Next gen graphics chips approach - were's the software?

[g__day]

Once again the new generation of 3d cards is approaching quickly. We are months away from R420 and NV40 - which may give us dazzling performance compared to our 9800 Pros and 5950 Ultras. But does anyone feel software that will utilise their capabilities is anywhere closer to appearing during the lifetime of these next generation of video cards than it did for their predecessors?

At least now it seems the industry has adopted scalable graphics processing in the most modern of 3d game engines (e.g. Source, Krass, CryEngine etc...). Engines that fall back to simpler shaders if they assess you don't have enough CPU/GPU 3d power to run incredible shaders. I imagine these engines will make it easier for game developers to also dial up the graphics load for any new DX9 or OGL card that appears that is detected and assessed as having performance capabilities way beyond what we have today - maybe???

But does anyone else fell we are getting closer to the time when leading edge cards' capabilities can actually be satisfactorily used in their lifetime?

Or asked another way if 2004 delivers us all of NV40, NV45, R420 and R500 - a possibility - will any game engine scale to actually push them? Are game engines and shader development teams good enough to deliver surperb shaders that are held in reserve simply waiting for detection of a card fast enough to run them? Or is development of such scalable shader effects simply viewed as too much wasted effort - even if the best 3d engines could rather elegantly detect them and rather simply deliver their effects?

What are your thoughts?

 

[Sxotty]

You can rest assured that some engines will be released so poorly coded that the perform poorly on the most powerful GPU's, so yes they will be pushed... but as to whether they will be pushed into providing their maximum capabilities I will leave that up to the fortune tellers.

 

[Tahir2]

My thoughts are I can play Need For Speed: Underground on a GF4MX Integrated GPU and boy am I glad.

My other thought is the fact that I am partly responsible for holding the 3D games technology industry back but I don't care! And I know I would look down on anyone with a DX7 gfx card if I had a DX9 graphics card in my system. Heh!

 

[DeathKnight]

I will look down upon thee.

 

[incurable]

What are your thoughts?

My thoughts are that:

a) I don't think it's likely that we'll see high-end GPUs pushed to their limits by contemporary games, at least not because of higher-quality content. Bad coding can & will do that, though.

IMHO, it just doesn't make much sense to give development times and support ressources to code that can't be reasonably run on any GPU on the market. Not to talk about the customer feedback ...

b) I want a high-end consumer graphics card with dual DVI out.
(totally OT you say? Nahhhhh. )

cu

incurable

 

[bloodbob]

HALO, tomb raider , misc other games and HL2 was due to come out in september but yeah they have had problems. The PS2.0 up take isn't that bad compared to PS1.0 and 1.3/1.4

 

[rwolf]

Look how valve spent 5x longer creating a custom code path for NV3X. We need graphics vendors to follow standards so there is one code path for all video cards.

 

[Hyp-X]

We could always add twice as many objects to our maps, and/or use twice as many texture layers to push new cards to their limits.

Altough I have a feeling that's not what you are looking for.

 

[Hyp-X]

Look how valve spent 5x longer creating a custom code path for NV3X.

I beleive they spent 5x time figuring out what to change to make the code faster.
Once you know what's slow and what's fast its pretty straightforward to optimize...

Too bad you have to spend your time researching stuff that the IHVs developer support should tell you...

 

[3dilettante]

Is there a consensus on whether PS 2.0 is being picked up faster than PS 1.x?

I'm personally a proponent of a more pessimistic school of thought, that no new graphics generation is considered well established until the hardware of the early adopters is inadequate.

I'm thinking that the current and soon to be current DX9 games will wind up looking downright antiquated before the real DX9 games come out, for which maybe only the 9800 XT will provide anything approaching decent performance at standard settings.

Once again, the thanks goes out to the early adopters: the ones who fund the advancement of 3d technology, by buying bleeding edge hardware that will ironically not cut it when the software that fully utilizes their featureset comes around.

 

[KPixel]

We need graphics vendors to follow standards so there is one code path for all video cards.

Standards in hardware will never be totally followed, simply because each vendor want to be different from its competitors, having something others don"t have.
What would be cool is if they all take the standard as minimum and try to add their specifics stuff on top; but that is also very difficult because :
- The standard is very high, so they have to make some choices
- And when they can raise the bar, the perf don't follow (see FP32 and 1024int of the NV30...)

So, to take full advantage of each (or some) high-end cards, we will always have to write "some" specifics code paths for those cards.

 

[Sxotty]

I'm personally a proponent of a more pessimistic school of thought, that no new graphics generation is considered well established until the hardware of the early adopters is inadequate.

I'm thinking that the current and soon to be current DX9 games will wind up looking downright antiquated before the real DX9 games come out, for which maybe only the 9800 XT will provide anything approaching decent performance at standard settings.

I agree 100% with the above stated.

 

[KimB]

Look how valve spent 5x longer creating a custom code path for NV3X. We need graphics vendors to follow standards so there is one code path for all video cards.

BS. The architectures begin development before the standards are adopted. What we need is a software interface that is more flexible, and can accomodate more different hardware easily (i.e. OpenGL's HLSL).

 

[Joe DeFuria]

What we need is a software interface that is more flexible, and can accomodate more different hardware easily (i.e. OpenGL's HLSL).

B.S.

We've "had that" with the original GL model and extensions. Doesn't seem to have helped "any more".

 

[jb]

g__day,

I just dont see how any game can be coded to push future tech. I mean we know how stuff seems to constantly change as far as standards go. The developers would have to assume that this effect will run in one manor but when its finally realized it may run in a different manor. So no I don't think that will ever be the case.....

 

[Brimstone]

The game STALKER: Oblivion Lost is going to use PS 2.0. The game will feature a DX 8 and DX 9 engine. The DX 9 version will support up to 3 million polygons per frame.

A week or so ago they released a new trailer showing off the light and shadowing effects in this game. I thought the trailer was impressive showing off a day and night cycle.



Day & Night cycle trailer link page for STALKER

 

[Hyp-X]

showing off the light and shadowing

You mean it shows that the shadows don't change as the lightsource (sun) moves, don't you?

 

[g__day]

jb

I agree with you on what would be required and the risks to be taken. Each new cycle of technologys offers more - it doesn't simply obliterate its predecssors. I am sure a few game developers know where the major IHVs are going under NDA.

One scalable choice would be simply put more complex objects on screen - swarms of weak and strong monsters rather than one big one all the time (Serious Sam vs Doom 3 approach) to push more polygons.

Lighting effects might prove tricky to capitalise on more grunt.

If you wanted to position a game so it was ready for a next generation of video upon its release you'd have to say - I'd like to do these 5 common things - prioritise a little, work out fall backs when your detected card can't do one or more of the five - and leave them there ready for when the cards powerful and fast enough appear.

You could test it by a developer having a selectable affects control panel. Maybe you'd either disable some effects to lighten the load on your processor whilst you look at new effects you've imagined, or maybe you'd accept 3 fps while your examining your visuals at an exact point in a game (a command to toggle these effects on and off - no matter what speed your current h/w gives).

Rather than just code Dawn Fairy or Vulcaan or a monkey or bear - divert your talent to releasing and ever growing 3d shader library of special effects for all game developers?

Just my wish to getting to Nirvana sooner. I'm on my 20 video card and its probably only the Voodoo 1 that full delivered what it originally promised. Since then h/w precedes s/w by 2 development cycles. Maybe its creeping closer to one generation cycle with advanced games and a plethora of DX9 capable cards on the market.

 

[micron]

g_day...nice to see you at THG

 

[nobie]

IMO, there won't even be games pushing the current generation of cards very far when the new generation ships. The current cards should be perfectly capable well into 2004...