Some usual funny nonsense from THG...

[T2k]

I just saw this posted over @ R3D... (I've stopped reading T's Garbage long time ago.)

Because of the ongoing discussions regarding "legal" or "illegal" driver optimizations, some people lost trust in NVIDIA. The new critics from Valve won't help NVIDIA to get this trust back. But it's also up to Valve now to offer a solution and a more in-depth explanation to the million owners of NVIDIA cards as to why their cards perform so badly with Halflife 2 at the moment. Of course, it's also in their interest to sell copies of the game to those people.

ROTFLMAO...

Taken from THG's conclusion: http://www20.tomshardware.com/graphic/20030912/half-life-07.html

 

[Bjorn]

...But it's also up to Valve now to offer a solution and a more in-depth explanation to the million owners of NVIDIA cards as to why their cards perform so badly with Halflife 2 at the moment.[/b] Of course, it's also in their interest to sell copies of the game to those people. [/i]
.....

You missed this part though:

That is one side of the story. The other is that NVIDIA now has to prove that their cards are really ready for DX9 games - as they promise with the upcoming Detonator 50 driver in their response to Valve.

 

[T2k]

...But it's also up to Valve now to offer a solution and a more in-depth explanation to the million owners of NVIDIA cards as to why their cards perform so badly with Halflife 2 at the moment.[/b] Of course, it's also in their interest to sell copies of the game to those people. [/i]
.....

You missed this part though:

That is one side of the story. The other is that NVIDIA now has to prove that their cards are really ready for DX9 games - as they promise with the upcoming Detonator 50 driver in their response to Valve.

No, I didn't but seems you do: it's just more cosmetics, nothing else. (I don't wanna say BS...)

On the other hand: he couldn't be dead serious about Valve... why Valve should have offer the solution? Especially after they DID put a lot of effort spending 5x more time to create a goddamn' lowered precision version codepath for NV's FX-family only?

This is preposterous.

EDIT: typo

 

[Bjorn]

On the other hand: the couldn't be dead serious about Valve... why Valve should have offer the solution? Especially after they DID put a lot of effort spending 5x more time to create a goddamn' lowered precision version codepath for NV's FX-family only?
This is preposterous.

I agree, Valve has already offered a solution. They have DX9 mixed modes, they have DX8 paths that you can use with the slower FX versions, what more can they do ?

But why make things worse then they are by leaving out imo critical information when quoting something ?

 

[nonamer]

THG has point believe it or not. HL2 benchmarks have currently left a bad taste in my mouth because the performance for the NV cards against the ATI cards in HL2 are at or below the theoritical minimum for a pure pixel shading tests. Then you have very well respected sites like Anandtech release some impossible numbers (how could a 9600 stay with 0.1 fps at every resolution?!) I've mentioned before that due the heavy PS nature of HL2 and the weak PS performance of the NV3x series, I expect the NV3x's to perform slower, but not like this. Something just isn't adding up here, and I'm reserve judgement on the issue for now.

 

[Bjorn]

(how could a 9600 stay with 0.1 fps at every resolution?!)

Just looked at this and i agree. That seems rather strange. And it doesn't match up with Tom's numbers. Probably a simple mistake though.

 

[OpenGL guy]

THG has point believe it or not. HL2 benchmarks have currently left a bad taste in my mouth because the performance for the NV cards against the ATI cards in HL2 are at or below the theoritical minimum for a pure pixel shading tests.

This doesn't make sense at all. Pixel shading tests have shown that there is a large performance difference, HL2 is no different.

 

[T2k]

But why make things worse then they are by leaving out imo critical information

That's the key: imo - not mine...

 

[Evildeus]

Is it the good forum? Good citations though

 

[Oompa Loompa]

I just saw this posted over @ R3D... (I've stopped reading T's Garbage long time ago.)

But it's also up to Valve now to offer a solution and a more in-depth explanation to the million owners of NVIDIA cards as to why their cards perform so badly with Halflife 2 at the moment.

I'm not sure what you find so interesting about this. If people with NV3x cards discover that they can't run HL2 properly, they are going to complain to Valve, not to the company who built their graphics card (let alone Nvidia). Valve has already done their best to make the game run well, and I'm sure that's the answer they will provide.

I don't think Tom's article is critical of Valve, nor does it let Nvidia off the hook.

[shrug]

 

[Doomtrooper]

Valve has done their part, they have spent 5X the amount of time working with Pseudo DX8/9 hardware. Miracles can't be done here, people are asking 'blood from a stone'....Nvidia must fix their hardware...since we know that isn't going to happen the article should have been alot more critical of the IHV vs. the Dev.

Especially the cheats exposed, screen capture....no mention of that at THG..as usual. The same site that harps all the time about 'talking to devs'

 

[Borsti]

I wrote some comments on why I chose this conclusion in the THG community Forums:

http://www.community.tomshardware.c...0607&page=0&view=collapsed&sb=5&part=all&vc=1

I also talked about the optimizations and cheats Valve is worried about:

http://www20.tomshardware.com/business/20030911/index.html

Borsti

 

[epicstruggle]

I wrote some comments on why I chose this conclusion in the THG community Forums:

http://www.community.tomshardware.c...r=510607&page=0&view=collapsed&sb=5?=all&vc=1

I also talked about the optimizations and cheats Valve is worried about:

http://www20.tomshardware.com/business/20030911/index.html

Borsti

Thanks for posting your side. Just to clarify something to myself, did anyone from valve officially say which driver version was (trying to) output(ing) higher quality screenshot?

later,
epic

 

[Dave Baumann]

Again, Valve were listing general optimisations that have been seen - they did not specifically relate any of these to any one benchmark or driver release, but rather these are the types of things that have been seen.

 

[WaltC]

...But it's also up to Valve now to offer a solution and a more in-depth explanation to the million owners of NVIDIA cards as to why their cards perform so badly with Halflife 2 at the moment.[/b] Of course, it's also in their interest to sell copies of the game to those people. [/i]
.....

You missed this part though:

That is one side of the story. The other is that NVIDIA now has to prove that their cards are really ready for DX9 games - as they promise with the upcoming Detonator 50 driver in their response to Valve.

What I thoroughly dislike the most about T's style of commentary is that it is constantly seeking to create equalities between events when no such equalities exist. For instance, when they glibly talk about "cheating" among IHVs as though they are all equally guilty at present and so there's no sense in getting into specifics. The IHVs are currently not equal in that regard and such commentary is dishonest. In this case, he's saying Valve needs to do something in the future to optimize for nV3x as though it has equal weight with nVidia proving it can do something legitimate to improve its products' DX9 performance. The truth is that Valve has already done what it can do in that regard--and most likely, nVidia has, too. But if anyone has work yet to do regarding these matters it is certainly nVidia, not Valve. For some reason T wants to paint everybody and everything with the same brush which results in nothing but false or highly misleading commentary. Very, very strange style. It is as though the author wants to place himself above the events he's reporting, and so he creates artificial equalities and downplays and minimizes and generalizes to such an extent that he contributes nothing of value to the discussion.

 

[WaltC]

Again, Valve were listing general optimisations that have been seen - they did not specifically relate any of these to any one benchmark or driver release, but rather these are the types of things that have been seen.

Yes, that was my impression as well. The only specific I saw was when he was asked about which game it was that failed to render fog...to which he replied HL2 (or so I read.) I still see his commentary as more foundational for helping people to understand why the DX9 performance disparities exist between IHV products running HL2. I think he pretty much summarizes the feelings of a lot of us--that nVidia has lost trust with its drivers--even among developers. He'd seen the 50's and did not like them, and recommended against their use, even though nVidia declared the 45's "invalid" for use with HL2. Declaring a set of existing, official, publicly available DX9 drivers for a DX9 product line "invalid" for a DX9 game--a game which the existing Catalysts run on the DX9 code path in the game without difficulty--has to be one of the most bizarre things I've ever seen.

 

[Bjorn]

I thought this was rather interesting:

The only thing they explained is the fog. NV told me that it´s a bug and it only happens on a 5600 because the driver they used was only tested on 5900 that time. That info might be correct or incorrect.

If this is correct, then i would like to know what other problem the Det 50's have since the only thing that we know of now is the fog problem. But, i guess that both the game and the drivers will be released soon....

 

[Dave Baumann]

If this is correct, then i would like to know what other problem the Det 50's have since the only thing that we know of now is the fog problem.

They are not publicly released drivers...?

 

[Bjorn]

If this is correct, then i would like to know what other problem the Det 50's have since the only thing that we know of now is the fog problem.

They are not publicly released drivers...?

That's true

 

[WaltC]

I thought this was rather interesting:

The only thing they explained is the fog. NV told me that it´s a bug and it only happens on a 5600 because the driver they used was only tested on 5900 that time. That info might be correct or incorrect.

If this is correct, then i would like to know what other problem the Det 50's have since the only thing that we know of now is the fog problem. But, i guess that both the game and the drivers will be released soon....

Here's the thing about that which I find puzzling...nVidia sends betas to Valve with the instructions to use them to run HL2 instead of nVidia's currently released DX9-supporting Dets, and declares those drivers (45.xx) "invalid" for use with HL2. Why are the 45.xx's invalid?

The presumption seems to be that the 45.xx's have not been "optimized" for HL2 and are therefore invalid for use with it. The presumption is also that the 50's are, in fact, optimized for HL2 and should be used with HL2. This implies that nVidia is privy to some HL2 code in some fashion, otherwise optimizing its 50.xx driver set for HL2 would not be possible.

So, how is it that nVidia was able to optimize the 50.xx driver set for HL2 and recommend its use over the 45.xx driver set, but yet not catch the fact that the 50.xx's have a "bug" which doesn't render fog properly (if at all) in HL2? I would presume that if they were able to optimize the 50.xx's for HL2 to the extent claimed--such that they disown their existing driver set for use with the game--they would have noticed the problem with fog prior to sending the drivers along to Valve with their recommendation to use those drivers.

I used to float the comment--as a joke--that nVidia floated cheat-laden betas to developers to gauge which cheats they would discover and which they wouldn't, so that they could leave in the overlooked cheats and claim that the "found" ones were "bugs in the beta." Now I'm wondering if that was as much a joke as I originally thought...