should actually make Raytracing a nearly realtime
>even.  (They should be able to easily break current encryption
>standards)  I am personally extrememly excited about this avenue of
>advancement.
>
>Steve

Don't you all guys think that by the time quantum computers and so are
available POV-Ray 9.3j will also be available (free!) with things like
true (integral) volumetrics, variable iors, backwards photon tracing,
true causics, trun diffuse interreflection,  wave behaviour,
wavelength-dependent scattering, diffraction, interference, true
dispersion, troposhperic refraction, hydrometeoric scattering, Doppler
effect, Einstein red deviation, photoelectric effect, luminance, heat
colors etc. The list goes on. Even if such machines are available in
three months, I can imagine Chris Young rubbing his hands with a big
smile on his face, and a post in povray.general like "With CPU
preformance so greatly increased since 3.1a we can at last include
some year-old ideas that have already been developed and coded but
proved inadequately slow for the public 3.1 release. Expect a quick
3.14159265 release featuring..." and the list I just mentioned.

Just daydreaming...

Peter

Post a reply to this message

They've created some gates...not computers, but I believe the NOT
operation, and I think they were close to both AND and OR when I last
checked (could have gotten them by now) So it is only a matter of
time... Every indication is that there is no physical or theoretical
barrier to quantum computers, it will just take time.  I GUESS I can
wait.

Steve


portelli wrote:
> 
> I think very simple quantum computers have been made.  With only a few
> 'bits'.  Maybe I'm mistaken but I remember reading about them in
> Scientific American.
> 
> Spider wrote:
> >
> > What I've read points at the fact that a biological computer would
> > increase its efficiency by each "generation" and thus, in a short range
> > thing(addition, and so on) would loose, but in a longer term equation
> > would gain on the quantum computer. (count all decimals to pi, the
> > gemoetrical sum of 1 to sqrt(-1) and so on... (yes, I know that neither
> > of them are possible, but, whatta heck :-) )
> >
> > //Spider
> >
> > Stephen Lavedas wrote:
> > >
> > > I'm not at all sure that is true... since each op on a biological
> > > computer takes so much longer, I believe the numbers I've heard are that
> > > quantum computers should be able to break 128bit encryption in 20
> > > minutes or so... If I recall, DNA computers require like an hour for a
> > > basic math operation, so while the difference may be insignificant, at
> > > least gaming on a Quantum computer will rock.
> > >
> > > Steve
> > >
> > > Spider wrote:
> > > >
> > > > The leap to a biological, quad computer will be far better at destroying
> > > > encryptions, but too slow for games.
> > > >
> > > > just a note.
> > > >
> > > > //Spider

Post a reply to this message

hehehe I hope... but Quantum computers are such a GIANT leap forward,
that I don't think we can conceive of a rendering application that would
take long enough to bother one...Current computers would take 1000's of
years to crack RSA encryption...With quantum computers cracking it in
under and hour (I remember 20 minutes vaguely, and someone else posted
11 seconds) the jump is incredible... Have you ever started a 1000 year
render?  I don't think so... (except maybe that one Spider sent... it
never did finish the bounding boxes on my computer)

Steve

Peter Popov wrote:
> 
> On Tue, 02 Feb 1999 21:01:35 -0500, Stephen Lavedas
> <swl### [at] virginiaedu> wrote:
> 
> >Actually, you're right... the leap to quantum computers in the distant
> >but forseeable future should actually make Raytracing a nearly realtime
> >even.  (They should be able to easily break current encryption
> >standards)  I am personally extrememly excited about this avenue of
> >advancement.
> >
> >Steve
> 
> Don't you all guys think that by the time quantum computers and so are
> available POV-Ray 9.3j will also be available (free!) with things like
> true (integral) volumetrics, variable iors, backwards photon tracing,
> true causics, trun diffuse interreflection,  wave behaviour,
> wavelength-dependent scattering, diffraction, interference, true
> dispersion, troposhperic refraction, hydrometeoric scattering, Doppler
> effect, Einstein red deviation, photoelectric effect, luminance, heat
> colors etc. The list goes on. Even if such machines are available in
> three months, I can imagine Chris Young rubbing his hands with a big
> smile on his face, and a post in povray.general like "With CPU
> preformance so greatly increased since 3.1a we can at last include
> some year-old ideas that have already been developed and coded but
> proved inadequately slow for the public 3.1 release. Expect a quick
> 3.14159265 release featuring..." and the list I just mentioned.
> 
> Just daydreaming...
> 
> Peter

Post a reply to this message

Stephen Lavedas wrote:
> 
> hehehe I hope... but Quantum computers are such a GIANT leap forward,
> that I don't think we can conceive of a rendering application that would
> take long enough to bother one...Current computers would take 1000's of
> years to crack RSA encryption...With quantum computers cracking it in
> under and hour (I remember 20 minutes vaguely, and someone else posted
> 11 seconds) the jump is incredible... Have you ever started a 1000 year
> render?  I don't think so... (except maybe that one Spider sent... it
> never did finish the bounding boxes on my computer)
hey, wich one was that ??? (bad memory... Remind me)

//Spider

Post a reply to this message

Stephen Lavedas wrote:
> 
> hehehe I hope... but Quantum computers are such a GIANT leap forward,
> that I don't think we can conceive of a rendering application that would
> take long enough to bother one...Current computers would take 1000's of
> years to crack RSA encryption...With quantum computers cracking it in
> under and hour (I remember 20 minutes vaguely, and someone else posted
> 11 seconds) the jump is incredible... Have you ever started a 1000 year
> render?  I don't think so... (except maybe that one Spider sent... it
> never did finish the bounding boxes on my computer)
> 
> Steve

The 1000 year render is the one that renders every cell in the
human body - one by one. You can watch as a real person grows
from initial conception (tee hee) to birth of the finish form.

-- 
Ken Tyler

tyl### [at] pacbellnet

Post a reply to this message

Margus Ramst wrote:
> 
> Well, do a simple test. Render a scene with all the effects you want,
> representing a typical frame of the game you have in mind.
> Now do some math to figure out, how many times faster your system would have
> to be to render it, say, 30fps.
> Believe me, the multiplier will be quite staggering.
> I figure a relatively simple scene with 640x480 AA, radiosity, atmosphere,
> reflections etc. would take _at least_ 10 minutes, that's 600 seconds. To
> render it 30 fps, your system would need to be... computing... 18,000 times
> faster. That's a LOT.

Considering Moores law
(computing power doubles every 18 months), then a simple
calculation tells us that computers will be that fast in
14.15 Moore-periods (log2(18000)), that being 254.5 months,
which are only 21 years.

Hold your breath.... ;-)

Of course this only applies if Moores law will last that long.
It is considered as coming to an end around 2015.....

Markus
-- 
"Denkst Du vielleicht über den nächsten Satz nach, den Du sagen willst?
 Ich nicht eine Sekunde!!!" H. Heinol in Val Thorens

Post a reply to this message

I think quantum computers won't be very efficient in raytracing.
Since the thing they thrive on is calculating several possibilities
at once... Oh, well, a thought strikes me: They will be able to
compute all imaginable raytraced scenes _at_once_

This would be the end of raytracing. :-(

STOP QUANTUM COMPUTING AT ONCE!!!!

Markus
-- 
"Denkst Du vielleicht über den nächsten Satz nach, den Du sagen willst?
 Ich nicht eine Sekunde!!!" H. Heinol in Val Thorens


Stephen Lavedas wrote:
> 
> hehehe I hope... but Quantum computers are such a GIANT leap forward,
> that I don't think we can conceive of a rendering application that would
> take long enough to bother one...Current computers would take 1000's of
> years to crack RSA encryption...With quantum computers cracking it in
> under and hour (I remember 20 minutes vaguely, and someone else posted
> 11 seconds) the jump is incredible... Have you ever started a 1000 year
> render?  I don't think so... (except maybe that one Spider sent... it
> never did finish the bounding boxes on my computer)

Post a reply to this message

Hasn't Moore's Law already been broken?  

Markus Becker wrote:
> 
> Margus Ramst wrote:
> >
> > Well, do a simple test. Render a scene with all the effects you want,
> > representing a typical frame of the game you have in mind.
> > Now do some math to figure out, how many times faster your system would have
> > to be to render it, say, 30fps.
> > Believe me, the multiplier will be quite staggering.
> > I figure a relatively simple scene with 640x480 AA, radiosity, atmosphere,
> > reflections etc. would take _at least_ 10 minutes, that's 600 seconds. To
> > render it 30 fps, your system would need to be... computing... 18,000 times
> > faster. That's a LOT.
> 
> Considering Moores law
> (computing power doubles every 18 months), then a simple
> calculation tells us that computers will be that fast in
> 14.15 Moore-periods (log2(18000)), that being 254.5 months,
> which are only 21 years.
> 
> Hold your breath.... ;-)
> 
> Of course this only applies if Moores law will last that long.
> It is considered as coming to an end around 2015.....
> 
> Markus
> --
> "Denkst Du vielleicht über den nächsten Satz nach, den Du sagen willst?
>  Ich nicht eine Sekunde!!!" H. Heinol in Val Thorens

Post a reply to this message

Except the experts say quantum computers would only be good for certain
things.  Like searching mega databases or cracking keys.  Some
applications would not run well on a computer will more than two bits. 
I don't know how rendering would work here.  I read that a quantum
computer would search a database by looking at all of the entries at one
time.  So maybe one could render every pixel at once?!

Stephen Lavedas wrote:
> 
> hehehe I hope... but Quantum computers are such a GIANT leap forward,
> that I don't think we can conceive of a rendering application that would
> take long enough to bother one...Current computers would take 1000's of
> years to crack RSA encryption...With quantum computers cracking it in
> under and hour (I remember 20 minutes vaguely, and someone else posted
> 11 seconds) the jump is incredible... Have you ever started a 1000 year
> render?  I don't think so... (except maybe that one Spider sent... it
> never did finish the bounding boxes on my computer)
> 
> Steve
> 
> Peter Popov wrote:
> >
> > On Tue, 02 Feb 1999 21:01:35 -0500, Stephen Lavedas
> > <swl### [at] virginiaedu> wrote:
> >
> > >Actually, you're right... the leap to quantum computers in the distant
> > >but forseeable future should actually make Raytracing a nearly realtime
> > >even.  (They should be able to easily break current encryption
> > >standards)  I am personally extrememly excited about this avenue of
> > >advancement.
> > >
> > >Steve
> >
> > Don't you all guys think that by the time quantum computers and so are
> > available POV-Ray 9.3j will also be available (free!) with things like
> > true (integral) volumetrics, variable iors, backwards photon tracing,
> > true causics, trun diffuse interreflection,  wave behaviour,
> > wavelength-dependent scattering, diffraction, interference, true
> > dispersion, troposhperic refraction, hydrometeoric scattering, Doppler
> > effect, Einstein red deviation, photoelectric effect, luminance, heat
> > colors etc. The list goes on. Even if such machines are available in
> > three months, I can imagine Chris Young rubbing his hands with a big
> > smile on his face, and a post in povray.general like "With CPU
> > preformance so greatly increased since 3.1a we can at last include
> > some year-old ideas that have already been developed and coded but
> > proved inadequately slow for the public 3.1 release. Expect a quick
> > 3.14159265 release featuring..." and the list I just mentioned.
> >
> > Just daydreaming...
> >
> > Peter

Post a reply to this message

portelli wrote:
> 
> Except the experts say quantum computers would only be good for certain
> things.  Like searching mega databases or cracking keys.  Some
> applications would not run well on a computer will more than two bits.
> I don't know how rendering would work here.  I read that a quantum
> computer would search a database by looking at all of the entries at one
> time.  So maybe one could render every pixel at once?!
> 
Ok, stop it here, we shouldn't yet develop software for a nonexistant
platform. Let's wait until we have a hardware defining list before we
speculate, ok ?
This thread is rather off-topic right now... :-)
do some fireworks or something.*smile*

//Spider

Post a reply to this message

http://members.xoom.com/darrelltroy/jesusinside.gif

Post a reply to this message

portelli wrote:
> 
> Hasn't Moore's Law already been broken?

No, it still holds. Perhaps you meant what I mistakenly
wrote ;-). Not the processor _speed_ but the number of
transistors doubles every 18 months. Together with the
MHz increase this more than doubles the speed of the
processors. And, as far as what I have read, it will
go on until approx. 2015, then it will reach some
principal problem concerning quantum physics.
But the cute technicians at <insert_favourate_manufacturer
here> will have invented some new technique to keep
Moore's law going...

Markus

-- 
"Denkst Du vielleicht über den nächsten Satz nach, den Du sagen willst?
 Ich nicht eine Sekunde!!!" H. Heinol in Val Thorens

Post a reply to this message

Yeah, that problem with wavelengths reaching magnitudes less than the
size of the electrons being moved (or at least too close to be usable)
it should actually be getting quite difficult very soon.

Steve


Markus Becker wrote:
> 
> portelli wrote:
> >
> > Hasn't Moore's Law already been broken?
> 
> No, it still holds. Perhaps you meant what I mistakenly
> wrote ;-). Not the processor _speed_ but the number of
> transistors doubles every 18 months. Together with the
> MHz increase this more than doubles the speed of the
> processors. And, as far as what I have read, it will
> go on until approx. 2015, then it will reach some
> principal problem concerning quantum physics.
> But the cute technicians at <insert_favourate_manufacturer
> here> will have invented some new technique to keep
> Moore's law going...
> 
> Markus
> 
> --
> "Denkst Du vielleicht über den nächsten Satz nach, den Du sagen willst?
>  Ich nicht eine Sekunde!!!" H. Heinol in Val Thorens

Post a reply to this message

Stephen Lavedas wrote:

> Yeah, that problem with wavelengths reaching magnitudes less than the
> size of the electrons being moved (or at least too close to be usable)
> it should actually be getting quite difficult very soon.
>
> Steve
>
> Markus Becker wrote:
> >
> > portelli wrote:
> > >
> > > Hasn't Moore's Law already been broken?
> >
> > No, it still holds. Perhaps you meant what I mistakenly
> > wrote ;-). Not the processor _speed_ but the number of
> > transistors doubles every 18 months. Together with the
> > MHz increase this more than doubles the speed of the
> > processors. And, as far as what I have read, it will
> > go on until approx. 2015, then it will reach some
> > principal problem concerning quantum physics.
> > But the cute technicians at <insert_favourate_manufacturer
> > here> will have invented some new technique to keep
> > Moore's law going...
> >
> > Markus
> >
> > --
> > "Denkst Du vielleicht über den nächsten Satz nach, den Du sagen willst?
> >  Ich nicht eine Sekunde!!!" H. Heinol in Val Thorens

this is just until Quantum computer processors come out ( at least 20 years
hence). they have ONE prototype working (only can pick an answer to a
preprogrammed question 400x the fastest super computer) but the problem is
right now the quantum computer weighs 3 tons (like the original mainframes
which PCs are based on), maybe soon we will have megafast computers based on
quantum physics like the one stated above....

psychomek

Post a reply to this message