Why can't NVIDIA or ATI finance a game specially for their

[beginner16]

hi

I don't know a first thing about graphic programming , but I was wondering why can't Nvidia or Ati finance the making of game developed specially for their latest and most advanced product?
Perhaps they would spend few millions on it,but once people see that there is an actual game with the same graphic quality as one in demos that nvidia and ati are so keen on displaying,I'm shure they'd be standing in lines to buy their lates card and be able to actually play such a game

I realise very well that it's up to programmers to make such a game engine and with it a game, but game programmers will do things that are in their own best interest,which in this case is backward compatibillity with older cards

In the end both of these two companies are very well aware that quallity of graphic displayed in their demos won't be available to the end user for many moons to come

Perhaps naive of me,but after ordinary people made both of these tow companies rich it's perhaps time these two should give something in return,and enable end user to enjoy game with graphics comparable to ...and as I said before,I'm shure it would benefit them also

 

[trinibwoy]

Re: Why can't NVIDIA or ATI finance a game specially for the

why can't Nvidia or Ati finance the making of game developed specially for their latest and most advanced product?

Haven't you heard about Doom3 and HL2?

 

[MasterBaiter]

Re: Why can't NVIDIA or ATI finance a game specially for the

why can't Nvidia or Ati finance the making of game developed specially for their latest and most advanced product?

Haven't you heard about Doom3 and HL2?

*rimshot*

I would guess because they are not in the business of making games? Besides, what better way to get your name out there than funnel some money (+ free hardware/tech support) to game devs to help implement features that benefit your customers, and then get some advertisements for your product in a nice flashy screen just before the game starts up. The fact that most customers are seeing some sort of such screen on most 3D accelerated games @ this point seems to me that they really don't have to bother making anything more than a tech-demo for their product.

 

[Socos]

Well in many ways they do that already, however, the investment in a real game is huge. Much more than the few million you reference. I believe HL2 was upward of $50 million. And if you are looking at basically cutting the number of people you can sell it too in half by limiting the hardware it can run on it is not really a good business proposition.
They develop demo's and I am sure that is not cheap either. But to make a game with all the latest bells and whistles that is a no go.

 

[MrGaribaldi]

And if you are looking at basically cutting the number of people you can sell it too in half by limiting the hardware it can run on it is not really a good business proposition.
They develop demo's and I am sure that is not cheap either. But to make a game with all the latest bells and whistles that is a no go.

The demo's they are creating are pushing the hardware as far as it can go., so I'm not sure you could expect a game with similar detail.
Can you imagine game with the same level of detail on every character as Nalu, good textures etc., which would run with more than a single digit fps on the latest hardware? And then you add in physics, AI and the rest.

So I agree that they should stick with the demos that wow's us, and rather help the game devs.


Besides, I'm getting annoyed enough as it is with games having "special" content for different cards (Tiger Woods? NWN?)....

 

[Laa-Yosh]

Well in many ways they do that already, however, the investment in a real game is huge. Much more than the few million you reference. I believe HL2 was upward of $50 million.

Half-life 2 is perhaps the most extreme game in terms of budget... and as such, a very bad example. Most games still cost "only" a few million dollars, like 2-5; and most Eastern European developers have to work from even smaller budgets.

 

[no_way]

Can you imagine game with the same level of detail on every character as Nalu, good textures etc., which would run with more than a single digit fps on the latest hardware?

Plus, at any point in time, there isnt much of the "latest hardware" on the market, so your initial userbase would be very very small.

But indeed, pushing better middleware for more efficient rendering is what they ought to be doing and are doing with their SDKs, toolkits, whitepapers and utilities.

 

[3dilettante]

How about that game Evolva? It had a demo that came with some vendors' Geforce 256s. I think the full version came out a while later.

I played the demo, it wasn't particularly interesting. It touted the use of T&L, but overall I wasn't too impressed. The utility of funding an entire game solely to demonstrate a card is proportional to the number of people who remember games like Evolva.

There's no money to be made creating a game that can only be run on a single vendor's newly introduced card.

 

[GwymWeepa]

I made a similar thread a few weeks ago and people here shit all over it lol. The type of game I had in mind was some simple, mind-less space shooter that wouldn't require too much programming on the AI and physics side, it would mostly be a graphically intense arcade game that could evolve and expand with each card release (new levels, missions, ships etc), when totally new architectures were introduced, the whole game could have its graphics updated with new effects. I still think its a good idea, but what it boils down to is...they just don't have the time and resources to spend on such a project, better to leave it to the game producers.

 

[beginner16]

I played the demo, it wasn't particularly interesting. It touted the use of T&L, but overall I wasn't too impressed. The utility of funding an entire game solely to demonstrate a card is proportional to the number of people who remember games like Evolva.

But the point is if a game would be looking as good as one of their demos it would such a huge leap forward in terms of graphic display that lots of people would be willing to pay for those cards.
Shure if they were to make a game that is only slightly better looking than todays games it would go un-noticed(like evolva did),but ...

Was evolva such a leap forward in terms of visual quality?!I didn't think so

Gotta say...I think games still look kinda shitty.
Perhaps objects cast shadows in more real world like manner,but the environment and characters still look like they've been made out of cardboards.Nothing life-like about them.Textures are still shitty etc.If you pick up PC-gamer from few years back and look at in-game pictures,you can hardly see any difference between then and now.Yes,when you actually play the game the movements and such are more realistic,but textures are still crap and only slightly better then they were 3 years ago


Them bastards, and that includes programmers and guys at nvidia and ati

 

[bloodbob]

Nvidia I know in the past has and the sales were rather poor.

 

[3dilettante]

I played the demo, it wasn't particularly interesting. It touted the use of T&L, but overall I wasn't too impressed. The utility of funding an entire game solely to demonstrate a card is proportional to the number of people who remember games like Evolva.

But the point is if a game would be looking as good as one of their demos it would such a huge leap forward in terms of graphic display that lots of people would be willing to pay for those cards.
Shure if they were to make a game that is only slightly better looking than todays games it would go un-noticed(like evolva did),but ...

Was evolva such a leap forward in terms of visual quality?!I didn't think so

Gotta say...I think games still look kinda shitty.
Perhaps objects cast shadows in more real world like manner,but the environment and characters still look like they've been made out of cardboards.Nothing life-like about them.Textures are still shitty etc.If you pick up PC-gamer from few years back and look at in-game pictures,you can hardly see any difference between then and now.Yes,when you actually play the game the movements and such are more realistic,but textures are still crap and only slightly better then they were 3 years ago


Them bastards, and that includes programmers and guys at nvidia and ati

Evolva's quality was pretty poor in relation to the demos because all of a sudden they had to worry about a game to go with the pretty visuals.

No game is going to look like Nalu or Ruby throughout on the hardware their demos were designed to advertise.

I'm not sure some games devote as many polygons and shaders to the entire screen as are devoted to either the aquatic or futuristic vixens' cleavage. Well maybe the entire screen, but that's probably pushing it.

If there's going to be any consistency to the visual quality, something has to be compromised to keep acceptable performance.

 

[beginner16]

Evolva's quality was pretty poor in relation to the demos because all of a sudden they had to worry about a game to go with the pretty visuals.

No game is going to look like Nalu or Ruby throughout on the hardware their demos were designed to advertise.

I'm not sure some games devote as many polygons and shaders to the entire screen as are devoted to either the aquatic or futuristic vixens' cleavage. Well maybe the entire screen, but that's probably pushing it.

If there's going to be any consistency to the visual quality, something has to be compromised to keep acceptable performance.

But if even demos for nvidia's cards (geforce 3 I think it was called ) of three years ago look 10 times better than current games,then the customers are getting cheated. And as I've said before , when opening PC-gamer from three years ago and looking at in-game pictures,you can hardly see the difference between games of yesterday and those of today.
Sometimes I can hardly believe when I read threads in this forum...while hardware power in the last 4 years had increased by the factor of 10+,you can hardly see that in games.And yet , noone in this forum complains about it ( this wouldn't be so strange if not for a fact that most of you are experts in this field )

 

[Himself]

I like the idea, doesn't have to be a full game, but say a one level FPS or something, wouldn't be too far off a demo, but you'd get more of a feel for what the maximum you could expect from the hardware would be in a bit more realistic terms.

 

[jb]

Keep in mind it takes a long long long time to make a game. First you have to make the engine and thats where are the new features of the card will come into play. Once you have a solid engine then you can go and make the game on-top of that. Granted you can do some things in parallel but the engine has to be in a semi-stable form before you can do most of your work. And with out some planning and talent how good of a game will it be? I mean we can make a DX9 Pong with all of the new features but its still just Pong.....

 

[KimB]

Keep in mind it takes a long long long time to make a game. First you have to make the engine and thats where are the new features of the card will come into play. Once you have a solid engine then you can go and make the game on-top of that. Granted you can do some things in parallel but the engine has to be in a semi-stable form before you can do most of your work. And with out some planning and talent how good of a game will it be? I mean we can make a DX9 Pong with all of the new features but its still just Pong.....

Well, for one, engines can be licensed. There are some modern engines, like the Unreal Engine 3, that are flexible enough to easily work with most new features that will become available in new hardware. It may actually be concievable, for instance, for nVidia to make a custom UT2k4 level (or, later on, a UE3-game level) that would showcase, say, more complex shaders made available by new hardware. It might be a bit more challenging to deal with overall rendering system changes, like what would be required for using a FP16 framebuffer, but nVidia may be able to work out a deal with Epic where they get to also ship a new renderer with the custom level (I think only people who license the engine typically get access to the renderer source these days....).

 

[jb]

Well, for one, engines can be licensed. There are some modern engines, like the Unreal Engine 3, that are flexible enough to easily work with most new features that will become available in new hardware.

Thats very true they could cut time down buy doing this. But depending on the engine it may or may not be easy. You can hack most anything in, but will it work well is another story.

It may actually be concievable, for instance, for nVidia to make a custom UT2k4 level (or, later on, a UE3-game level) that would showcase, say, more complex shaders made available by new hardware.

Also a very good possiblity. However the normal tools we have don't expose that kind of controll...at least what I can see of it. Still waiting for that section of the UDN to get unlocked

It might be a bit more challenging to deal with overall rendering system changes, like what would be required for using a FP16 framebuffer, but nVidia may be able to work out a deal with Epic where they get to also ship a new renderer with the custom level (I think only people who license the engine typically get access to the renderer source these days....).

Yeap I am sure they could if they want to pay. But at that point, does it become cheaper for NV (or any IHV) to just have the game developer just add the effect for them (ie like Farycry and PS3.0)? Speaking of cost, the UT2k3/4 engine goes for $500,000. How much more does someone want to spend on the engine and development time to make a game off that engine for a game? Games are not a cheap to develop as a tech demo...and if you want a good game it will take longer (ie more money).

To me it does not make any sense at all for the IHVs to do this. Instead they would be better off getting the developers to add thier features, save the money buy using the tech demos and let the game developer due what they do best. But if they want to release a custom map for a game, I am all for it...however that VIA Quake map needed some tweaking

 

[Sonic]

The reason pictures in PC Gamer magazines look the same as they did three years ago is that PC Gamer uses crap pictures to begin with.

There are companies that makes games for their own hardware. Microsoft did it a while back when it made a set top box with technology based off of PC's. I remember it had an Nvidia chip in and also a crippled Pentium 3. I think the name of it was Xbox but I can't be sure.

Basically, the economics aren't there for a graphics company to make games for their own chips. We can lee it at demos to show off the power of the cards. No real reason to make a game. If they started making games specifically for their chips they might as well enter the console business. But that won't happen since console gamers don't want to keep upgrading every 6 months. There really is no point to it.

 

[KimB]

Thats very true they could cut time down buy doing this. But depending on the engine it may or may not be easy. You can hack most anything in, but will it work well is another story.
...
Also a very good possiblity. However the normal tools we have don't expose that kind of controll...at least what I can see of it. Still waiting for that section of the UDN to get unlocked

Definitely. So in making the demo you'd simply use industry-standard tools or design your own and interface them with the game's editor (note that nVidia currently designs many developer tools). For example, the Unreal editor has, for a long time now, had the capability to import stuff from 3DSMax (and I think some other programs as well).

Yeap I am sure they could if they want to pay. But at that point, does it become cheaper for NV (or any IHV) to just have the game developer just add the effect for them (ie like Farycry and PS3.0)? Speaking of cost, the UT2k3/4 engine goes for $500,000.

This is the cost for a developer who desires to sell a retail game using the engine. I don't see why some special case couldn't be made here. That is to say, I doubt the demo would be sold, but rather bundled (note that I am still using the term "demo," but now more in the context of a game demo than a tech demo).

To me it does not make any sense at all for the IHVs to do this. Instead they would be better off getting the developers to add thier features, save the money buy using the tech demos and let the game developer due what they do best. But if they want to release a custom map for a game, I am all for it...however that VIA Quake map needed some tweaking

Which they certainly try to do. I'm just not sure it's always the most effective way to do things. For example, how long did it take Crytek to finally get the PS 3.0 and HDR stuff out for FarCry?

 

[AlNom]

There are companies that makes games for their own hardware. Microsoft did it a while back when it made a set top box with technology based off of PC's. I remember it had an Nvidia chip in and also a crippled Pentium 3. I think the name of it was Xbox but I can't be sure.

Hm... I think it was like one of those black boxes found in airplanes. Able to withstand gunfire and the sort too.

If they started making games specifically for their chips they might as well enter the console business. But that won't happen since console gamers don't want to keep upgrading every 6 months. There really is no point to it.

Bingo!

Right now, You've only got the entire R3xx R4xx, NV3x, NV4x chips to optimize for, and even then, there are some differences. And then there are the older lines of hardware.....

Not happening

There is probably a problem with load balancing too. If you have very few specific cards for the PC, you'd still have a wide range of CPU speeds... How do ensure the person is getting the best performance for their system as such?