Nv30 / R300 - The mighty are totally dependent on the weak!

[g__day]

A curious thing is going to happen over the next 9-15 months. The cream of the first DX 9 cards have appeared with as usual absolutely no software written for their new features and capabilities. But remember these cards will run all games written for 2-3 generations old video cards quite well.

But when will games written for the high-end cards appear? The answer is simply when there is enough of a user base to justify their development costs. This user base will probably consist of 5% of high end bleeding edge video card adopters and 95% from OEM or value purchasers who buy the value NV3x or Radeon cut down cards.

Think about that - unless the value derivatives of the leading edge cards take off the leading edge cards themselves are doomed to obscurity - they are white elephants.

Amazing but obivious. I feel the pace of technology will push on regardless - always. But if you get a bit or alot of value out of say NV30 going forward will be almost entirely dependent on NV31 and NV34 etc being a huge success.

So as the title says, "The mighty are so dependent on the success of the weak".

 

[Dreadnought]

Thats all pretty true - but we've been here b4. The GF4 MX has significantly retarded the development of games using DX8 capabilities.

However, if Nv don't repeat the same mistake with the FX/MX we could (considering the mainstream ATI parts) see far more widespread DX9 hardware - than we saw with the changes to DX8/8.1

 

[MuFu]

What you don't tend to see so much in the PC market, as opposed to that of consoles, is games selling graphics hardware. For example, a lot of people buy PSX for Gran Turismo and many buy XBox for Halo. I guess we're still waiting for the day when hardware upgrades will surge forward riding on the back of titles that fully exploit their features. While logistically it sounds quite far-fetched, I suppose the recent pushes involving Cg and RenderMonkey are a big step towards such a phenomenon.

Maybe GPU companies should allocate resources to their own small teams developing games that can work on a title with their hardware in mind from the word go*. I'd love to see "launch titles" for graphics cards.

MuFu.

*of course allowing compatibility with other hardware. Crap... I just thought of 3Dfx and Glide! That's inevitably where I am going to end up with this argument, doh...

 

[Dreadnought]

What you don't tend to see so much in the PC market, as opposed to that of consoles, is games selling graphics hardware.

I don't think thats true in the general sense at all - Pick a good number of major games and you can see very visible signs on forums and in certain niche hardware sales of spikes related to release dates. Thats been the case since at least the Q1/voodoo pairing..

 

[Nagorak]

Thats all pretty true - but we've been here b4. The GF4 MX has significantly retarded the development of games using DX8 capabilities.

However, if Nv don't repeat the same mistake with the FX/MX we could (considering the mainstream ATI parts) see far more widespread DX9 hardware - than we saw with the changes to DX8/8.1

The thing is Nvidia doesn't care about the development of 3D graphics (nor do ATi for that matter), they just care about making money. You say the GF MX was a mistake, but for them it wasn't. It's cheap, it makes them a lot of money, and if it holds back 3D graphics, what do they care?

 

[demalion]

Isn't this basically the same thing as your previous thread, g__day?

In any case, it inspires me to the same reply:

demalion[/url]]
Hmm...pixel and vertex shaders are a major change in the way of doing things. T&L on card was a change before that (less of one I'd say). Enhanced precision is not...it is just a quality enhancement of the existing way of doing things.

That is not to say that the difference between the two isn't significant, but that significance in is what the hardware is capable of, not what the developer is required to do to achieve it. So, as far as just that aspect of DX PS AND VS 2.0, I don't think we'd necessarily have as much delay as we've had for PS 1.x and VS 1.x.

 

[Snyder]

What you don't tend to see so much in the PC market, as opposed to that of consoles, is games selling graphics hardware. For example, a lot of people buy PSX for Gran Turismo and many buy XBox for Halo. I guess we're still waiting for the day when hardware upgrades will surge forward riding on the back of titles that fully exploit their features. While logistically it sounds quite far-fetched, I suppose the recent pushes involving Cg and RenderMonkey are a big step towards such a phenomenon.

Haven't we had that already? Remeber Origin...

 

[demonic]

What you don't tend to see so much in the PC market, as opposed to that of consoles, is games selling graphics hardware. For example, a lot of people buy PSX for Gran Turismo and many buy XBox for Halo. I guess we're still waiting for the day when hardware upgrades will surge forward riding on the back of titles that fully exploit their features.

Perphaps NV should get together with ID and bundle D3 with GFFX.

 

[g__day]

demalion:

I am impressed with your memory! Its a spin off from that post. The original post was about a lack of software for the high end cards; six months on and nothing at all has changed! This post is about the dependency between high and low end cards as being the key driver for next generation software development. And if the low end card fails to gain an enormous user based then everything is stuffed.

Your post reply is basically saying shaders V2 are easy to implement if you have previously done a 1.0 version. Trouble is not too many folks have developed shaders of any version it seems.

So the long wait is on.

 

[dksuiko]

But when will games written for the high-end cards appear? The answer is simply when there is enough of a user base to justify their development costs.

What's new? This has been the trend for the past few years.

 

[Dreadnought]

You say the GF MX was a mistake, but for them it wasn't. It's cheap, it makes them a lot of money, and if it holds back 3D graphics, what do they care?

Obviously I meant 'mistake' from the games developers & users perspective, I appreciate not for Nvidia. Indeed to take that a stage further I wonder how many ppl noticed the still excellant return Nv made from the TNT2 in the last (Nov7) Nv earnings call.

Still, I feel the MX was the result of the weak competiton. The 4200 release shows that Nv is perfectly prepared to offer better cards when its needed to counter ATI. Oh course I could be wrong, but Nv would seem to have little room considering the strong top-bottom range from ATI to produce a quite so crippled a MX version this time around.

 

[demalion]

....

Your post reply is basically saying shaders V2 are easy to implement if you have previously done a 1.0 version. Trouble is not too many folks have developed shaders of any version it seems.

I'm saying that implementing support for DX 9 can go hand in hand with implementing support for DX 8. In fact, if the tools execute well enough and soon enough (HLSL), they can be the same thing. Not an earth-shattering revelation, but I think it important to point out because I think this will allow adoption to happen rather rapidly compared to DX 8 adoption.

Some benefits can be had about as easily or maybe even more easily (or atleast it seems to me, depending on the graphics engine) as utilizing n-patches/TRUFORM (specifying floating point precision for buffers, 10 bit per component color output), while others (full range dynamic lighting calculations, longer shaders than can be effictively done with DX 8 ) will perhaps take longer as it might require more fundamental changes.

So the long wait is on.

While the wait for DX 9's official release will likely slow this down, I really think we'll see some benefit pretty quickly.