Sunday, July 20th 2008

GPU PhysX Doesn't get you to 3DMark Vantage Hall of Fame Anymore

With NVIDIA releasing their GeForce PhysX drivers, users of the PhysX accelerating GeForce cards were at an advantage over their Radeon counterparts, reason being that in a certain CPU test routine of the 3DMark Vantage benchmark, the physics processing abilities of the computer are tested, and since the physics API used happens to be PhysX, users of GeForce get higher scores despite not having a physics processor device such as an Ageia PhysX card. This differs from a real-life scenario where a GeForce accelerator does both graphics and physics and the overhead of physics processing affects the graphics processing abilities.

The relation of GPU acceleration for gaining higher 3DMark scores in physics tests has been controversial to say the least. Futuremark has now decided to update its Hall of Fame to exclude all results using PhysX on a GPU, simply because this was not how they intended it to work. It has also been updated to organise the results better for easier comparison. You will be able to use GPU physics processing to get a 3DMark score, you will not be able to make it to the Hall of Fame using it. You can use an Ageia PhysX card to assist your 3DMark score to make it to the Hall of Fame, as that's how Futuremark intended PhysX processing scores to assist your final scores. Source: NordicHardware
Add your own comment

62 Comments on GPU PhysX Doesn't get you to 3DMark Vantage Hall of Fame Anymore

#26
mullered07
the whole point futuremark are making is that the physX test in 3dmark vatage were made for the cpu not a gpu and it doesnt represent real world gaming as the nvidia gpus are only being used for the physx in the test and not for rendering graphics at the same time which is what would happen in a real world scenario (ie gpu would be rendering graphics and physx at the same time)

and ati gpu are fully capable of doing physx only they would have to create there own api as cuda belongs to nvidia (who didnt create it either before the nvidia fanboys start)
Posted on Reply
#27
Kursah
mullered07, post: 893981"
the whole point futuremark are making is that the physX test in 3dmark vatage were made for the cpu not a gpu and it doesnt represent real world gaming as the nvidia gpus are only being used for the physx in the test and not for rendering graphics at the same time which is what would happen in a real world scenario (ie gpu would be rendering graphics and physx at the same time)
This is what dissapointed me most of FM...they knew this was coming, couldn't have been completely blind to it and should've done another physX test that stressed modern GPU's on the PhsyX and Rendering at the same time for a more proper score...sure it might've been headscratching at release, and maybe they didn't know or didn't have the proper tools...if not then an update will hopefully truly be in the works to alleviate this. I'd like to see a test just for simultaneous Rendering and PhysX GPU...I don't care as much about benches or scores, but to tout a bench that should be better at predicing modern performance, it should have this test in it.

Hopefully we'll see that happen! :toast:
Posted on Reply
#28
candle_86
personally whats the issue, the GPU was designed to act as a Physx card, so whats the issue you dont have to spend 100 bucks for the card now?
Posted on Reply
#29
Darknova
candle_86, post: 894028"
personally whats the issue, the GPU was designed to act as a Physx card, so whats the issue you dont have to spend 100 bucks for the card now?
That's not the issue.

The issue is that a CPU physics test is being offloaded onto the GPU. In Futuremarks rules this is NOT allowed. Therefore its technically cheating.
Posted on Reply
#30
Megasty
Darknova, post: 894033"
That's not the issue.

The issue is that a CPU physics test is being offloaded onto the GPU. In Futuremarks rules this is NOT allowed. Therefore its technically cheating.
I don't care about the cheating. I just want them to get it right for the games.
Posted on Reply
#31
Darknova
Megasty, post: 894037"
I don't care about the cheating. I just want them to get it right for the games.
Same lol, but that's the topic of this discussion is it not? :)
Posted on Reply
#32
Megasty
Darknova, post: 894045"
Same lol, but that's the topic of this discussion is it not? :)
Yeah, but that is FM's problem. Its plain to anyone to see who is using PhysX & who isn't. The GPU score isn't affected by it but the overall score is. FM turned a blind eye to it at first but its still cheating, so they had finally step in, especially when it was making such a ridiculous difference with the CPU score :shadedshu
Posted on Reply
#33
H82LUZ73
this is from Driver Heaven`s R700 preview test setup pic.......notice the blue card?
Posted on Reply
#34
mullered07
candle_86, post: 894028"
personally whats the issue, the GPU was designed to act as a Physx card, so whats the issue you dont have to spend 100 bucks for the card now?
read my last post :slap:

no one is moaning the gpu can do physics but in the cpu physx test the gpu is soley using physx, its not rendering anything else like it would do in game so therefore doesnt represent real gaming/benchmarking(in a game the gpu would be used to render the game aswell as using physx thus taxing it more ), if people are to stupid/ignorant to actually read the article and understand what fm are on about they really shouldnt be commenting on the topic at hand
Posted on Reply
#35
Hayder_Master
guys i just want to ask what is the last physics card release , last one i see it is asus p1 , anything new , and i see someone say the physics on nvidia gpu better than physics card is that true
Posted on Reply
#37
Hayder_Master
as i told before , only a program can help physics on ati , just like some guy do it before and go more than p22000 score in vantage with 4870 , and nvidia got the guy and he work with nvidia now , aslo i sure for something which is it my frien want to buy a program from nvidia to help him render 3d max projects on gpu , we now it is more fater than cpu , and am sure physics is same thing
Posted on Reply
#38
1c3d0g
This is why I dislike Futuremark and their stupid benches so much. It's just some rabid fan boys trying to measure who's e-penis is the biggest, but at the end of the day, what did they really "win"? Even if they get the highest score, they're still retarded...nobody with an ounce of sanity wastes so much time and energy into such a pointless benchmark.
Posted on Reply
#39
mullered07
i must say 3dMark vantage looks like a piece of shit to me anyway, god knows who they had coding that for them but that first test "jane nash" looks absolutely awful, i think all they have really done is concentrate on sm3/4? shaders and giving everything that gay unrealistic looking glow and thought bollox to the rest. im not impressed and feel its more of a tideover until 3dmark09/10? comes out.
Posted on Reply
#40
Evo85
People get waaaay to bent out of shape over this kind of thing. Must be a E-peen thing.. :slap:

I have always look at benchmarking as a way to judge changes I make to my system. OC this,
add a better cooler there, run Futuremark, see what the difference is. What matters to me is real world performance. As long as my system is capable of running what I throw at it, and is rock solid I dont really pay to much attention to the numbers...
Posted on Reply
#41
warhammer
3dMark vantage is not real world gaming or performance..
Posted on Reply
#42
farlex85
warhammer, post: 894227"
3dMark vantage is not real world gaming or performance..
Real world gaming is just another benchmark, and is subject to the same biases and differences that 3dmark is. I really don't know why that argument always gets brought up. Just b/c a game is popular doesn't make it a better bench than a program like 3dmark.

I personally find it a little hard to believe fm didn't intend for this sort of effect on scores, as physX is built into the final and difference making test. PhysX being there completely changes the way the test is run. Did they not know nvidia was putting physX on their cards? Did they think this would just be for dedicated physics cards? I doubt it. The bench really is done poorly, and planned very poorly. That last test should be a seperate category for physics, calling it a cpu score is a cause for all the frizzy. I also agree w/ mulder, I think it doesn't look very good at all. They need to figure out something new to accomodate this changing graphics processing arena.
Posted on Reply
#43
mullered07
farlex85, post: 894288"
Real world gaming is just another benchmark, and is subject to the same biases and differences that 3dmark is. I really don't know why that argument always gets brought up. Just b/c a game is popular doesn't make it a better bench than a program like 3dmark.

I personally find it a little hard to believe fm didn't intend for this sort of effect on scores, as physX is built into the final and difference making test. PhysX being there completely changes the way the test is run. Did they not know nvidia was putting physX on their cards? Did they think this would just be for dedicated physics cards? I doubt it. The bench really is done poorly, and planned very poorly. That last test should be a seperate category for physics, calling it a cpu score is a cause for all the frizzy. I also agree w/ mulder, I think it doesn't look very good at all. They need to figure out something new to accomodate this changing graphics processing arena.
its mullered :laugh::toast:
Posted on Reply
#44
farlex85
mullered07, post: 894302"
its mullered :laugh::toast:
:laugh: Sorry. I confused you w/ a paranormal investigator. :D
Posted on Reply
#45
btarunr
Editor & Senior Moderator
H82LUZ73, post: 894057"
this is from Driver Heaven`s R700 preview test setup pic.......notice the blue card?

PhysX driver allows you to use either the Ageia card or GeForce physics, not both. So if they use a test bed with a Ageia card, it makes for an even (fair) ground in which to test R700 against GTX 280 or any other card, since GTX 280's physics abilities won't be used by the driver, in each test, the PhysX card adds to the score.
Posted on Reply
#46
Wile E
Power User
I think it's silly to not include the Physx scores, so long as you are using FM approved WQHL drivers.

Banning gpu Physx in Vantage is like banning quad core cpus in 06. Going from a 3Ghz dual core to a 4GHz quad doesn't change your gaming experience at all, but it sure as hell boosts 06 scores.

I think if they are gonna ban hardware that gives an "unfair advantage" in their benchmarks because "it doesn't reflect real world scenarios", then they should do it with all of there benchmarks.

It's not the decision that upsets me, it's their lack of consistency in this. If you apply an "unfair advantage" rule to one bench, you should do it to all of them.
Posted on Reply
#47
tkpenalty
Wile E, post: 894475"
I think it's silly to not include the Physx scores, so long as you are using FM approved WQHL drivers.

Banning gpu Physx in Vantage is like banning quad core cpus in 06. Going from a 3Ghz dual core to a 4GHz quad doesn't change your gaming experience at all, but it sure as hell boosts 06 scores.

I think if they are gonna ban hardware that gives an "unfair advantage" in their benchmarks because "it doesn't reflect real world scenarios", then they should do it with all of there benchmarks.

It's not the decision that upsets me, it's their lack of consistency in this. If you apply an "unfair advantage" rule to one bench, you should do it to all of them.
Think of this. One system with one 8800GT. When you're gaming I doubt the GPU will simultaneously process physics while processing graphics. FM basically are saying that and thats why the GPU score is invalid.

Okay so lets say you have one 8600GT for Physics and one 8800GT for Graphics.

Going from 3Ghz to 4Ghz DOES give performance gains with Core 2. GRID goes from moderately smooth 30-40 FPS up to 50-60 FPS.
Posted on Reply
#48
Wile E
Power User
tkpenalty, post: 894517"
Going from 3Ghz to 4Ghz DOES give performance gains with Core 2. GRID goes from moderately smooth 30-40 FPS up to 50-60 FPS.
I'm sorry TK (and no offense intended), but I don't believe that. I think you are simply mistaken. I would need proof. I've already seen countless benchmarks that say otherwise, even in some of my own testing.
Posted on Reply
#49
tkpenalty
Wile E, post: 894532"
I'm sorry TK (and no offense intended), but I don't believe that. I think you are simply mistaken. I would need proof. I've already seen countless benchmarks that say otherwise, even in some of my own testing.
Consider the chipset, and memory that you are using first. This RAM doesn't yield good memory performance; latencies are fairly high with rather weak read/write.

I take no offense, and in your case its different its like comparing apples to oranges.
Posted on Reply
#50
Wile E
Power User
tkpenalty, post: 894547"
Consider the chipset, and memory that you are using first. This RAM doesn't yield good memory performance; latencies are fairly high with rather weak read/write.

I take no offense, and in your case its different its like comparing apples to oranges.
Well, my system doesn't suffer from high latencies and low read/write speeds. But that's neither here nor there, I'm not speaking purely from my own personal experience. I'm speaking about the multitude of benchmarks available in the wild that do not support your claims. The only 2 exceptions are games like SupCom, and where your res is so low the gpu isn't fully stressed.

If you are running at that low of a res, your framerates are gonna be so high to start with, that the frames given by OCing your cpu won't be noticeable at all, unless you are specifically measuring them. AKA: Going from 200fps to 240fps.
Posted on Reply
Add your own comment