Joe DeFuria said:
Demalion: Agreed.
I wonder if Beyond3D is going to post their own Full Analysis of 3D Mark (beyond their introduction.) I honestly don't know what their full evaluation of it is, but the reputation of this site's quality, technical competance, and perception of fairness may have a bigger impact on "the public's" view of 3D Mark than review sites...
I would also love to see B3D not only analyze and give concluding thoughts on 3D Mark...but also on NVIDIA'S PR....
My concern here is what sort of "analysis" could B3D do beyond what they've already done which wasn't merely an opinion on the "pros" and "cons" of 3DM'03? I think that 3D Mark is what it is--and the authors of the benchmark have done a credible job explaining why they did what they did--and so the benchmark will stand or fall based on its popularity. I think B3D did such a superlative job reporting on 3D Mark '03 precisely because they injected not a whisker of opinion into it--B3D simply reported what it was without making value judgements. I don't see how B3D could improve on that. Excellent reporting. Far above THG and [H]--not even in the same league.
In the end it won't matter what nVidia or Kyle or THG or FutureMark or ATI or B3D say about it--it sinks or swims based on what it is and whether or not people like using it and derive any perceived utility from it.
I am frankly mystified beyond belief by those whose complaint about the benchmark is that it is synthetic (it's always been synthetic and certainly isn't any *more* synthetic simply because 3D Mark has stopped paying lip service to the Max Payne engine), and those who complain that the benchmark is "too advanced" and there "aren't any games programmed like it" available right now. They said exactly the same things about 2001 when it was first published, as I recall. Those are not valid reasons for opposing the benchmark IF you have previously supported one of its earlier versions, IMO.
This is all very ironic for me because I have never "liked" 3DMark in any form--and can actually state I've never even run it! *chuckle* I didn't have to run it to know I didn't like it because it was synthetic.
But that was not really why I have always personally disliked it. My biggest reason for disliking it was furnished by the sites who gave its results such a prominent degree of importance! These sites often suggested that one ought or ought not to buy 3D hardware based on results collected by 3D Mark! That's why I've always had a personal beef with the benchmark--not necessarily because of what it is. What it is is a synthetic benchmark no more "evil" than any other synthetic benchmark. The "evil" part comes into the picture by way of 3rd parties who misuse the benchmark in certain subtle ways to commercially promote one brand over another. I think everybody in these forums knows what I'm talking about.
I've read several misinformed comments--not just in the B3D forums but all over the Internet--in which people state that nVidia has "always supported" 3D Mark until now. Good grief, I can clearly recall nVidia making the *exact same negative comments years ago* about FutureMark's 3DMark benchmarks--that nVidia was "opposed" to them because in nVidia's opinion the benchmarks did not represent the "actual" way 3D games were being written or were likely to be written. I recall these public, official remarks from nVidia at the time the nVidia/3dfx wars were taking place--even when, curiously enough--it became evident that the benchmarks were favoring nVidia's hardware approach over that of 3dfx's. At the time, I wondered more than once about these remarks, as they were ones I could fully agree with and yet felt were somehow out of character for nVidia to be making.
Then a little later I realized what nVidia was doing--on one hand condemning the benchmarks while on the other optimizing their drivers for them and seeking to turn 3DMark into a nVidia asset to whatever extent became possible. nVidia was playing both sides of the situation, in otherwords--it could appeal to the people who didn't approve of synthetics being made pivotal in the evaluation of the efficacy of 3D Hardware, and on the other hand it could exploit the benchmark to the maximum extent possible in order to facilitate sales of its hardware. Really, it's not a bad policy from an intellectual point of view--it's a smart one, actually. But it is also a policy rife with duplicity which just caused me to disapprove of the use of 3D Mark benchmarks in 3D hardware reviews all the more.
As far as I'm concerned, the only thing that's changed currently with respect to nVidia's public statements or its actions (as we can see nVidia is still optimizing drivers around the benchmark regardless of what it has said about it), is the fact that this newest version of 3D Mark is not so easily manipulated into supporting congruent aspects of nVidia 3D technology--not only nv30 technology, but much of nVidia's earlier technology. Ironically, although still as synthetic as it has ever been, the '03 version of the benchmark actually seems to be *more accurate* in illustrating how 3D hardware runs in actual 3D games than its predecessor ever was! *chuckle* And it is this point, I think, which has led nVidia to not only publicly lambast the FutureMark benchmarks *yet again,* but *this time* has led to nVidia severing all its relationships with FutureMark--relationships it had sustained even while simultaneously expressing a hypocritical distaste for the benchmark and its artificial nature.
The real difference this time, I think, is that 3D Mark serves more to point up weaknesses in the nVidia 3D approach than it does to underline what strengths nVidia thinks its technology has, and so nVidia has decided to put some distance between itself and 3DMark--or at the very least to offer an apology as to why its vaunted, expected-to-ship--soon and extremely rare, hard to find GF FX Ultra technology can barely keep pace with the 6-month-old DX9-compliant hardware from ATI built on .15 microns and running 175MHz slower. If not quite a scapegoat, nVidia is certainly attempting to paint 3DMark as "software which isn't giving our products a fair shake." At the same time I note with much humor nVidia is still optimizing heavily for 3DMark...
I read one statement by nVidia recently which seemed to say that nVidia thought that optimizing its drivers to support synthetic 3D benchmarks was a "good thing"--since even though 3D gamers could not *now* see any profit from those optimizations they would at sometime in the future be able to do so when 3D games were written which supported the same features the 3DMark '03 benchmark supported and the drivers were optimized to support as well. This at least is the rationale provided by nVidia to explain and sanitize its practice of optimizing drivers for synthetic 3D benchmarks. But wait! Isn't nVidia making a liar out of nVidia in saying something like this? *chuckle*
Sure looks like it to me, since on the one hand nVidia has disowned the 3D Mark '03 benchmark saying that it is not an accurate representation of how 3D developers are writing 3D games and so we aren't likely to see the 3D Mark approaches show up in 3D games anytime soon; yet on the other hand nVidia states that optimizing benchmarks for 3D Mark is a "wonderful" practice because it means that when 3D games begin including the kinds of features 3D Mark '03 supports *then* the users of their "optimized" drivers will see tangible benefits. It's almost too funny for words--it's a direct contradiction in terms--and it is literally nVidia making a liar out of nVidia...
If it's indeed true, as nVidia has stated, that nothing in 3DMark '03 will ever see the light of day in a 3D game, then obviously no "optimizations" nVidia makes for the 3D Mark '03 benchmark--so that its products achieve better "scores"--will mean a hill of beans because of that. So which is it, nVidia? *chuckle* (What a mess the company has gotten itself into--I hope nVidia wises up and fires the whole PR staff--they've made a royal mess out of almost anything with the nVidia name on it. Untruths and misrepresentations have a predictible habit of ensnaring the very people who seek to use them to deceive others.)
It's abundantly clear to me, anyway, that nVidia is tripping all over itself trying to "weasel" out of whatever jam it thinks it has stumbled into. The more it talks the worse it gets. Accordingly, how can anybody give sites like THG the time of day? It's hard to even give them the benefit of the doubt, since they so obviously and thoroughly are in pursuit of another agenda altogether, an agenda the specifics of which are hidden from all but themselves. I saw a little blurb on [H] tonight in which Kyle wanted everyone to know that there had been an awful lot of "misunderstanding" going on and that *he never meant to state or imply* that his 3D hardware reviews would *leave out* 3D Mark 03 in its entirely--only that 3D Mark '03 would not be used to in the weighting of the "final score" a product gets in a review. (Wow--wouldn't 'ya know I thought that's what he'd always been doing?) Why, if anyone thought that he ever meant to *abandon* 3DMark '03--how on earth could they have arrived at such a silly notion? *chuckle* How indeed. I haven't seen this much doubletalk and equivocation since Clinton left office....
Hmmmm....maybe nVidia should hire the ex-President to be a mascot for its nv30 products?....
(Maybe he's already been hired and *that's* what this is all about.)
