Here We Go Again: Modders Reenable ATI/NVIDIA PhysX Combination

Here We Go Again: Modders Reenable ATI/NVIDIA PhysX Combination

Last week, we covered NVIDIA's decision to disable PhysX support if the NV drivers detect the presence of an ATI product. Jump ahead seven days, and the utterly predictable has happened—end-users have found a way around NVIDIA's lockout and have re-enabled the one-two punch of ATI graphics and NVIDIA PhysX. Quick, kids: Can any of you guess what happens next?

Over at NGOHQ, forum member Regeneration is claiming that the NVIDIA drivers published after the 185.85 WHQL series (this includes 186.18 through 191.07) also disable Ageia's original PPU add-on board if a non-NVIDIA GPU is installed. We've yet to see concrete proof that this is true; if you're one of the handful of people who own an Ageia PPU and don't mind testing it, let us know what the results are. Presumably, even if this is occurring, it's a bug—removing functionality on what amounts to another company's product is ballsy, even for the boys in green.

Disabling ATI cards is one thing, turning this puppy off is something altogether different...

The market's quick correction to NVIDIA's ATI driver lockout could pit the two sides in a constant game of one-upsmanship that NVIDIA is virtually certain to lose. There are ways for the manufacturer to assert control in the short term, such as requiring a specific driver to play a newly released game, but there's no way to maintain that lock long-term unless the company is willing to spend additional resources constantly plugging the holes the online community will be gleefully drilling into the company's "CUDA only" requirement. We understand why NVIDIA wants to keep PhysX to itself, but all the wishing in the world won't build a hack-proof driver. For now, the company has two options: Accept reality, and spin the option as a value-added feature (thereby pumping up PhysX and possibly encouraging its adoption), or battle the modders tooth and nail.

More Fermi. Less fighting.
0
+ -

Nvidia will almost certainly patch it out with their latest drivers, so the best we can hope is that the good buys keep fighting the good fight and trying to hack around it. What's sad is that Nvidia is actually hurting sales with this. People with a Radeon might buy a 9600GSO or something for Physx, but now they can't.

Physx, CUDA, GPGPU, Nvidia is stressing these because they are the only reasons to buy Nvidia outside of if you have an SLI board only. By this I mean the Nvidia equivalent always costs more (ex. 260 to 4870) but performs pretty much the same.

0
+ -

Lol, Nvidia is making a fool of themselves.

Crisis Causer, Nvidia has usually had the most powerful cards on the market. Even now the 295 is the best one, albeit it's a dual GPU solution and sucks much more power than the 5870.

They need to concentrate on the 300 series and let the rest sort out for itself.

0
+ -

"Nvidia has usually had the most powerful cards on the market. Even now the 295 is the best one, albeit it's a dual GPU solution and sucks much more power than the 5870"

True, but they've priced them out of reach of allot of people's means and they always have.

That's why ATI sells so well. They always have a product that is barking at the heels of NVIDIA's best offering, but is far less expensive.

Those of us with little to spend see allot of beauty in those affordable, capable 5870 cards.

0
+ -

>> "Disabling ATI cards is one thing..."

The drivers never did, nor now do, that.

People are blowing this out of proportion and implying nefarious intentions where there are none.  Apparently nVidia's seeing PhysX bugs when their cards are used as PPU's in combination with non-nVidia GPU's.  So, nVidia made the drivers disable PhysX support on the *NVIDIA* card if a non-nVidia GPU is being used... simply to avoid having to support a billion combinations for which nVidia can't possibly do QA testing.

That's it.  They're not doing anything that pushes others out, they're making their own product less useful for the few odd people that may have mixed mfg cards installed, for the benefit of more focused support.

If someone wants to use a hack to re-enable mixed-mfg configs, more power to them.  nVidia probably won't even do much to prevent this, since anyone using the hack would already be aware that they can't get any nVidia support for such a configuration.

0
+ -

3vi1,

We've asked NVIDIA to provide even a single case where using an ATI card + NV card for dedicated PhysX created a problem. As of this writing, they have not--and if you read their press statements, they're pretty clear that this is a move to protect their bottom line, not their product stability.

0
+ -

Well that's very interesting information - you should have put that in the article!

I never assume bad intentions when there's a simpler technical explanation, but it does seem odd that they didn't give you guys any reference-examples of problems they might be avoiding supporting.

0
+ -

is it too much to ask that they grow up and quit cutting out customers and market share for the illusion of control?

Login or Register to Comment
Post a Comment
Username:   Password: