The Rydermark Thread (TM)

off topic

trinibwoy said:
Nvidia has certainly earned their guilty until proven innocent reputation ...

Sigh. I'll never have Trini figured out.
Several months ago I was prepared to nominate you for most improved poster of the year (due to a dramatic mellowing of what previously were overly green tendencies). Then you seemed to have a relapse (your new X1900 didn't live up to your expectations or something?) And now this statement.. Ah, well. Can't peg certain people down, I guess.

:p

ERK
 
geo said:
Ach! My bad. :cry: I combined two eps there (there was one where he did something on a motorcycle in the parking lot at Al's).
He was going for the record rubbish can jump for a tv show. The 1st episode ended just prior to the jump attempt and the next started dealt with the aftermath.
 
  • Like
Reactions: Geo
IgnorancePersonified said:
He was going for the record rubbish can jump for a tv show. The 1st episode ended just prior to the jump attempt and the next started dealt with the aftermath.

Bingo! :D
 
ERK said:
Sigh. I'll never have Trini figured out.
Several months ago I was prepared to nominate you for most improved poster of the year (due to a dramatic mellowing of what previously were overly green tendencies). Then you seemed to have a relapse (your new X1900 didn't live up to your expectations or something?) And now this statement.. Ah, well. Can't peg certain people down, I guess.

:p

ERK

Rofl! I took that as a complement in case you were wondering :LOL: Thanks, you made my day! But if you were to take a step back you would notice that those overly green tendencies subsided in tandem with the withdrawal of those with correspondingly excessive red tendencies. When I first joined this place was a tad "unbalanced" and I like to be different :devilish: That in addition to the fact that I was an arrogant snotty nosed brat back in 04 and I am just a little less snotty nosed now. I blame it on the more "mature" folk around here (you know who you are :p )

And for the record the only thing I've been unhappy about with the X1900XT was the noise, which was promptly rectified with a Vf900-cu. I wouldn't have bought it if I thought there was a better card out there now would I ;)
 
Last edited by a moderator:
Sharkfood said:
I'd say that's a very biased perception of what was presented.

It explains that the makers of Rydermark benchmark have mentioned there is a forced lower precision occuring with NVIDIA hardware and/or drivers.

If you don't consider relaying information as "evidence".. then I'd expect the same harsh-line be applied in the future also be applied similarly. That was my point.
Relaying information like this is certainly not evidence. There are easy ways to prove this with a relative degree of certainty, and L'Inq did no such thing. Having a pretty significant accusation that could easily be verified and not doing so makes me approach this with serious skepticism.

This is the very double-standard that I strived to emphasize... this stance has changed dramatically.

The point being, a forum post by some anonymous source ("SlayahODeath") of liguistics attuned towards "(IHVC) is suxorz doodz!" can spur front-page, widespread website link attention.. comprehensive screenshot analysis, side-by-side or java applet tests, sample programs, code snippets and white-paper quality definitive resources.

A pretty specific accusation from a website, in reference to a benchmark, and with substantially more information from which to get a feel of the accusation is now getting the brush-off, character defamation attempts, and arguments that deflect away from the main accusation.
Eh? "Significantly more information?" Fudo said, "Rydermark guys say they can't get FP32 from NVIDIA cards. At all." Now, there are plenty of ways to test for this, but the claim is so outlandish and so easily noticed without any special tests that the burden of proof is on them. If somebody wants to find something with a Big Lighting Shader and compare it against an X1900, feel free, but I don't buy these accusations right now.

I'd start with the accusation that wasn't touched or disproved, but instead precisely the deflection tactics described above- what's the deal with FP32? Is FP16 forced, defaulting or incorrectly being applied when referencing/requesting FP32? And if so, under what conditions?

Yes, I agree the burden of proof lies on the accuser.. but this has *never* been the case here or anywhere else in the past.

A good change in ethic? Perhaps.. I'm just interested why the "good change" has suddenly occurred... and not too hopeful as I've only seen differing application of ethics in the past 10 years that vary only by the IHV in question.
Burden of proof has always been on the accuser except in questionable situations. E.g., NVIDIA 3DMark performance suddenly doubles with a single driver revision to a level that isn't really theoretically possible--yeah, I'd definitely say that's worth investigating. When Dave broke the UT2003 AF optimization thing, he did so with screenshots. ExtremeTech did the same with 3DMark. Nothing has changed here. We've got no evidence of widescale wrongdoing (like the article suggests), no bizarre benchmark results or IQ problems. Every other time you mentioned later in the thread, those conditions existed. So no, this isn't "Hey guys, we're a bunch of NV fanbois!" This is a "we're not going to sit here and take another ridiculous bit of rumormongering and speculation by the Inq seriously" thread. Having a domain name doesn't mean people are or should ever be listening to you--that's the case with Fudo. He simply doesn't have a clue.

(PS: you don't request FP32. You always get FP32 with ps_3_0 unless you specifically flag a shader as _pp. ps_2_0 is the same, except you can get FP24 as well.)

But fine! Let's take the damn thing seriously just for the sake of fucking argument, in the hopes that this puts it to bed once and for all.

Fudo said:
Nvidia doesn’t leave you any choice, it's claimed. You simply cannot turn 24 or 32 bit precision on, you are always locked at 16 bit. From a developer and artistic perspective this is really unacceptable but will buy you a few frames here and there.


Developers have also informed us that they have no way to stop Nvidia doing this. The only way is to make the community aware, and that can change some minds. There is more to come and we will try to get you screenshots to really see the difference. µ
Both paragraphs heavily imply ("no way to stop NVIDIA" and "always locked at 16 bit") that this is on in all applications. Let's call the hypothetical forcing of FP16 for all pixel shaders situation A. But, just to stretch the argument even further, let's look at a second possible situation, where NV devrel got their hands on a version of Rydermark and had the driver team implement cheats for it (to force FP16). This is situation B.

To confirm situation A, we can do one of two things. If we think it really is in all applications and isn't being overly sneaky, we can do a simple IQ comparison between a G71 and an R580 and look for artifacting resulting from rounding in the G71 shots. That takes all of... uhh... an hour or two. If we're really worried about NVIDIA being sneaky and not forcing FP16 with some shaders in order to preserve IQ (e.g., instead of "some shaders to use FP16" like in the days of yore, they have "specific shaders to use FP32" and everything else gets FP16), we can write a really basic GPGPU application and look for reduced precision there. That will take slightly longer--say, six hours, just for laughs--but it's still not a big deal, especially that there's so much GPGPU code out there today. And considering a great deal of GPGPU research is being done on NVIDIA hardware and GPGPU applications, as a whole, are incredibly concerned with precision, we probably would have heard about this from then a long time ago.

Situation B: We know how NVIDIA performs application-specific optimization/cheats/whatever-the-hell-you-want-to-call-them. I assume that Unwinder's statement back in... 2003? 2004? still holds true, and NVIDIA is performing a hash on the Direct3D window name and the filename in order to detect the application (and prevent another Antidetect from being written, and defeat Quack3.exe-style tricks). So, if you're a developer and you have the source code... uhhhh... change it and recompile, dumbass. Do you have FP32? If B was true in the first place, yeah, you would.

Of course, there's situation C: this only happens on NV3x hardware. In fact, I think it might force FP16 (or FX12 for NV3x<5), and I don't really have a problem with that. Nothing is going to run with FP32 because of the register penalties and such and still have any sort of reasonable performance, regardless of res, so the only people you're hurting are possibly GPGPU people. Then again, if you call up NV and tell them you're a GPGPU developer, they'll probably just send you new stuff. And you want DX10 anyway.

So yeah. In the absence of evidence, can this please go away. If you want to write your own tests, feel free. If it helps you sleep at night, go for it. If you prove me wrong, well, I'll happily eat my hat and call NVIDIA a bunch of bastards. I've just been given a ridiculous accusation (in terms of how broad it is) with no evidence; ergo, I have no reason to believe it.
 
The Baron said:
So yeah. In the absence of evidence, can this please go away. If you want to write your own tests, feel free. If it helps you sleep at night, go for it. If you prove me wrong, well, I'll happily eat my hat and call NVIDIA a bunch of bastards. I've just been given a ridiculous accusation (in terms of how broad it is) with no evidence; ergo, I have no reason to believe it.

I think we've reached the "Fudo, put up or shut up" stage on this one. This crowd would be happy to analyze his "put up" if he'll bring it forward.
 
Meh, I dunno how it took 5 pages :rolleyes:
I got there about half way through the first sentence of the article when I noticed there were no accusing screenshots.
Without any evidence the accusation is by definition baseless.

I vote Barons situation C or a variant including bits of B along the lines of 'the developers are dumbasses that have been using the wrong (NV3x) compile target'
 
Well, if Fudo's not just totally making this crap up, here are some possibilities of what I think happened:

1. Fudo confused FP32 framebuffer support with FP32 register storage. When the developer commented that the NV4x couldn't render to FP32 (true if you want to do blending), Fudo thought it'd be a great thing to write about.
2. The developer was lamenting nVidia's recommendations about using _pp as much as possible. He said something along the lines of, "you can't select FP32," (since it's the default) and Fudo got confused.
3. The developer stupidly had a global precision hint enabled, then started complaining about precision.
 
The Baron said:
(PS: you don't request FP32. You always get FP32 with ps_3_0 unless you specifically flag a shader as _pp.
Technically, this is an incorrect statement. I'd be mad as hell if I got 32-bits automagically if I never wanted or specified it.

The Baron said:
But fine! Let's take the damn thing seriously just for the sake of fucking argument
Come now, any argument regarded as serious doesn't deserve a F'ing. :)

Everyone is entitled to their opinions and express them (eg. I kinda understand Sharkfood's comments from a "influenced by past conduct" standpoint but I also agree with almost everything you posted).

This never started out as a discussion about technology but about company practices and I know how emotional these kinda topics can be for certain folks. I've checked and found that Sharkfood seems inclined to participate in threads where there seems to be some wrongdoing by one specific IHV; it indicates a certain non-preference (as opposed to preference) and IMO there's nothing really wrong with that but I think we should just not get overly worked up too soon.

Personally, I think this thread should be locked to prevent any further animosity, and be re-opened if and when The Inq (or someone else) provides more details. I repeat, this isn't about 3D technology or even hardware but about right or wrong practices or simply about suppositions.

Can we return to a topic about 3D, like when the hell will we ever see soft stencil shadows? :) :)
 
Last edited by a moderator:
Parousia said:
I've checked and found that Sharkfood seems inclined to participate in threads where there seems to be some wrongdoing by one specific IHV;
Yes. And not for what you feel it's indicative of, but the very point I was making.

When any *other* IHV is accused of wrong doing, it's always a 38-50 page thread accompanied by web site front page links, side-by-side screenshots, java applets, code snippets, sample test programs, IHV and developer interviews, and comprehensive lists of applications effected.
 
Parousia said:
Technically, this is an incorrect statement. I'd be mad as hell if I got 32-bits automagically if I never wanted or specified it.
Er, no, not in DirectX 9, nor in OpenGL. In SM3.0, if an assembly instruction does not have the _pp modifier, it must be done at 32-bit precision. In SM2.0, it must be done at at least 24-bit precision.

The _pp modifier, on the other hand, is merely a hint. The instruction must be carried out at greater than 16-bit precision.

One never specifically asks for any level of precision for 3D hardware. Only minimum precisions are specified.

Edit: The only time precision is explicit in D3D/OpenGL is for input/output formats (i.e. textures and rendertargets).
 
Sharkfood said:
When any *other* IHV is accused of wrong doing, it's always a 38-50 page thread accompanied by web site front page links, side-by-side screenshots, java applets, code snippets, sample test programs, IHV and developer interviews, and comprehensive lists of applications effected.
Oh, please. It's more like when there is any reasonable basis for an accusation of wrong doing, we see lengthly threads and whatnot.
 
Sharkfood said:
Yes. And not for what you feel it's indicative of, but the very point I was making.

When any *other* IHV is accused of wrong doing, it's always a 38-50 page thread accompanied by web site front page links, side-by-side screenshots, java applets, code snippets, sample test programs, IHV and developer interviews, and comprehensive lists of applications effected.

That is simply not true. Show us the discussions you talk about, links please.

One thing's for sure: they all get their asses kicked when there is something to look for in the first place :!:
 
I did the somewhat obvious thing and spent a couple of hours this morning testing existing (D3D9 only) shaders in games and authoring simple shaders of my own, analysing both fxc output and instruction streams output by the runtime, and some image-level checks, and there's no evidence whatsoever that there's any kind of global _pp force happening.

So unless the driver is ignoring the instruction modifiers, but also somehow using black magic to cover up somewhat obvious areas where you'd expect FP16 pixel shader precision to show a difference to FP32 (I was using a shadow mapping demo, if anyone cares) as it ignores, there's no issue (that I can detect) here.

I'd say fixable driver bug if something is actually up, or developer error.
 
The Baron said:
Of course, there's situation C: this only happens on NV3x hardware. In fact, I think it might force FP16 (or FX12 for NV3x<5), and I don't really have a problem with that.
I do (magnified large textures would show obvious banding if you used FP16 texture coordinates), but there certainly is no unconditional forcing of FP16 on NV3x going on. Conditional changes are of course much harder to detect.
 
Parousia said:
Technically, this is an incorrect statement. I'd be mad as hell if I got 32-bits automagically if I never wanted or specified it.

Excuse me, but. . .huh? The default for ps_3 isn't FP32?
 
Parousia said:
Technically, this is an incorrect statement. I'd be mad as hell if I got 32-bits automagically if I never wanted or specified it.
Unless you ask for a half, you get full precision.

And technically you don't tag a shader with __pp, B. It's an instruction modifier, not a 'shader' modifier.

Edit: although you can force the modifier per shader with the Gpp compiler flag to fxc (cheers, Chris, for the reminder). However the developer has to make sure that's the case at compile time, since it's not the default.

PS, Chris, I can't reply to your mail (Recipient address rejected: Account/Mailbox has expired due to inactivity.) Doh!
 
Sharkfood said:
Yes. And not for what you feel it's indicative of, but the very point I was making.

When any *other* IHV is accused of wrong doing, it's always a 38-50 page thread accompanied by web site front page links, side-by-side screenshots, java applets, code snippets, sample test programs, IHV and developer interviews, and comprehensive lists of applications effected.
Eh. There are plenty of examples of both.

Then again, this is the first time I've heard somebody say B3D was biased in favor of NVIDIA :p
 
geo said:
Guys, Fonzi is throttling the motorcycle here. . .
Weird. I finally look up "jump the shark" not a week ago. If this thread contributes nothing else, at least we've brushed up on our TV critic jargon.

Interesting to note is how involved some us get to right perceived wrongs. I found my way back to AT for basically the same reason trini ratcheted up his participation here, and this same reason is used to explain a lot of (IMO) over the top behavior in certain other fora. Of course, trini and I were the only truly righteous ones in a sea of misguided heathen. It's a fine line b/w anti- and reverse-fanpersonism, though, and if I hadn't similarly (to trini) matured in character, I'd worry about crossing it, eventually. ;)

dizietsma said:
Fuad's piece failed to reach impartial precision and ended up with just partial precision.
:smile:

Finally, as to Fuad's typically insightful outspeak, what's the claim? (I'm reluctant to click on what appears to be a gimmick.) That NV has only recently switched to forcing _pp on anything and everything? B/c (to add to the many reminders) I remember Mandelbrot hijinx (mmm, precisiony) showing FX12/FP16 through FP32 from way back when, not to mention the oodles of custom pp/fp benchmarks we've seen from HW.fr, Xbit, and iXBT. Both prove, along with everyone else's indignant links, that NV could do both FP16 and FP32 at the drop of a hint since basically the beginning of time (or DX9, whichever is closest). So all I can think is that the claim purports that new NV drivers have only recently forced everything down the _pp path, which (if true) just sounds like human error more than foxy guile.

So, yeah, I just wanted to link to that pic of the Fonz. Oh, and props to Rys for his research. Also, laffs at Rys for being gullible enough to engage in said research when a mocking and/or indignant link would have sufficed. ;^)
 
  • Like
Reactions: Geo
Back
Top