NVIDIA, MSI, and Lucid speak about Hydra delay

[Rollo]

(and PhysX lock out)

http://www.brightsideofnews.com/news....aspx?pageid=0

Well, there you have it. I had been told by email a couple weeks ago NVIDIA had nothing to do with Lucid's delay, and that they had nothing against it.

Which makes sense. I'd think NVIDIA would embrace Lucid because anything that gets people to buy multiple NVIDIA cards is good for NVIDIA and their partners.

[badass]

(and PhysX lock out)

All well and good but there's nothing to stop them from having a tickbox that anbales the Physx when an ATI card is present and warns the user that it is an unsupported configurration.

[Biscuit]

Well its a start... i think the Fuzion will be my next motherboard

I still dont buy the whole testing configurations' thing for locking out the AMD cards. All they need to do is say 'we do not officially support this configuration' then its the consumers problem and not the corporations

[Pob255]

Unfortunately many people don't bother to read or pay attention to that sort of stuff.

If Ati brought out a new set of drivers that caused a conflict with PhysX on a 2nd nvidia card you can bet a lot of people will blame nVidia reguardless of the fact they ticked a box that read 'we do not officially support this configuration'

[Syllopsium]

what a load of rubbish. what next - Nvidia disabling certain network cards or sound cards because they havent been tested? Tough luck if it did conflict - its the job of each manufacturer to maintain device independence

[scaryjim]

Given that separate PhysX boards were available, and could be used with any GFX cards, long before NVidia bought and locked down PhysX to its hardware, I see no reason to believe that an ATI driver change would suddenly "break" PhysX, any more than any other driver would. If that's *genuinely* the reason NVidia block PhysX in ATI-containing systems, then why don't they do this for any other hardware type - "Audigy soundcard? Your drivers might break PhysX! No PhysX for you...". I'm sorry, as a CYA story it has a vaguely plausible ring to it, but in real-world terms it makes no sense. Are NVidia claiming they have tested every possible hardware combination *apart* from those containing ATI graphics cards? Because I find that highly unlikely...

[Pob255]

Well that because the audigy sound is still running on seperate hard ware, not running through the graphics card.
And it could happen, if the other graphics drivers are trying to make a call to the same address as the graphics card running the physx is using.
Ok it's unlikely and such a blanket call would be highly sloppy codeing but it could happen.
They don't have to test "every" combination, just the major ones and adding ati card's into the mix would near enough double the test load.

I agree it's a poor execuse, but it is understandable, it's a covering there own butts decision, however the general stiuation made worse by past nvidia practises and attitudes.
It's from there past attitude that this whole "it was nvidia I tell ya" Hydra situation started.

[badass]

Unfortunately many people don't bother to read or pay attention to that sort of stuff.

If Ati brought out a new set of drivers that caused a conflict with PhysX on a 2nd nvidia card you can bet a lot of people will blame nVidia reguardless of the fact they ticked a box that read 'we do not officially support this configuration'

Maybe a registry hack to enable then. Make it suitably difficult to do so that only enthusiasts will bother.

[Syllopsium]

It's not understandable. It's a convenient and bogus excuse full stop.

What the heck does 'if the other graphics drivers are trying to make a call to the same address as the graphics card running the physx is using' mean? You're not a programmer, are you?

PK

[shaithis]

The Physx lock-out is a BS excuse and they know it. They are just trying to force a second nVidia purchase.

As others have said, let us enable it as an unsupported feature.......instead of running 3rd party hacks to get it working.

What do nVidia think is worse? Them allowing an unsupported config or using a 3rd-party hack to enable the feature?

[jim]

At the end of the day, if they're being beaten on price and performance fronts by ATI then the only things they've got to hang on with are CUDA, physx and fanboyism.

By restricting physx to nVidia-only cards, they'll ensure that people look to the GTX285 and 295 for top-notch performance + physx, netting them some nice profit, rather than the far better value solution of a dirt cheap 4870 with a second-hand GeForce 8800, netting them nothing.

Until they can match ATI on performance and price again, I don't think anything's going to change re: physx unless they're given no choice in the matter.

[scaryjim]

Hmmm, consumer law test case? "I bought this NVidia graphics card as it promises PhysX support but when I put it in my computer the PhysX doesn't work" - isn't that kind of false advertising? Or do all of NVidia's product descriptions say "Amazing PhysX effects - but only if you're running the card in an NVidia appropved hardware configuration!"?

[NVIDIA_tap]

Hi All, my name is Tom Petersen and I work for NVIDA. I did the interview on BSN and I help manage SLI and PhysX here at NVDIA. Reading this thread has been interesting and I thought I would dive in and give you guys a chance to ask me some questions about all this. I hate to see folks speculating about our motives and thought maybe I could help explain why we do what we do.

So keep it fair and reasonable and ill try to respond on this thread.

[Biscuit]

ok - why exactly cant Nvidia allow the functionality of Physx with an ATI card and step over the unfortunate 'unable to test all the configurations' hurdle using a simple 'I Understand this configuration is not guaranteed to work as it has not been officially tested by Nvidia?' tick box. Surely this would cover your need to actually test an insane number of set ups all and would be way better for all of us 'gamers', which is what Nvidia seem to express as such a vital part of their job ethic.

Please dont coat us in mumbo jumbo sales or legal speak here...

[sammorris]

I know for a fact nvidia won't want Lucid to go ahead. Nvidia owner means another nvidia card for SLI, unless they're on Lucid, in which case they're open to switch brands. ATI will also be a bit miffed by this. It also means no SLI licensing for motherboards, no nvidia proprietary chipsets being sold, and no more beefing up TWIMTBP for high-end games - if a game is demanding and heavily nvidia-biased, just get a cheap nvidia card to be the primary and beef up the rest of the performance with a bigger ATI card, job done. No two ways about it, without accusing either company of anything, there's clearly motive there.
I was initially strongly taken aback by the website semiaccurate, until after reading the rest of their articles it became apparent just what the site was about. I know I'm a little biased but good lord, there's some serious fanboyism going on there.
However, what is also almost set in stone is that if Lucid gets released to market, I, an ATI owner, will be buying an nvidia card - Fermi or No Fermi, a sale they would not have had otherwise.
Quite frankly, I am thoroughly disillusioned with nvidia's business practices and have been since the Geforce FX series. Whilst this isn't enough to absolutely stop me buying a geforce, ATIs were never bad enough to force me into going with nvidia graphics over that 'prejudice'. Today nothing has changed, I still don't like nvidia, their products are, with a few exceptions, of a high standard. It's the company itself that put me off. When all is said and done though, being able to use both manufacturer's cards in tandem in the same PC solves such a ridiculously high number of problems I've come kissing nvidia's toes (Again, provided the Lucid boards are a success - I guarantee I'm far from alone in praying that'll be the case)
Summary of my thoughts on this:,
Semiaccurate's accusative article comes out: I believe nvidia are partly responsible for the delay
Nvidia deny reports: I still believe nvidia are partly responsible
Preliminary Benchmarks released of Lucid Platform: I no longer believe nvidia are responsible

The figures speak for themselves. The system does work, and nvidia's end of it is doing quite well. Fact is, the system was clearly shown to have bugs and some scaling issues. Knowing this as fact over rumour, there is a perfectly logical reason for the system to have been delayed, and for nvidia to have nothing to do with it. Now here's hoping the system works fully with dual GPUs, and that it's released earlier in Q1 rather than later...

[NVIDIA_tap]

@Biscuit - I think we can all agree that NVIDIA can choose to not spend our investment dollars to QA ATI + PhysX correct? My experience has been that if we don't QA someting inevitably it will break. When it breaks comes the real issue.

Let's say the function breaks because of a game update, or a dirver update, or a PhysX update, or a windows patch - and it starts ransdom crashes in a host a physX titles. This will result in a few things:
1. Users get unhappy and look for a fix
2. Game developers get unhappy about using PhysX
3. NVIDIA is put back in the line of fire for nefarious actions against the ecosystem.

Since we don't support it (and could easily and inadvertantly be the cause of the break) we create dissatisfied customers since NVIDIA would have no plans to fix it.

Instead of this, NVIDIA has choosen to proactively communicate to the world that we don't expect this to work long term. We knew some folks would be disappointed now, but it is my belief that we are saying a lot of pain later. As more and more PhysX titles come out it will become more and more difficult to turn off features. In effect we are taking our medicine now.