[H] Benchmarking Future Ed.

DaveBaumann said:
Whether or not you or other feel he has brought this on himself is beside the point, you are being rude to B3D by bringing these types of attacks into this forum.

I have not posted anything rude in this post, I stated quite clearly what the issue was, in fact you point me where I did and I will edit it.
Funny how Extremetech or Beyond3D or any other site like AMDMB.com that has a different opinion than [H] gets labelled as low credability.
 
Dt, I wasn't specifically singling out anything you've said - I've not actually read this entire thread yet, however in the last few pages I just noted that two people have the responce "Kyle is an *ss..." which I feel isn't particularily constructive. Your responce that "he did it so we can" is also not really the type of attitude that should be taken here.
 
DaveBaumann said:
Pete said:
A little more civility would suit this forum well.

The level of vitriol is fairly irksome. I'd certainly hope that we could maintian a higher degreee of converstation about the points at hand here without resorting to the less constructive comments.
Sorry, it's just the frustration...but I shouldn't waste bandwidth on it here. :(

My apologies, I will stop such behavoir and conduct meself accordingly.
 
Another ramble

Once upon a time I had a thread here with a detailed discussion of my opinions on criticism. I don't know what happened to it, and my inability to reference it has irked me on many occassions in the past few months. :(

What I feel is relevant to this thread is the part of the discussion when I mentioned evaluating whether your criticism is worthwhile to post. In this context, you should know a few things about where I'm coming from on the matter:

I have an extremely negative opinion of some of the things Kyle has said.

I've expressed that opinion (hopefully while avoiding using words like "ass") in the forums.

I try to avoid expressing that opinion just for the sake of expressing that opinion, and try to reserve commentary for when I'm discussing something to do with the forum topic (which isn't Kyle Bennet, but 3D technology EDIT: pardon me, Kyle is on topic for "3D industry"...that should be "isn't calling Kyle Bennet an 'ass', but 3D industry related discussion" :oops:).

How does this relate to what I'm trying to say? In short, my view is that: the problems with what Kyle says are directly related to 3D technology and the discussion of the industry, but Kyle being an "ass" or "egomaniacal", or whatever label you prefer, is not.

Perhaps thinking of things in that perspective will shed some light on what might possibly be wrong with some posts (not people, not opinions, but the posts the people made and the method of expressing their opinions it represents). Namely, don't confuse calling him an "ass" with discussing the problems with what he has stated....though they might be closely related factors in your thoughts about him, one doesn't hold water as a continued discussion thread in this particular forum.
Also, please don't equate Kyle's selective quotes and criticism with this request do to superficial resemblence....consider my words independently, whatever they end up being worth to you.

...

Yes, this post breaks the "on topic" rule :-?, but, believe it or not, I covered that in the missing thread as well....basically, I'm hopeful it will be food for thought for some, and help in further support achieving what (I think) Wavey is asking for.
 
DaveBaumann said:
Dt, I wasn't specifically singling out anything you've said - I've not actually read this entire thread yet, however in the last few pages I just noted that two people have the responce "Kyle is an *ss..." which I feel isn't particularily constructive. Your responce that "he did it so we can" is also not really the type of attitude that should be taken here.

My apologies, my paranoid tendencies are getting the better of me :LOL:
 
heh

Doomtrooper said:
DaveBaumann said:
Dt, I wasn't specifically singling out anything you've said - I've not actually read this entire thread yet, however in the last few pages I just noted that two people have the responce "Kyle is an *ss..." which I feel isn't particularily constructive. Your responce that "he did it so we can" is also not really the type of attitude that should be taken here.

My apologies, my paranoid tendencies are getting the better of me :LOL:

I'd like to see what you think after reading this> http://biz.yahoo.com/prnews/030624/sftu011_1.html

:devilish:
 
I am beginning to think that Nvidia's PR department is doing this just to piss us off. :devilish:
In the vein of Dave's request my wife always says separate the deed from the doer.
 
Re: heh

Jima13 said:
I'd like to see what you think after reading this> http://biz.yahoo.com/prnews/030624/sftu011_1.html

:devilish:

I can hear blood vessels popping all over the forum. This is just a PR blurb put out by Nvidia, using selective quotes from the websites that are being used as an extenstion of that same PR department.

Look at the qualfication statment at the bottom. "Based on a biased test from a biased website, using an unreleased game on unreleased hardware, in our own specially prepared test against an unprepared competing product with faulty drivers." FACT!

It's no wonder Nvidia are so good at that Marketing Spin - it's the only product they've been offering for the last year. They can put it alongside the award for "Best GPU of 2002" for the NV30. How about we get some products rather than marketing blurbs, eh, Nvidia? :LOL:
 
1) It's obvious my magnanimous mood was mistimed in light of Kyle's recently re-exposed egomania in lumping Rev in with the vast Radeon-wing consipiracy against him. Well done, Kyle, for kicking the legs out from under those who would extend you courtesy over cuts.

2) That nVidia PR is almost nauseating. Curious how they didn't use THG's Doom 3 numbers at the same settings, which showed the opposite picture: the 9800P beating the 5900U. Either way nV explains it looks bad for them: drivers mucking up AF disproves their faultless driver mumbo-jumbo, and faulty testing pretty much disproves all scores. I await ATi's response quoting GamePC's entire eVGA 5900U review (in which the OC'ed eVGA 5900U loses to the 9800P basically across the board), as well as some of FS's and UKGamer's recent 5900/U reviews. Kyle faults FM for being all about tha Benjamins, and yet it's painfully obvious nVidia and ATi are after those same dead Presidents (or, if you're in Europe, nondenominational landmarks ;)). nV's IQ pronouncements are particularly irksome, though. I'll grant them their card has the raw horsepower to bench at the top, but it certainly has not been proven to have the IQ to match. I like how careful they are: "In addition to stunning performance, the GeForce FX 5900 GPU features...." Note that a literal interpretation puts those "features" in a separate category from nV's "stunning performance." :)

3) And Lars has posted his take on the matter over at THG. It's an even-handed editorial touching on every recent battle in the war against marketspeak. He makes two interesting points. One, that ATi ("A 3D chipmaker who prefers to remain unnamed") was responsible for the first (ExtremeTech, B3D, eventually FM) investigation into 3DM03 cheating (clip planes, etc.)--the same "unnamed" IHV who tipped TR off to the nV 3DM03 aniso iregularities. Two, that nV's "adaptive" aniso is, according to nV's own replies, only "adaptive" if nV spends time "adapting" it to an application (paging radar1200gs ;)). Worth reading, IMO.
 
Goddammit Crystal! Now you guys made me read the [H]ard essay! ;)

Anyway, all I read is a number of different ways of stating that they now 'think' that synthetic benchmarks are useless, like:

We have already seen the world of synthetic benchmarks crumble around us leaving a bunch of useless data that one might argue was worthless to begin with.

... and a little later:

If these synthetic benchmarks are as valuable as some might have you to believe, then why can’t those people use those utilities to conclusively tell me what is going to be the best card to play Half Life 2 or DOOM3?

Reality call: Everybody know/knew that we have to be careful about what you conclude from a benchmark - synthetic or not. Those synthetic benchmark can be quite usefull to meassure specific parts of performance (and I assume Hardocp is talking shaders here).

If the reviewer doesn't understand the data results - or think that they represent the performance of a whole engine - I can see why one would claim them to be "useless". But I would hazard the guess that Kyle have been in the business long enough to understand that the quotes above are really bollox for all but the ignorant.

BTW: For those who few think that Kyle is totally pro-nVidia I suggest that you read some of the reviews at [H]ardocp. I seem to remember the NV3X line getting some pretty harsh word on the CineFX/DX9 shader performance. Just another hint to my point. [/quote]
 
Can we all wave to Kyle to show we still *love* him really? ;)

Well I gotta admit, I'm generally a fan of the 'eveyones' entitled to his own opinion'. If that's his opinion, that's his opinion. A lot of the posters may not feel it's justifiable, but that's besides the point. It's Kyle's opinion and more power to him for putting it up on the world wide web.

Following that chain of thought it's my opinion that Kyle's editorial (what I've read of it) is incorrect. Sure Synthetic Benchies aren't perfect but what testing methodology is?

I have no problem games being benched on custom code paths as long as it's clearly stated and/or they tested without reference to a competitors card.

that yahoo article is just plain weird though.
 
Heathen said:
Can we all wave to Kyle to show we still *love* him really? ;)

Well I gotta admit, I'm generally a fan of the 'eveyones' entitled to his own opinion'. If that's his opinion, that's his opinion. A lot of the posters may not feel it's justifiable, but that's besides the point. It's Kyle's opinion and more power to him for putting it up on the world wide web.

Following that chain of thought it's my opinion that Kyle's editorial (what I've read of it) is incorrect. Sure Synthetic Benchies aren't perfect but what testing methodology is?

Everyone can have an opinion, but a lot of them will be incorrect, often due to ignorance of the facts. I can have the opinion that the moon is made of green cheese, but that doesn't mean it is.

Heathen said:
that yahoo article is just plain weird though.

It's not an article - it's a press release. The source is listed as Nvidia.
 
Quote from Brents missive...
" Lastly, the timedemo stuff. Yeah, it’s not perfect. The evidence points to some foul play in timedemos, can’t deny that. And we know it’s not perfect either. In fact we have already been discussing the fact of using private timedemo’s made by us and only used by us. We have some thought going that we will probably make 2 timedemo’s per game. One that is private, and one that is public, therefore the private one acts as our control. We will be able to see if any foul play happens. There are many things we are considering, and believe me we are working to get better honest and fair benchmarking. "

Found your post as a whole enlightening and interesting. I do have a couple of observations about the above.

On the face of this, It sounds like a good idea, and that it might be made to work with one caveat. I would have to be able to trust the Website and it's editors to not leak the Private Demo to any IHV's and my "A" list of those trusted Web sites is quite short at this time. I am not particularly trusting when it comes to People who may have a vested interest in how things transpire. This is not directed at you personally. What may happen without your knowledge is something else though, that you have no control over.
Trust must be earned, and once having been earned , if broken, becomes even harder to garner. I am quite saddened by this whole affair as the trust I had for a certain Hardhitting, raw, (and believed to hold unbiased content) Web site has been broken. There have been some very good questions asked here by the serious posters as to the motives of a certain web editor, and an explanation and some substantiation for his stance in a reasoned manner has not been forthcoming. Instead he chooses to single out the one liners that suit his purpose as he slams and catergorizes this web site based on those few chosen remarks that are ignored here by the serious minded just as he supposedly does. He still avoids the valid issues raised here by a few, with rebuttals like that instead.


Now as I read it, the [H] stance is to change the way benchmarking will be done, which on the surface seems good, but I have to wonder who is directing this endeavor. It sounds like Nvidia PR speak almost from word one. [K] prognosticates about the future of 3D programming, (Divergent Code Paths and all game developers aligning themselves with a specific IHV) as if it is a given, but this sounds exactly like Nvidia speaking. This is how Nvidia would like to see things go. It just seems like more of the Nv party line.

Some editorial excerpts:
" Quite simply, NVIDIA has changed the rules of synthetic benchmarking. Synthetics were understood to be free standing utilities that were not optimized for. We think that is how the hardware community saw these tools. But let’s look at the facts and legalities. As in the case of 3DMark03, we don't think there was anywhere in their EULA that specified you could not “optimize†for the benchmark. We don’t think “cheating†was defined. Like many other synthetic benchmarks, everyone understood the rules of the game, but no one defined them, as they possibly should have been."

This sounds like pure justification for Nvidia cheating... And now he's a Legal Expert too! but He's right! "everyone understood the rules of the game" It's a matter of having the integrity to play by those rules Fairly that have come into question and the integrity of Web sites who ignore the real issues.

"NVIDIA is optimizing for benchmarks quite clearly and that is something we will have to deal with"

Why do we HAVE to deal with this? Why do we HAVE TO put up with cheating? Why are they above the law. Has Nvidia ever heard of the word Ethics? How about [K]?

" Apples to apples benchmarking is becoming more difficult. "

Only more difficult because Nvidia is making it so.

"The fact of the matter is that benchmarking the same code path on competitive video cards will not give us an accurate representation of gameplay on one of the two major cards out there.""

Because Nvidia chose not to follow DX9 standards? So where Nvidia leads the sheep should follow? Sorry, but they screwed themselves by believing that they held more power in the industry than they thought. To hear them talk, 9 out of 10 Games are being developed on Nvidia platforms with it soon to be 10 out of 10, implying that the Developers are Only using Nvidia Hardware. and if you don't own an Nvidia card, no future games will run for you?

" What we mention above with DOOM3 is only going to become more widespread and not necessarily in such a cut and dried fashion. We are going to see D3D games that behave differently within the same API."

Silly me, I thought the Game/API told the hardware what to do, not the other way around? I guess not in Nvidia's world.

"As the video card companies move forward with developing tighter bonds with game developers and publishing houses, there is no doubt we are going to see more hardware specific effects in games. Depending on your hardware, your game experience may differ. To what extent is yet to be seen, but we think that this alone will show that the days of apples to apples is coming to a close as the proprietary hardware technologies diverge."

The world according to Nvidia... But I thought that DX9 was the standard, Not Nvidias interpretation of it via non-standard code paths using sub-standard precision?

"While this is just speculation, we think that NVIDIA moving some of their high-end business to IBM will facilitate this happening more rapidly as each faction of the video wars becomes a bit more shielded from each other’s technology.
"

Most definately speculation, and why would hiring IBM to Fab promote divergence of Video technology? More thinly veiled Nv PR to justify divergence?

Kyle continues to aide and abet Nvidia in it's long term goal to dominate the Video card market through Proprietary means. His stance that divergent engines is a given, and that that is Ok is scary. Splitting the Game developers into ATI or Nvidia camps hurts the industry and especially the Gamer. I want the Game developers to code to a standard that will run well with good IQ on any capable piece of harware I see fit to use.

Proprietary API's= bad
Standard API's =Good!

All I have to to do to remember this is recall getting a Diamond Stealth S220 with the Rendition 2100 chip some years back. Worked pretty good, But it didn't look very good on any number of games because They either ran in software mode, or used Glide. The one I remember best was Janes Longbow Gold. Glide only or 640X480 software mode. yuck! Luckily I had a few Rendition "Optimized" games too, like Nascar Racing. And when coded for, The Rendition card was no slouch, and looked pretty good too.

anyway... The point being that I didn't like seeing cool games out there that didn't take advantage of my Hardware. Playing a game at reduced settings and or with poorer IQ because My hardware wasn't supported generally made me pissed. Why not code for both? Ding! the light goes on! duh! because it is more work!
I believe in the K.I.S.S. principle. Keep it Simple Stupid! Lets keep it simple and straightforward with one or 2 Standards, say DX9/OGL, and build our Hardware to run those fast without resorting to special Codepaths etc. Seems simple enough. That way all the games run on all the Hardware. Why regress back to Glide days... wait, didn't the 3Dfx crowd migrate to Nvidia. Has 3Dfx conquered from within and are now running Nvidia "The Way it's meant to be Run"? :oops:

These are my opinions and mine alone. They do not represent in any way the Opinion of any B3D staff. ;)
 
Pete said:
And Lars has posted his take on the matter over at THG. It's an even-handed editorial touching on every recent battle in the war against marketspeak. He makes two interesting points. One, that ATi ("A 3D chipmaker who prefers to remain unnamed") was responsible for the first (ExtremeTech, B3D, eventually FM) investigation into 3DM03 cheating (clip planes, etc.)--the same "unnamed" IHV who tipped TR off to the nV 3DM03 aniso iregularities. Two, that nV's "adaptive" aniso is, according to nV's own replies, only "adaptive" if nV spends time "adapting" it to an application (paging radar1200gs ). Worth reading, IMO.

Pretty good read Pete. Thanks for pointing it out. I really liked the end of his article. Where he talks about reveiwers being pawns for the IHV's PR machine, by publishing spoonfed stories without adequately researching them, in order to be the first to publish a story. (Man, that was one heck of a run on sentence.)
 
Back
Top