Intresting...

[Teasy]

Anyway, if you remember my arguments, I stated that the Radeon 8500's PS 1.4 wasn't really much of an advantage for consumers because it wasn't going to be used much, and when it was used, it would almost invariably be for a performance increase (and a performance increase that wouldn't bring it past the GeForce4....). If you turn my words around, you'll see that I was stating pretty clearly that the GeForce3's programmability was limiting the advancement of pixel shaders.

If you didn't put such a negative spin on ATI tech and such a possitive spin on Nvidia tech then people wouldn't have to turn your words around.

The fact is that when Nvidia were holding back the 3d industry you spin it one way by basically saying:

"Nvidia don't support this new feature, but because they don't support it won't be used much, therefore the feature is not an advantage for ATI"

And when ATI do the same you basically say:

"ATI don't support a feature that Nvidia will support, because of this the feature won't be used much, so ATI are holding back the 3d marked, bad, bad ATI"

 

[Simon F]

Teasy (et al),
It's best just to let it go. To paraphrase a well known proverb:
"Empty vessels make the worst signal to noise ratio"

 

[Teasy]

Simon F, I've never heard of that proverb before, did you just make that up?

BTW that was my first post in this thread, so I don't know why you singled me out there...

 

[Mariner]

He's probably worried that you were going to start betting other vital body parts!

 

[Guest]

Yes, I'm fully aware of flow control in the vertex shader portion. The NV30 will certainly have this as well (flow control in the vertex shader portion, at least, is in the Cg white papers...sorry about the misinformation previously that flow control was also there for the PS...).

Can I ask you how you know what nvidia hardware in the future will be? On more then one occation you have claimed how much more powerful the NV30 will be and seem to know intimate details about it. Or are you just assuming/pumping about how your favorite company will respond? You seem to have insider info on future nvidia hardware .... in other words I have doubts about the credibility of your claims with regards to hardware that is at best in its alpha stage. In which case nvidia should regard you as some sort of potential security issue. Otherwise you are just full of it and I am getting tired of reading your crap.

You pump nvidia every chance you get and take cheap shots at ATI every chance you can. Bla, if you are trying to convince people about how superior nvidia is then you are sorely mistaken in your endeavor as you have failed miserably in my case.

 

[Teasy]

He's probably worried that you were going to start betting other vital body parts!

Not at this stage in the thread... maybe later

 

[Reverend]

It has come to a point where words fail to describe what I think of Chalnoth so I won't except to say that NVNews should be embarrased. His flow control and 8x1 vs 4x2 "points" are absurb.

Lack of flow control isn't a limiting factor, based on what we know so far about the R300. The CPU remains the best place to execute really complex flow-dependent algorithms. The vertex shaders are best suited for more straightforward per-vertex operations. The one practical area where flow control might be useful is for iterating over per-bone influences when rendering a skeletal animated mesh, but it's really questionable whether the overhead over flow control will be offset by the reduction in workload. Long term, it may be better to have more flow control possibilities, but we need to keep in mind that the brute-force predictability of a GPU is the reason it's radically faster for rendering than a CPU. Lack of flow control in a GPU is not hindering the progress of 3D, not at all.

The 8x1 vs 4x2 distinction is invisible to software developers; its only measurable consequence is the performance of the hardware, and looking at the R300's performance, what possible complaints can anyone have?

 

[RussSchultz]

Its too expensive for me to convince my wife that I should get one.

 

[alexsok]

GOD, this thread has turned into something rather unpleasant...

My opinion about this is that ATI has delievered! Their R300 chip is definetly their best chip to date and the biggest advancement anyone ever made (if you look at all the previous generation of chips from both ATI and NVIDIA, you'll see that I'm right).

I admit, I prefer NVIDIA over ATI, but I'm not a fan of either one of them! Advancements are all I'm intrested about, as I'm sure most of us are!

NV30 will be announced 4 months after R300, so it's obvious it will be somehow better than R300 (maybe it's flow control in PS, unlimited instructions in PS or something else).

For now, R300 is the best hardware avaiable right now, so it's just plain silly to talk about NV30 beating the hell out of R300, when R300 is here & now and NV30 is still months away.

This chip exceeded all of my expectation, both from the technology standpoint & ATI in general.

Also, as many people noted here already, these features will take sometime to implement in games and even if NV30 will have flow control in ps pipeline and unlimited instructions, it will be a standard by the time developers begin to implement it.

I stress again, R300 has delievered, ATI has the best hardware right now, they did it BEFORE NVIDIA, so I doubt that "the biggest advancement in 10 years" will have anything sagnificant that ATI R300 doesn't have, NVIDIA are in trouble.

Both Companies lead the market from all standpoints, they both RULE, so as far as I'm concerned, i don't care who will come out with flow control in ps or unlimited instructions or real time raytracing or real time radiosity lighting first. The most important thing is for these advancements to be made in the first place!

 

[Sharkfood]

I think it's funny how a product that will be available in the next month or so is being spuriously labeled "inferior" to a product that will be out just in time for Christmas.

How come I didnt see the same kinds of rumor mongering of unknown featuresets for the R300 in all the GF4 reviews?

 

[OpenGL guy]

if NV30 will have flow control in ps pipeline and unlimited instructions, it will be a standard by the time developers begin to implement it.

Personally, I hope NV30 supports unlimited instructions: Then you won't see a single pixel and everyone will buy R300s instead!

 

[fresh]

Nobody gives a rat's ass about flow control in pixel shaders. It's not even part of the DX9 spec. You can do most "flow control" function in a pixel shader with a texture lookup. You wanna talk about limitations, talk about how the GeForce series implements clip planes! It uses a multitexture + texgen HACK to generate a clip plane using a 2x2 alpha texture. So you can't have a user defined clip plane AND multitexturing at once. What a hack!

The R300 kicks ass, plain and simple. Can't wait to receive my board and play with the new features

 

[KimB]

Can I ask you how you know what nvidia hardware in the future will be? On more then one occation you have claimed how much more powerful the NV30 will be and seem to know intimate details about it. Or are you just assuming/pumping about how your favorite company will respond? You seem to have insider info on future nvidia hardware .... in other words I have doubts about the credibility of your claims with regards to hardware that is at best in its alpha stage. In which case nvidia should regard you as some sort of potential security issue. Otherwise you are just full of it and I am getting tired of reading your crap.

You pump nvidia every chance you get and take cheap shots at ATI every chance you can. Bla, if you are trying to convince people about how superior nvidia is then you are sorely mistaken in your endeavor as you have failed miserably in my case.

I have already posted about all the evidence that I have. What I say here has nothing to do with my association with nvnews, and you can be certain that anything I state in the future at that website will be much more guarded than what I state here (i.e. I tend to post a lot more speculation here than I ever would in any sort of official article...much of what I say here, and on other forums, is just musings that I haven't given much thought to, etc.). Here, I do not represent nvnews in any way, shape, or form. Nor do I ever intend to. As far as I'm concerned, I'm here at the forums to learn more about 3D, and to express my personal feelings.

Anyway, I am sorry about the misinformation on nVidia's use of flow control in the pixel pipeline. Given that there is no mention of it in the Cg specs, there's not much reason to believe that the NV30 will have any. As far as unlimited-size programs are concerned, there is evidence that the NV30 will not have them. Namely, remember JC's comment that auto-multipass is essential? I doubt he'd say that if he knew that hardware would be coming out soon that would make it pointless.

As a side note, if auto-multipass in HLSL's is ever achieved, there is a very real possibility of limited flow control in any pixel shader.

I suppose it is possible that I have been too hard on the R300. Perhaps my hopes were just too high, but I will tell you that I'm pretty certain that this is the direction that technology is heading (though I'm also certain that I have no idea about a number of other future 3D technologies...).

 

[Ascended Saiyan]

The only gripe that I have with ATI & this card is that they won't have any support for Win 98 & I'm betting that all next generation cards will be the same.

 

[Pete]

Yeah, the trouble with (I think it was) FSAA'ing 16-bit games is worrying. They'd better make sure they can run people's old games as well as the new stuff. In fact, compatability with old games may be the biggest plus for unified drivers.

 

[KimB]

Well, *hopefully* ATI will have such compatibility issues with old games solved in the month or so that they have until release. It's not really worth talking about driver problems until the product is actually released.

Still, ATI has never given me much reason to have confidence in their ability to solve driver problems in a timely and effective fashion.

 

[Fred]

Yea the only thing thats holding me back from preordering a R300 is stability issues (and the fact that evidently 98 is not supported). Call it the 'annoyance' factor which goes into tweaking drivers/dling patches to get older games to work.

I'll give it a month or so to see how things turn out, before biting the proverbial bullet.

 

[OpenGL guy]

The only gripe that I have with ATI & this card is that they won't have any support for Win 98 & I'm betting that all next generation cards will be the same.

This is not an official company statement, but I use Win 98 just fine for testing I believe the Win Me drivers will work ok on Win 98. I can't say if the same is true for future DirectX versions though.

Edit: I just noticed I didn't mention the Win Me drivers, sorry.

 

[phynicle]

whu da hell runs the currently most powerful next generation graphics card on windows 98?

 

[KimB]

To be fair, most D3D games still run best in Windows 98, and I have a couple of games that don't want to run properly under Windows 2000.

Still, I won't go back to Win98...at least Win2k is somewhat stable...