Friday, May 28th 2010

NVIDIA Removes Restriction on ATI GPUs with NVIDIA GPUs Processing PhysX

NVIDIA has reportedly removed the driver-level code which restricts users from having an NVIDIA GeForce GPU process PhysX with an ATI Radeon GPU in the lead, processing graphics. Version 257.15 Beta of the GeForce drivers brought about this change. Possible commercial interests may have played NVIDIA's previous decision to prevent the use of GeForce GPUs to process PhysX with ATI Radeon GPUs, where users could buy an inexpensive GeForce GPU to go with a high-end DirectX 11 compliant Radeon GPU, thereby reducing NVIDIA's margins, though officially NVIDIA maintained that the restriction was in place to ensure Quality Assurance. The present move also seems to have commercial interests in mind, as NVIDIA could clear inventories of GeForce GPUs at least to users of ATI Radeon GPUs. NVIDIA replenished its high-end offering recently with the DirectX 11 compliant GeForce 400 series GPUs.

Update (28/05): A fresh report by Anandtech says that the ability to use GeForce for PhysX in systems with graphics led by Radeon GPUs with the 257.15 beta driver is just a bug and not a feature. It means that this ability is one-off for this particular version of the driver, and future drivers may not feature it.
Source: NGOHQ.com
Add your own comment

276 Comments on NVIDIA Removes Restriction on ATI GPUs with NVIDIA GPUs Processing PhysX

#251
Regeneration
NGOHQ.COM
newtekie1Actually, if you believe Regeneration from NGOHQ, it is a big conspiricy by nVidia...because nVidia has so much power they can get anti-virus makers to put false positives in their software...:rolleyes:
That wasn't just a regular false positive! It was a gamer-virus-false-positive. Jeeeee!! What a coincidence.

Instead of checking up those false positive reports, both sites rush to censor the word NGOHQ from their sites and warn users not to download anything from our site.

Those sites seem to think they should report any news that hits any site without investigation, hasn't reported the latest PhysX development - like all the other sites, and yet they report some random false positive virus responses to known wrappers.

Just recently Gabriel Torres from Hardware Secrets got blacklisted by Nvidia, because he refused to be their puppet.

So yes, I believe someone has intentionally tricked Symantec. Who? I don't know. But if it looks like a duck, sounds like a duck, good chances that... it is a duck.
Posted on Reply
#252
Mussels
Freshwater Moderator
newtekie1Shhhh...

It isn't the mod-maker's fault, all those anti-viruses are just shit.:rolleyes:
IMO... yeah, the ones that showed it as a virus ARE shit AV's. kaspersky, nod32, arent in there. neither is MSE for that matter.


still, with so many shit AV's on the market, it is best to pack things in a way they wont whine about.
Posted on Reply
#253
Wile E
Power User
MusselsIMO... yeah, the ones that showed it as a virus ARE shit AV's. kaspersky, nod32, arent in there. neither is MSE for that matter.


still, with so many shit AV's on the market, it is best to pack things in a way they wont whine about.
Exactly. I trust that it's clean, but many wouldn't. Put that up on Demonoid, for instance, and it would get nuked the instant somebody posted those VT results.

I agree, most of them aren't very good AVs, but that many hits is going to cause controversy.
Posted on Reply
#254
Unregistered
so how exactly this patch work?

and btw is this patch effect my save game?
#255
GenL
newtekie1You mistunderstand me.
I believe i understand you. Such rants always could be biased (i mean all sides - fud is everywhere, you know!), but everyone have to agree that it came from HardOCP at first. Yes, i disagree with some points Regeneration posted at NGOHQ about this, as i stated in my comments there, although they all seem to be valid for most users who don't deeply understand how AV software works. It's not a big deal.
The accusal without any proofs is what i really refuse to understand. Moreover, the original author of those claims refused to even try to get any proofs after i contacted him. So i'll just stick with a fact that all those tech-sites relations are serious business and i don't really need/want to understand their policies.
Wile Eyou really need to use a different packing method.
newtekie1It isn't the mod-maker's fault, all those anti-viruses are just shit.
I suggest everyone who wants to understand my point about this, to spend some time and read a topic at Rage3D: www.rage3d.com/board/showthread.php?t=33962732
My standpoint is fully explained in my posts there.
wahdangunso how exactly this patch work?

and btw is this patch effect my save game?
How it works, is explained just under the screenshots there at the original topic: www.ngohq.com/graphic-cards/17716-batman-arkham-asylum-msaa-fix.html
No, it shouldn't affect your savegames.
Wile EI want to see more comparison screen shots with and without in game AA, on both brands of cards at the same settings.
I would like to see it too. I can't do it by myself, sorry.
Posted on Reply
#256
newtekie1
Semi-Retired Folder
RegenerationThat wasn't just a regular false positive! It was a gamer-virus-false-positive. Jeeeee!! What a coincidence.

Instead of checking up those false positive reports, both sites rush to censor the word NGOHQ from their sites and warn users not to download anything from our site.

Those sites seem to think they should report any news that hits any site without investigation, hasn't reported the latest PhysX development - like all the other sites, and yet they report some random false positive virus responses to known wrappers.

Just recently Gabriel Torres from Hardware Secrets got blacklisted by Nvidia, because he refused to be their puppet.

So yes, I believe someone has intentionally tricked Symantec. Who? I don't know. But if it looks like a duck, sounds like a duck, good chances that... it is a duck.
Look at the god damn results, there are like 20 anti-viruses that report the false positive. Use a different wrapper that doesn't trip AVs, and stop all this BS nVidia is out to get us crap. Symantec certainly has no reason to bow down to nVidia and put false positives in their software because nVidia wanted them to. It is completely idiotic to even suggest that the false positive was caused by anything other then your use of a poor wrapper.

And Gabriel Torres deserved to be blacklisted. His reviews are complete shit, he wants nVidia to keep giving him free samples, but won't even put CUDA and PhysX in a feature list when asked. I wouldn't give him free shit either.
Posted on Reply
#257
Regeneration
NGOHQ.COM
newtekie1Look at the god damn results, there are like 20 anti-viruses that report the false positive. Use a different wrapper that doesn't trip AVs, and stop all this BS nVidia is out to get us crap. Symantec certainly has no reason to bow down to nVidia and put false positives in their software because nVidia wanted them to. It is completely idiotic to even suggest that the false positive was caused by anything other then your use of a poor wrapper.
You twisted my post. False positives usually look like this:

www.virustotal.com/analisis/59261f8a261705fab4b95723449c1d4e442ccc23cedec214b355c04157061fdb-1274174744

It is pretty simple to take a binary, bundle it with a trojan and submit it to AV vendors. Anyone can do it.
Posted on Reply
#258
Wile E
Power User
RegenerationYou twisted my post. False positives usually look like this:

www.virustotal.com/analisis/59261f8a261705fab4b95723449c1d4e442ccc23cedec214b355c04157061fdb-1274174744

It is pretty simple to take a binary, bundle it with a trojan and submit it to AV vendors. Anyone can do it.
If I didn't already know, test and trust the patch myself, and have a basic understanding of AV and packing, 11 hits earns it a nuke in my book. I won't mess with an 11 hit file unless I have no other choice, then I sandbox it first. 4 or 5 hits on the lesser AVs, and I might take a chance, start getting into double digits, and I'm looking for something else if at all possible.

Somebody donate him a better packer if need be, and the controversy disappears altogether. Shit, I'd donate to get him a better packer that doesn't throw a million false positives.
Posted on Reply
#259
GenL
Wile ESomebody donate him a better packer if need be, and the controversy disappears altogether. Shit, I'd donate to get him a better packer that doesn't throw a million false positives.
That's what i tried to tell at Rage3D - it really has nothing to do with money, as it won't help here.
Posted on Reply
#262
TRIPTEX_CAN
So I got some "new" shiny to play with. I'll be testing them tonight individually (rig in specs) but I'm a little worried about power draw. I think I'll most likely underclock the cards to around 500Mhz core. Do you guys think my system will be able to cope? Keep in mind my 5970 runs stock clocks.

Also assuming both cards check out and run perfectly there is a chance one will go up for sale if anyone is interested they can PM me.

Posted on Reply
#263
OnBoard
TRIPTEX_MTLI think I'll most likely underclock the cards to around 500Mhz core. Do you guys think my system will be able to cope?
Underclock memory as low as it goes and if you underclock core make sure it's not linked with shaders. Although it might have plenty of performance on lower shaders too. And yes, should cope. PhysX doesn't peak the cards power consumption.
Posted on Reply
#264
newtekie1
Semi-Retired Folder
Wile EIf I didn't already know, test and trust the patch myself, and have a basic understanding of AV and packing, 11 hits earns it a nuke in my book. I won't mess with an 11 hit file unless I have no other choice, then I sandbox it first. 4 or 5 hits on the lesser AVs, and I might take a chance, start getting into double digits, and I'm looking for something else if at all possible.

Somebody donate him a better packer if need be, and the controversy disappears altogether. Shit, I'd donate to get him a better packer that doesn't throw a million false positives.
You know, I would think it would be good practice to run any file you are going to distribute through that checker, if it generates false positives, then fix it before making it public...and of course don't fly off the handle with conspiricy theorys about everyone running a scam against you with nVidia at the top when reports start coming in about the false positives...

And isn't the 7zip packer free?
Posted on Reply
#265
dadi_oh
OnBoardUnderclock memory as low as it goes and if you underclock core make sure it's not linked with shaders. Although it might have plenty of performance on lower shaders too. And yes, should cope. PhysX doesn't peak the cards power consumption.
My GT240 with only 96 shaders is used less than 60% even in Batman Arkham Asylum. In UT3 Physx levels it tops at about 25% GPU usage. Fluidmark benchmark only gets it to about 45% usage. That suggests to me that your 112 cores on the 8800GT would be just fine running at a lower clock to reduce temps and fan noise (those single slot 8800GT coolers are noisy). And as suggested I don't believe memory bandwidth plays much of a role at all so you could downclock that a bit too.
Posted on Reply
#266
TRIPTEX_CAN
OnBoardUnderclock memory as low as it goes and if you underclock core make sure it's not linked with shaders. Although it might have plenty of performance on lower shaders too. And yes, should cope. PhysX doesn't peak the cards power consumption.
Thanks, I've never clocked an NV card before so I'm not really sure where to start. I have the feeling I can use EVGA precision with any NV GPU so I'll probably start with that. From what I understand physx power draw is pretty low and physx doesn't even saturate PCI-e x1.
Posted on Reply
#267
newtekie1
Semi-Retired Folder
dadi_ohMy GT240 with only 96 shaders is used less than 60% even in Batman Arkham Asylum. In UT3 Physx levels it tops at about 25% GPU usage. Fluidmark benchmark only gets it to about 45% usage. That suggests to me that your 112 cores on the 8800GT would be just fine running at a lower clock to reduce temps and fan noise (those single slot 8800GT coolers are noisy). And as suggested I don't believe memory bandwidth plays much of a role at all so you could downclock that a bit too.
I just did an interesting quick write up on this exact subject.

Using a 9600GT LP, which is already underclocked from stock 9600GT specs, with only 64 shaders, there was actually no performance benefit of having it in my main rig as a dedicated with a GTX470, tested with Batman:AA.

Underclocking the 9600GT as low as the sliders would let me go in MSI Afterburner dropped the FPS about 3-5FPS.

So even an 8800GT is probably overkill, even if everything is underclocked as low as possible.
TRIPTEX_MTLThanks, I've never clocked an NV card before so I'm not really sure where to start. I have the feeling I can use EVGA precision with any NV GPU so I'll probably start with that. From what I understand physx power draw is pretty low and physx doesn't even saturate PCI-e x1.
I'd try using the latest version of MSI Afterburner(1.6.0 Beta 6), it works with any nVidia cards like Precisions, but seems to work better with the latest drivers. I had some problems getting the latest version of Precision to read the clock speeds properly and actually overclock my GTX470 when I moved the sliders.
Posted on Reply
#268
TRIPTEX_CAN
I wonder if it will conflict with my 5970 though. Unless the latest AB 1.6.0 has a drop box to choose which card to configure... I use AB to clock my 5970 and I cant see how I would possibly differentiate between GPUs in the program.

The older precision should recognize the 8800GT regardless of drivers no? Unless in the case of your 470 precision worked until you changed drivers then that's another issue.
Posted on Reply
#269
newtekie1
Semi-Retired Folder
TRIPTEX_MTLI wonder if it will conflict with my 5970 though. Unless the latest AB 1.6.0 has a drop box to choose which card to configure... I use AB to clock my 5970 and I cant see how I would possibly differentiate between GPUs in the program.

The older precision should recognize the 8800GT regardless of drivers no? Unless in the case of your 470 precision worked until you changed drivers then that's another issue.
I only ever tried with the latest drivers, so I'm not sure, it could be the card or the drivers.

Afterburner has a place in the settings to switch GPUs, though I don't know if it will apply the different profiles/clocks to both cards at the same time, or if it will only work on one card at a time.
Posted on Reply
#270
TRIPTEX_CAN
newtekie1I only ever tried with the latest drivers, so I'm not sure, it could be the card or the drivers.

Afterburner has a place in the settings to switch GPUs, though I don't know if it will apply the different profiles/clocks to both cards at the same time, or if it will only work on one card at a time.
OK, I'll look for it but I plan to keep the 5970 stock for most of the tests tonight so I think AB should work.
Posted on Reply
#271
TRIPTEX_CAN
One care smoked when I powered up the system and when the other card was installed the system did boot but wouldn't display video.
Posted on Reply
#272
OnBoard
TRIPTEX_MTLOne care smoked when I powered up the system and when the other card was installed the system did boot but wouldn't display video.
Eew, now you have to merge them together to get one working card :)
Posted on Reply
#273
dadi_oh
TRIPTEX_MTLOne care smoked when I powered up the system and when the other card was installed the system did boot but wouldn't display video.
Probably not related but I did need to make dummy plug for the Physx card. Otherwise the nvidia driver disables it.
Posted on Reply
#274
erocker
*
dadi_ohProbably not related but I did need to make dummy plug for the Physx card. Otherwise the nvidia driver disables it.
If your monitor has two inputs you can just hook up the extra input to the PhysX card. Works for me. :)
Posted on Reply
#275
TRIPTEX_CAN
The card didnt post video from boot so I don't think drivers were the issue. I believe the second card is functioning but I didn't try it in the primary slot.

meh , i might try again later
Posted on Reply
Add your own comment
May 2nd, 2024 03:02 EDT change timezone

New Forum Posts

Popular Reviews

Controversial News Posts