[FUDie]

Originally Posted by radar1200gs

He was asking about the partial precision hint. That is clearly defined in the DX9 specification and has nothing to do with internal precision.

Are you trying for a record on how many times you can be wrong in a single thread? "Partial precision" refers to _pp modifiers on pixel shader instructions. Of course this relates to internal precision. Why else would the spec say that partial precision must be at least FP16?

Good grief!

-FUDie

[MuFu]

Originally Posted by radar1200gs

He was asking about the partial precision hint. That is clearly defined in the DX9 specification and has nothing to do with internal precision.

Apart from the fact that it does, of course.

Edit - oop, beaten to it.

[mozmo]

I don't buy this whole driver cp is too complex thing, for christ sakes how hard is it put a checkbox to enable/disable trilinear optimisations? It it confusing? No. Will it clutter up the cp? Hell no. For christ sakes they add that silly shader effect option to the cp and they're concerned about cluttering up the cp and confusing users. Fact is a large number of people use 3rd part cp tools like ATI tool, rage3d tweak because the cp lacks options. High end users crave options, we want to be able to tweak everything.

It's fairly obvious they're dodging full trilinear options so that they're products look faster, again it's all just deceptive marketing so that they win that performance crown. Unfortunately success in the computer industry is all about market perception, ie who is percieved to have superior products. A very small number will be able to afford these products that claim tech superiority. The rest of market settles with the crippled slow cards, but be swayed because the company xyz has the highest performing high end product.

[radar1200gs]

precision is precision. Internal or external as is bandwidth. It all adds up.

Pixelshaders require bandwidth also. Using the _PP hint can reduce that bandwidth.

We have heard plenty of coders who post in this forum state that they have to be careful when coding not to run into the limits of R3xx's shading bandwidth.

[Tim Murray]

...

no. it can't be.

I think he thinks that the difference between FP16, FP24, and FP32 is the color depth at which the things are rendered. hell, reading his last several posts, it's the only explanation that makes ANY sort of sense. (okay, I know, I shouldn't try to make sense out of his posts, it's bad for me, but...)

[radar1200gs]

The difference is how much data has to be moved around the chip at any one time. FP16 halves FP32's requirements in that regard and allows for more register space which can assist scheduling.

[FUDie]

Originally Posted by radar1200gs

precision is precision. Internal or external as is bandwidth. It all adds up.

Pixelshaders require bandwidth also. Using the _PP hint can reduce that bandwidth.

No, shaders require no bandwidth whatsoever. What the _pp hint can do is reduce internal storage. This has nothing to do with bandwidth but how many pixels the shader unit can have in flight. Internal bandwidth from the shader to this storage is "free" since the thing should be designed to run at full speed. In fact, if you are using _pp, you are essentially wasting this "bandwidth" since you will have data lines that are not being utilized. Judging by experiments done on NV3x, the storage can be accessed at full speed in either full or partial precision mode (shaders with few temporaries ran fast), but the number of pixels in flight was reduced in full precision mode due to a lack of storage.

Texture lookups use bandwidth, but the _pp modifier will have absolutely no effect on this bandwidth because the data format of the texture is what determines how much bandwidth it requires.

We have heard plenty of coders who post in this forum state that they have to be careful when coding not to run into the limits of R3xx's shading bandwidth.

Never saw this mentioned by anyone. Some developers said they ran into R3xx's shader instruction limits but that has nothing to do with bandwidth, but how long shader programs can be.

Why do you insist on defending a position which is indefensible?

-FUDie

[jvd]

Originally Posted by radar1200gs

precision is precision. Internal or external as is bandwidth. It all adds up.

Pixelshaders require bandwidth also. Using the _PP hint can reduce that bandwidth.

We have heard plenty of coders who post in this forum state that they have to be careful when coding not to run into the limits of R3xx's shading bandwidth.

well for bandwidths sake i guess ati should just force bilinear all the time and get rid of af ?

If the dev asks for 32bit fp then they should get it. Its not up to nvidia to choose when a dev gets what he asks for .

[FUDie]

Originally Posted by radar1200gs

The difference is how much data has to be moved around the chip at any one time. FP16 halves FP32's requirements in that regard and allows for more register space which can assist scheduling.

This bandwidth is free. End of discussion. Register space is something completely different. Your whole premise is flawed.

-FUDie

[radar1200gs]

Originally Posted by jvd

well for bandwidths sake i guess ati should just force bilinear all the time and get rid of af ?

If the dev asks for 32bit fp then they should get it. Its not up to nvidia to choose when a dev gets what he asks for .

Since when has nVidia recently forced precision changes upon us in drivers? That hasn't happened in a long time. Get over the past.

In contrast ATi is forcing AF & Trilinear optimizations upon you whether you want them or not.

[radar1200gs]

Originally Posted by FUDie

This bandwidth is free. End of discussion. Register space is something completely different. Your whole premise is flawed.

-FUDie

Nothing is ever free. Especially not register and memory bandwidth.

[MuFu]

Originally Posted by radar1200gs

Nothing is ever free. Especially not register and memory bandwidth.

But you were talking about "shader bandwidth" - something that isn't even a recognised term because it's a total non-issue (i.e. free).

[jvd]

Originally Posted by radar1200gs

Since when has nVidia recently forced precision changes upon us in drivers? That hasn't happened in a long time. Get over the past.

In contrast ATi is forcing AF & Trilinear optimizations upon you whether you want them or not.

have you played farcry recently ?

Not to mention how much else they do that we can't notice or haven't noticed .

[MuFu]

Originally Posted by beyondhelp

and WHY, if the 6800 is so Fine, is Nvidia still using _pp hints with it? I thought it was this all powerfull Full 32bit precision powerhouse -lol-

It can still benefit from the use of FP16 - decreased reg usage limitations (no way near as severe as NV3x though) and free normalisation. One would hope that the _pp hint is now being used in situations where it does not constitute "cheating".

[whql]

Originally Posted by radar1200gs

In contrast ATi is forcing AF & Trilinear optimizations upon you whether you want them or not.

So is nvidia. On they boards they have sold in the past 14 months they are only not forcing it for the 10 or retail retail 6800's that have been sold, with one driver release so far. Thats hardly a commitment.

[Pete]

Originally Posted by DaveBaumann

However, if people can spot the differences and they are highlighted hopefully they will recant on this.

Most 9600 or X800 owners won't have a chance to spot the differences (if there are any) without a software toggle for trylinear. Or does ATi mean for "people" to run two PCs side by side with different video cards, or to swap cards in the same rig? Screenshots can't fully capture how visible mipmap bands are in motion, so does ATi really expect ppl to create and compare videos of try and tri to spot the difference?

[FUDie]

Originally Posted by radar1200gs

Nothing is ever free. Especially not register and memory bandwidth.

Yes, it is free because of the chip's design. Register bandwidth is free, this has been proven even on the NV3x. Register storage is not free, but that's not what you were saying.

At the beginning you claimed that partial precision doesn't refer to internal precision, which it, of course, does. Then you said that pixel shaders use bandwidth, which they don't. Then you said internal storage is reduced, which is true, but has nothing to do with what you started off as saying.

You can't use one correct statement to justify a whole host of incorrect ones.

The NV3x is one example we have that benefits from partial precision. Is it because of "reduced internal bandwidth"? No. It's strictly because the number of pixels in flight is increased. The chip has enough internal data lines to support FP32 at full speed, this has been shown. What it doesn't have is a large enough internal register file to keep enough pixels in flight to hide all the latency when FP32 is used.

-FUDie

[FUDie]

Originally Posted by whql

So is nvidia. On they boards they have sold in the past 14 months they are only not forcing it for the 10 or retail retail 6800's that have been sold, with one driver release so far. Thats hardly a commitment.

And let's not forget the "special" 61.11 driver where the option to disable trilinear optimizations was "broken".

-FUDie

[cthellis42]

Originally Posted by jvd

I'm still wiating for problem cases though. People keep saying they see them but never post any proof. They just ignore me asking for and offering to host thier proof .

Kinda funny huh

Sometimes they also point to an article which shows the mathematical difference or made up software so YOU can see the mathematical differences, with a conclusion of "how can it NOT cause serious IQ problems?"

Oh, and sometimes they post pics with differences far smaller than the usual ones between differing techniques ATi, nVidia, and others always have in games with a bit "AHA!"

Hopefully that stuff will come in time, looking at a broad scale of games and situations. Regardless, I do think ATi should have made sure to keep people appraised of their "advanced trilinear algorithm(s)" and kept it optional, so there wouldn't be fuzziness on the situation and people would still be happy to "get what they want" while they try out something else for themselves. Heck, they'd get a lot more case testing done that way to refine the process, and gamers would judge at their leasure. There will be SOME differences, and each individual will react differently to them. (Though I imagine the vast majority would notice nothing at all.)

Otherwise, though, it's much like adaptive AF or getting stuck with only certain AA configurations where you prefer other methods--just something you have to live with.

Originally Posted by FUDie

Why do you insist on defending a position which is indefensible?

When you say things enough times, it's true!

[radar1200gs]

Originally Posted by whql

So is nvidia. On they boards they have sold in the past 14 months they are only not forcing it for the 10 or retail retail 6800's that have been sold, with one driver release so far. Thats hardly a commitment.

It's more of a commitment than ATi has made so far, and ATi has been optimizing since the 9600 was released.

[jvd]

Originally Posted by radar1200gs

It's more of a commitment than ATi has made so far, and ATi has been optimizing since the 9600 was released.

nvidia has been optimizing since the fx series was released !!!


Not only that but in the almighty 6800ultra they put in af optimizations !!!!!

Oh no radar what are you going to do. Your god has been doing it for almost 2 years !!!!!!

Oh wait your a troll so it wont matter to you .


Good day . I said good day

[radar1200gs]

Originally Posted by FUDie

And let's not forget the "special" 61.11 driver where the option to disable trilinear optimizations was "broken".

-FUDie

They are beta drivers. Wait for the official release, then crtiicize.

[cthellis42]

Originally Posted by radar1200gs

It's more of a commitment than ATi has made so far, and ATi has been optimizing since the 9600 was released.

...and nobody noticed. Though brilinear was noticed immediately, complained about immediately, and those concerns ignored stoicly. (And then it was applied universally to get around one of their own internal optimization guidelines! Whee!)

[Smurfie]

1 checkbox - "Disable all optimizations."

How difficult is that? Not much clutter, puts forth tons of goodwill. I used to hate Nvidia for their stupid brilinear optimization, but this time, I'm going with the 6800U because they give users the option to turn it off.

[Smurfie]

Originally Posted by jvd

nvidia has been optimizing since the fx series was released !!!

Now, I do kind of remember that ATI first optimized AF with the 8500. Angle dependent AF is optimization in my book, since my graphics reference books paint a different picture in the results of AF.

Nvidia going down the same route with adaptive AF is a step backwards, but here's to hoping we can expose their non angle dependent AF in their drivers again.