The Way its Meant to be Reviewed?

True, but I've really been worried that lately they've been sliding back into a bit of a fansite.

I truly hope I'm wrong, but I am worried.

 

To be honest I never paid real close attention to the "bias-meter" on sites until I became more closely involved with them so I'll just have to take your word that nV News was very biased. I wouldn't really be surprised though because ATI was crap back then...3DFX was on its last leg...so it would only be natural to be fanatic about a company that was doing it right and still alive.

I think the 8500 from ATI really changed the landscape. From that point on it was becoming clearer and clearer that we had a real Ford vs. Chevy fight on our hands and not a Ford vs. Yugo as it had been. So it was a natural progression of things. It wouldn't make any sense to give much credence to ATI at the time because they simply didn't deserve it...back then. Today is obviously a different story.

I'd like for you to elaborate on this over in our forums if you have time.

Thanks!

 

Yeah Digi...don't make Dave go get Tim's trout!

*skitters off at realization that /me deserves a few smacks meself*

 

Sorry, you already did answer it further back here. Thanks!

 

quote]
This is not what Carmack has said at all. What he said was that using the NV30 path, the NV35 was on par with the 9800 Pro in shader performance. However, using the ARB2 path, the 9800 Pro was about twice as fast in shader performance.
[/quote]
Actually those year old comments are not what i am refereing to. I am refering to what he stated in an interview last week. Hwe said that other than the specific case of FP32 Ati and Nvidia hardware performed at about the same level.

I dont have the link... I am sure one of the NVN guys can post it.

 

That's funny because that's not what Shadermark, Tomb Raider AOD, HL2 or Ashlimark are showing. I guess they're all wrong and JC is right?

-FUDie

 

RESPONSE. The word is spelled RESPONSE.

Spelling aside, my point was simply being a part of the beta program slants any opinions the author may have on the subject of synthetic benchmarks. I am not trying to argue, I am merely stating what many people must be feeling after reading the article.

I was directed here from another discussion board discussing the issues I am raising here.

Thanks for the time and response.

 

I'd say that's one of the weaker car analogies I've read, and I've read a few. The difference is that most people who buy a car test drive it in conditions they anticipate using it for most of the time. A road is a road, a traffic light is a traffic light, and things ain't gonna change much over the lifetime of that car. (Not to mention that most people already have a basic grasp of horsepower and cylinders and tire size--more is better.) Games aren't the same. Practically every other game has a different game engine and will perform differently. Look past the games available when you buy the card, and the landscape changes even further.

To the guest insinuating that B3D's 3DM beta membership biases them, I think you have it backwards. I view it more like B3D thinks synthetics are potentially valuable, and they joined the 3DM beta program to get more in depth with the program and perhaps to help influence its future direction. The B3D crew didn't change their minds about synthetics after they joined the beta program. On the contrary, as Dave said, B3D has been interested in using them before 3DM03, and B3D's participation in the beta program implies they're interested in using synthetics after 3DM03.

No doubt these "extra" editorials and articles on synthetic benches are more numerous and more visible due to nV's very public flaming of 3DM03 and the corresponding somewhat muddled reaction from other sites. I think the very questionable Doom 3 and Half-Life 2 "real" benchmark events also add to the question (answer, IMO) of a sythetic's relevance.

As for the latest Guest post about the FX being equal in everything but FP32, what's your point? I believe that full, not partial, precision is assumed in DX9. That's mostly a semantic distinction ATM, though, and I await some decent HL2 (and D3, to a lesser extent, given Carmack's comments) comparative screenies.

 

Well cool, at least I rate in some capacity...even if it is last. I do hate car anaolgies and they all suck...just couldn't resist, hadn't done one in a while.

 

[OT] reading threads where a lot of replys are from guests guests guests becomes annoying ... only "loged" members should be allowed to post/reply ...[/OT]

 

God save the Queen!

Heh, glad you took it so well. Note I said it was one of the worst I've read. I said nothing about the ones I wrote.

Edit: LOL, I just read Dave's reference to TR and checked out Diss' news post thread. Lo and behold, what should greet me on nothing less than the second post: a car analogy!

 

It's ironic that a big critique of 3DM03 was that it was poorly (inefficiently) coded, yet people are leaping at benching TR:AoD and Halo as real DX9 games despite the fact that they seem to be rather terribly coded themselves.

Am I right, folks? Eh? Eh?

/me ducks tomatoes and oversized boot

 

Well.. I think the Jury is out still on what the final version of HL2 will perform like on Nvidia hardware.

As for the others... there is no doubt that *initially* and in some cases *Still* perform better on ATi hardware. However that is not the type of cases i was addressing here. What is being adressed are the games comming out over the next 12 months. Those games will be hand coded by Nvidia, Favor FP16 and the specific instruction order they like etc etc etc.. You can already see it happening. As in the cases of Painkiller, STALKER, Doom-III and others where little comments of the Devs message boards give it away.

Thats the point. Synthetic benchmarks today show the raw performance on Untuned software yes. But will not reflect what the websites and benchmarks and results will be for the 10-20 super popular games each year.

 

That Critique or accusation was invented and pushed by Nvidia and Nvidia alone.

There are a great deal more people and professionals who feel it is a fine job of coding and represents the style of game code that will be seen in the future.

(that is not hand coded or bought off by Nvidia )

 

And an interview is not an editorial. Perhaps Reverend should have put some Q.'s and A.'s in there for you.

 

So nVidia's FX will be good for the 10-20 super popular games each year and you should expect sub-par performance on the hundreds of others?