Nvidia drivers give lower performance than AMD's due to high CPU overhead/inefficient

[Grooveriding]

Some recent tests have turned up some more information on what some reviews have shown prior. Nvidia's drivers use more CPU power than AMD's drivers do, resulting in lower performance and framerates.

A recent review using 3DMark 11 highlights this issue and Futuremark confirms that Nvidia's drivers are less efficient and chewing up more CPU cycles resulting in lower FPS than AMD's.


http://www.lostcircuits.com/mambo//index.php?option=com_content&task=view&id=97&Itemid=1

Futuremark's conclusion about the CPU overhead of Nvidia's drivers:

'While the main component of the Physics test is the rigid body physics simulation running on the CPU, the rendering portion is also still present in the test. It appears that the drivers from different vendors have different overheads when rendering a large number of objects and especially when the rendering is threaded to multiple cores. In your test case it looks like the display driver overhead with the GTX 460 is the cause here and the high number of threads also adds to the overhead. With a 4-core system the overhead would likely be smaller and the performance difference would also be narrower.'

 

[Ben90]

While I'm aware that Nvidia's drivers are more resource intensive, I believe there is more going on here than JUST that. I'll do some experimenting when I get home and let you guys know my findings.

 

[BFG10K]

This has been a known issue since Fermi’s launch. Also Fermi has poor OpenGL performance. As an example, my Radeon 6850 outruns my GTX480 in Doom 3 (2560x1600 with 8xSAA) and Quake 4 (2560x1600 with 2xMSAA).

 

[tviceman]

So if Nvidia's drivers were as efficient as AMD's, then the Fermi architecture would be even faster than it is now while simultaneously consuming less power?

 

[Fox5]

Maybe it's a trade off, nvidia is using the cpu to perform more graphics optimization, assuming the bottle neck rests with the gpu?

Not a big deal to me, most games don't make use of a quad core fully anyway, let nvidia use a core or two for their drivers.

 

[Grooveriding]

So if Nvidia's drivers were as efficient as AMD's, then the Fermi architecture would be even faster than it is now while simultaneously consuming less power?

Yes, seems nvidia has lots of room for improvement in catch up to AMD's drivers. Not only in getting their drivers more efficient to reduce CPU overhead, but also would be nice if they could improve SLI scaling to be as good as Crossfire scaling is with AMD's 6 series. ()

 

[Skurge]

While I'm aware that Nvidia's drivers are more resource intensive, I believe there is more going on here than JUST that. I'll do some experimenting when I get home and let you guys know my findings.

Thanks. It would be good to know if a 6850 would be better for a system with maybe a core2duo over a 460. Seeing as most review sites use heavily clocked i7s so you cant see the drivers would make a real difference.

 

[Obsoleet]

This is old news, it's been well known that in a post-Vista world (2007 or so), AMD has the industry's gold standard drivers.

 

[Grooveriding]

This is old news, it's been well known that in a post-Vista world (2007 or so), AMD has the industry's gold standard drivers.

 

[Wreckage]

So this is based off of 1 part of 3DMark11? Most likely a problem with 3DMark, I'm sure they will hot fix it.

Way to grasp at straws though.

 

[Obsoleet]

So this is based off of 1 part of 3DMark11? Most likely a problem with 3DMark, I'm sure they will hot fix it.

Way to grasp at straws though.

Thought you had a big problem with hot fixes though? Sounds like an unacceptable solution to a guy like you.

 

[LiuKangBakinPie]

Eh what profile did they use to run those tests on? Yes its got profiles for most apps benches and games. Then theres Nhancer so I say that test is a crock of dung in my neighbours garden which I left there this morning

 

[Grooveriding]

Eh what profile did they use to run those tests on? Yes its got profiles for most apps benches and games. Then theres Nhancer so I say that test is a crock of dung in my neighbours garden which I left there this morning

This is a known issue and your post is somewhat hard to decipher.

 

[LiuKangBakinPie]

Actually the security suites of KIS and Norton killed Vista and gave vista a bad reputation

 

[RussianSensation]

The thread title is misleading. Nvidia's drivers do not result in lower performance relative to the competition:

GTX580 has no single GPU competition.
GTX570 competes well with HD6970
GTX560 TI competes well with HD6950
GTX460/470 compete well with HD6850/6870.

The title should be changed to something like "NV's videocards are more dependent on CPU speed to extract maximum performance compared to AMD's cards." That's not the same as saying NV's graphics cards are slower than AMD's because their driver is less efficient.

The article itself is very weak in trying to portray this idea. Xbitlabs produced a far more comprehensive article on this topic many months ago.

 

[Wreckage]

http://www.guru3d.com/news/nvidia-grabs-71-of-discrete-graphics-card-market/

NVIDIA grabs 71% of discrete graphics card market

Hmmm... it does not take a physics professor to do the math on that one. :hmm:

The title should be changed to something like "NV's videocards are more dependent on CPU speed to extract maximum performance compared to AMD's cards." That's not the same as saying NV's graphics cards are slower than AMD's because their driver is less efficient.

Also add "in one test in 3DMark11"

 

[Barfo]

This has been a known issue since Fermi’s launch. Also Fermi has poor OpenGL performance. As an example, my Radeon 6850 outruns my GTX480 in Doom 3 (2560x1600 with 8xSAA) and Quake 4 (2560x1600 with 2xMSAA).

Funny, didn't nvidia used to lead ati/amd in opengl performance all the time before this?

 

[Grooveriding]

I meant it in the context of lower performance due to cpu overhead. Not overall. Your wording wouldn't fit.

There is also the context of Nvidia's drivers consume more cpu power than AMD's causing lower framerates due to taking cpu resources away from the game its self.

The thread title is misleading. Nvidia's drivers do not in result in lower performance relative to the competition:

GTX580 has no single GPU competition.
GTX570 competes well with HD6970
GTX560 TI competes well with HD6950
GTX460/470 compete well with HD6850/6870.

The title should be changed to something like "NV's videocards are more dependent on CPU speed to extract maximum performance compared to AMD's cards." That's not the same as saying NV's graphics cards are slower than AMD's because their driver is less efficient.

The article itself is very weak in trying to portray this idea. Xbitlabs produced a far more comprehensive article on this topic many months ago.

 

[SolMiester]

Who was it that said this forum is full of NV fanboys....LOL, what a joke!


This is neither productive nor helpful for the thread's discussion.

Please keep your postings on-topic and avoid the inflammatory rhetoric.

Moderator Idontcare

 

[RussianSensation]

I meant it in the context of lower performance due to cpu overhead. Not overall.

The article shows that NV's driver has a higher overhead compared to AMD's in 1 test. They specifically look at "Physics" 3dMark11 test to highlight this difference. Ironically, NV graphics cards utterly dominate AMD in any game that actually uses PhysX effects (i.e., some attempt at realistic physics effects). So do you really care how good a CPU is in running physics when hardly any modern games run realistic physics effects on the CPU? In other words, if they wanted to look at CPU dependence by brand, they should have compared various CPUs in actual games. At least this way, C2D/Q, Phenom II owners, for example, would have a better idea if AMD's cards are a better fit for their slower CPUs compared to say a SB platform.

 

[notty22]

Its strange how Nvidia gpu's are all part of the
3DMark 11 Top 20 (Performance preset)


AMD gpu's / cpu's are nowhere to be seen. Seems like AMD needs to do some work .

 

[Skurge]

The article shows that NV's driver has a higher overhead compared to AMD's. They specifically look at "Physics" 3dMark11 test to highlight this difference. Ironically, NV graphics card utterly dominate AMD in any game that actually uses PhysX effects (i.e., some attempt at realistic physics effects). In other words, if they wanted to look at driver overhead, they should have looked at actual games not a "Physics" test.

Thats an odd example you give especially since AMD cards don't run physx at all. Im not sure why you are disagreeing with the assessment when the article you posted confirms these findings?

Im not saying anyones drivers are better here, I just find it interesting.

 

[Grooveriding]

The article shows that NV's driver has a higher overhead compared to AMD's. They specifically look at "Physics" 3dMark11 test to highlight this difference. Ironically, NV graphics card utterly dominate AMD in any game that actually uses PhysX effects (i.e., some attempt at realistic physics effects). In other words, if they wanted to look at driver overhead, they should have looked at actual games not a "Physics" test.

They used the physics test because it makes the highest demand of the cpu while continuing to stress the gpu which helps to highlight this issue with Nvidia's drivers.

The following is not directed at you. But some posters in this thread continue to provide examples of fanboy whining. If you can only discuss positive aspects of your fanboy brand of choice, why not stick to those threads and stay out of the ones that cause you so much personal distress over the negative issues with your card of choice.

 

[RussianSensation]

Im not saying anyones drivers are better here, I just find it interesting.

I am saying that they show reduced CPU performance in a 3dMark11 physics test. But when was the last time you ran hardcore physics effects on your CPU? The whole point of testing driver overhead is to test CPU dependence in real world games. I don't disagree that NV's cards have steeper CPU requirements. I just think their example is meaningless for drawing any real world conclusions for actual games. The article I linked on the other hand shows exactly how the driver overhead impacts performance in favour of AMD.

 

[Obsoleet]

No. AMD drivers are better, many studies have proven this without a shadow of doubt.
That's why everyone does damage control stating they are "roughly the same, give or take". That's the best PR NV can put out there at this point in a losing battle in the driver quality competition.