Nvidia cheats on 3DMark with 177.39 drivers
Different workload, let's just pretend it's legit
NVIDIA IS VOILATING Futuremark rules with the latest PhysX drivers, and doing it in the usual sleazy way. The rules are simple, violating them isn't a trick, and doing so in order to pump up your numbers is the height of unethical behavior.
Well no, it isn't the height, this is, but calling Nvidia on unethical behavior is what you might deem a target-rich environment. That said, the company's behavior this time could be done in an ethical way, but it chose not to. The explanation needs a little background though, so bear with us.
3DMark Vantage has four major components, two CPU and two GPU. One of the CPU subtests is a physics-based test. The physics test is based on the Ageia PhysX API, a fairly widespread API in use by a large number of games. Between the time that 3DMark Vantage development was started and the time it was released, Nvidia bought Ageia.
The problem is that the PhysX DLLs, and for that matter, the whole API is now owned by Nvidia. In and of itself, this is not a problem, especially if the company involved had a history of honesty, integrity, and fair play. Nvidia has none of these attributes, and has a proven history of cheating on 3DMark.
To be fair, ATI has been caught at the same thing as well, but nothing lately, and Intel compilers come with curious optimisation defaults as well. No one is clean, but only Nvidia seems to take dishonesty as a corporate mandate.
So, with the latest driver, Forceware 177.39 drivers, Nvidia put its now in-house PhysX APIs into the drivers. Instead of it running on the CPU or on the PhysX chip, it is running it on the GPU. It owns the GPU and it's drivers along with the physics API and all those drivers. This is a dangerous situation.
There are two problems with Nvidia doing this, it isn't a legal driver for 3DMark, and it isn't even running the same program as others who run 3DMark. Either one is enough to preclude people from using those drivers and calling the results 3DMark scores.
If you look at the 3DMark Vantage Driver Approval Policy, section 3.5 clearly states, "Based on the specification and design of the CPU tests, GPU make, type or driver version may not have a significant effect on the results of either of the CPU tests as indicated in Section 7.3 of the 3DMark Vantage specification and whitepaper." When you run a CPU test on a GPU, it clearly violates the rules.
The other problem is that when you install the drivers, they replace the PhysX DLLs with a completely different set of DLLs. If you look at the Nvidia PhysX reviewers guide, a PDF that NV hands out to help people write up their newest toys, they say the following for UT3 installation.
That looks completely above board!
Note steps 4 and 5 that say "Uninstall the existing AGEIA PhysX v7.11.13 driver (installs with UT3 installation)." and "Install the new PhysX 8.06.12 driver." Same with 3DMark Vantage, and they offer the helpful hint of "GeForce PhysX is enabled in CPU Test 2. We recommend testing in Performance Preset for the best final score with GeForce PhysX. In Extreme Preset the score is mainly determined by the GPU score. A faster CPU Test 2 result will not make much difference."
This means two things, when you are running 3DMark Vantage with the 177.39 drivers, you are not doing the same work as every other driver running 3DMark Vantage. You are doing a completely different workload on 25 per cent of the tests. To rub salt into the wound, Nvidia then tells you that the Extreme preset, the one meant for high-end GPUs, doesn't show off the cheat sufficiently, so use one that weights it more heavily. What gall.
The end result is about a 10n per cent increase in scores, and a claimed 7.5x advantage on the physics subtest. You can see how they word it for yourself.
If the glove doesn't fit, you must acquit.
If you look at it as a whole, Nvidia is doing two things, first of all doing a different workload than the rest of the world and claiming it is the same, and second not following the Futuremark rules. This is nothing less than blatant dishonesty.
We asked Futuremark for a comment, and they referred us to the above Driver Approval Policy, and then to the approved driver page here.
Nvidia has not submitted 177.39 for approval, and likely will never do so because the chance of it being approved are something between zero and having to buy Futuremark. Until they submit a bad driver, no harm, no foul. The Futuremark policy is that if it isn't up on ORB, it isn't a 3DMark score, and that is quite sensible.
Nvidia engineers know that there is no way they can get this driver approved, so they don't try. They know they are not running 3DMark, but they don't even try to hide it. They are however disingenuously doing a different workload and trying to cynically pass it off as the same old workload. There is a word for this behavior, cheating. µ
Note: At time of this writing, the ATI Catalyst 8.6 drivers are not approved yet, the latest valid set is 8.5. Nvidia's latest approved is 175.16, and Intel isn't even submitting a set, it seems they don't want the world to know something. In any case, this does not preclude all three companies from quoting scores liberally with unapproved drivers.
Bad industry, no cookies for you.
Reviewers take note.
Comments
If you got it ,flaunt it
Nvidia owns Physx ,boo hoo ,if you got it flaunt it, you have every right to if you own something. This is deliberate obviously ,since Nvidia can flaunt those numbers ,wheres ATI cant ,simple as. Theres nothing new here and reviewers take an obvious note and move on.Oh Please
I am the biggest ATI fanboy on the east coast, nevertheless this article is bull. If Nvidia can offload the physx pipline off to the GPU and the user can see a dramatic gain in games without a loss of visual quality, I don't care how Futuremark has their legalese written about what is a valid score - there is significant and measureable improvement in physics calculations. In fact, someone who didn't have an Ageia card can now benefit in a way unheard of from all previous graphics drivers, as their card can do something that upon purchase, it was never intended to do!I will upgrade to the latest 4000 series this winter and stay true to my ATI fanboy status, but I will give credit where credit is due - Nvidia pulled a rabbit out of their hat with this one.
Bash the right people.
This is not something nvidia is doing wrong, this is the 3dmark people putting a CPU test in their suite and adding its result into a final score and then (many) reviewers using it to test graphics cards on themselves.They should have designed the test to separate the CPU/PPU and GPU related tests also in the final score, since they know what their test is predominately used for, namely testing graphics cards and far far less for testing the capacities of an entire system to run games (if ever).
Tests Compare
Will it Fly? Theory of cheat is good to know, NOT to be fooled. Ultie Tom Feels Vindicated in All Glory Possible, In Sense That Vantage is Vista ONLY.So There, Certain Raggettes need NOT Apply. However, Test is Merely Device to Create Comparisson of performances. If nvidia Test Different compnets of its software on radically differnt potential hardware parts, its NOT Same test, yet is that BAD?. Perhaps, yet perhaps deeper view into Vantage & Vista & Cards will be result, is that BAD? Perhaps not as much.
No TestBed Remains alone, although 3d has been ease of use, while game scores vary widely & high can fall to very low in different game. So Both & now: how does Nvidia Score indicate pushing score & harming play? or is it utilizing test from lost part, cpu into center of turmoil itself, gpu. migration of software test to harder GPUs' hardware Potential.Everyone knows Todays CPU/Main system is several years behind Game Card arena.
If you test all cards in all Vista types,thats one level, still if you thrown in Mass marketed computer with BIG Brand names, Scores go kaplunk, bottoming out into mans ancient trading Past. So Go Figure, its' One More help.
drashek
eh
As if 3DMark wasn't useless already.It's amazing that getting better performance now for free can be construed as a bad thing. Looks to me that aside from the hoops you have to jump through, the GPU solution is fair. When someone plays an actual game, they can take advantage of this. Why not in a benchmark?
Just because Futuremark didn't have the foresight to call a physics test a physics test and not a CPU test doesn't mean nVidia users should have to gimp their setup for a benchmark.
Was it cheating to use a PhysX card on this benchmark?
how is this cheating?
How anti-Nvidia can you be? If you had a AgeiaPhysX card installed would not you get a "inflated" score versus if said cad was not installed?If you have a game that used PhysX and it runs better on a Nvidia card now since they do the the PhysX calculations on GPU as compaired to the same game running on a ATI card that is not cheating on the part of Nvidia, it is a matter of ATI not doing the PhysX calculations.
If the purpose of 3DMark is to give you an idea of which GPU/CPU setup is faster overall and thus a better setup for gaming, would'nt it make sense to show that it produces higher scores with a Nvidia GPU that supports PhysX since a game that supported PhysX would also run faster?
3DMark, what an awesome game!
Nobody cares except you Charlie.Perfect
Now this is what I come here for, perfect article.Now lets all hold our breath untill the first nvidia fan boy comes flying in with excuses :)
And your point it?
To the writer of this article: did you zone out while writing this?Obviously, if you want to run physics from your GPU, then you would have to integrate the programming into your drivers.
But no, you don't want to run physics from your GPU and have better physics performance, because that would invalidate your 3Dmark score.
They innovate and improve overall system performance and you dislike their innovation because it temporarily conflicts with the process of making bragging rights?
So don't run 3D mark. Run actual games. So what?
Well...
Notwithstanding the ethics or lack thereof on Nvidia's part, doesn't this also raise the question of whether 3Dmark is any use at all if it's so easy to cheat?175.16 Drivers
I think it's worth pointing out that 175.16 which ARE verified, allow the PhysX panel/drivers to be installed too, apparently (I haven't tested I am on 177.39 myself). It would be interesting to see if the same boost comes in those drivers (which are also CUDA enabled), which with the panel installed, it should.But I don't quite get it, I mean is this not the point in the whole PhysX thing to begin with? Sure it's a test and they need rules, but its a Physics test specifically, and doesn't the score jump if you install an actual PhysX card into a system too? I fail to see the point really. Sure it maybe a 'cheat' in some senses, but it is doing what it was designed and coded for, which is physics calculations. It's not like they coded it specifically for that test to cheat, they have coded it so the GPU takes alot of the strain from the CPU which is supposed to be a good thing. It seems to work in the games that support PhysX hardware too, not just this 1 test.
Maybe I am missing something. I don't know.
Ladies and Gentlemen
I have actually come across some information that may surprise you, might not. Futuremark, prior to this article, had even contacted Charlie and flat out told him that Nvidia was not cheating. Yes, Nvidia needs to submit drivers for approval, but thats about the only part Charlie got right.This is yet another smear campaign by Charlie who holds quite a lot of hatred for the boys and girls at Nvidia.
Charlie's hatred of Nvidia is in fact, well, legendary. He never gets invited to events and he has his tantrums over it. He pretends to blow it off and say it doesn't matter to him, but then look what he does! My information is accurate folks. Right from the mouth of Futuremark. Charlie makes stuff up as he goes along. Somebody give this boy some Enphamil and a rattle!
I know you do this in part to get hits on your site, and at this point, I think the inquirer is too far gone for any sort of redemption. You'd have a long hard road to earn back whatever little credibility you once had. (And that was not much).
-Magnum Opus
How else are they going to prove that physx is better on the GPU??
I dont understand... how is this a different workload if the Physx API (& not its implementation) remains the same?The API implementation is part of the product here... and the same case applies to directx drivers which use JIT for generating native binaries.
If Physx runs faster on the GPU it should be run on the GPU. How else should the perf difference be made visible to the end user?
workload?
Does it actually do a different workload? (Meaning it actually skips some of the physics work) I mean if they bought the company and released faster drivers when using capable hardware, why is that wrong?I understand it's against their current driver policy. Maybe they (futuremark) should have included a separate physics section, then stressed the CPU with other core game logic like AI?
Also isn't Futuremark to blame here for not seeing this comming? We knew physics was moving to GPU for a while now. They should have updated their driver policy. There is nothing immoral about doing physics on the GPU? If Ghost Recon Advanced Warfighter runs faster now, that's good right?
When ATI gets HAVOC up and working I'd expect them to do the same. (maybe shame on Futuremark if they don't support HAVOC..)
Question....
So if I for example have a PhysXcard and update to these drivers, will my card no longer be used? Will the Vid card assume all physX calls?
Take those Vantage scores with a bucket o salt.
Armed with this knowledge, we now know that true performance cannot be measured by a systhetic benchmark alone but on a variety of games (as if we didnt already know that).Intel has nothing to do with this...
Why do you have to mention intel in almost every article even when they have nothing to do with it? Intel bought havoc, yadda-yadda. But intel doesn't really have any good graphics part. Oh, I know, they will have something gooood in the future. Yeah right, AMD and Nvidia will also have somethin way better than intel in the future. So stop pretending that intel has something up their sleeve coz' it's more like something up their "behind"!...and Intel isn't even submitting a set, it seems they don't want the world to know something...That's right, they don't want the world to know that they suck asz at the graphics game.
Rules are made to be Broken
Who cares!. SO what,10%,...jeez stop already. Your like school children acting as if Nvidia stole your lunch money.Everyone cheats as you so put it but how in god names do you go fourth and say its a corporate mandate? I'll tell you why. Intel and Amd have there feet so up your *stars* that you seem to just keep ranting and ranting on and on of how the green team besides there superior performance cards are the bad boys in all this because they can innovate and change the rules to gain better benchmarks. Well well i'm impressed Nvidia. Cry me a river Intel/Amd and inquire.
Rules are made to be broken. Man broke the sound barrier so if Nvdia can emulate a CPU with up to 10x the performance with PhysX API so be it. Give us the power Nvidia! yeah! This is why intel hate them..they know there CPU's are being replaced slowly in the gaming arena...
So what Nvidia has extra horse power to replace other test and let the GPU do it better...omg stop will you. Nvidia give consumers the power to play games at the best possible way.
Did you know on load a 65nm card gtx280 uses less power and is less hot than a 3870X2 55nm. Pawned again Noobs.
' check this out:
http://techreport.com/articles.x/14934/16
'omg stop nvidia stop. Humiliating people at this level. No wonder there is so much hate. Go on say ..uum wait for 4000 series. lol. The dudes haven't even done a gtx280 on 55nm yet. ATi know pretty well its going to be hard, that's why the inquire has started a propaganda campaign against the greens. Sorry but the benchmarks speak for themselves. Keep whining.
Technological edge is the key to success. Change is necessary to innovate even if it means changing the rules of the game if it increases performance. Swallow this real hard and don't forget it..
Physix on GPU - not good.
I think the problem is somewhere deeper, and it is not just cheating with different workload. As everybody knows 3DMark has always been used as a tool: a) to get a highest score on planet, b) to understand which way of improving visual (mostly) technologies the gaming industry chose for some 2-3 years. Using this knowledge se can estimate, what components of our system can be upgraded before others, and so on. So, nVidia with the current (unnoficial) set of drivers tries to make us beleive that it will help you get the highest score for you money and that gpu is not just all the visual candies you knew many years before. One of the CPU-subtests, that uses the PhysX workload is totally simple in visual way. And, as we know (because nVidia provided support for PhysX even for previous graphic card generation) there is no any special chip, that makes accelerating physics. All the workload is done via same old good shader processors. We have 99% of GPU busy with physics and 1% busy with actually rendering those simple planes and rings, barely textured. And you get ultra-boost for physics processing (~ 3 times faster then your daily dual core CPU). And of course - this score highly influences the final 3dMark Vantage score - be it Performance or Extreme profile. But if you remember, 3DMark serves to predict some average game stuff rendering. And that is why this test is going oh so wrong. If you take modern game (with modern visual engine), say Crysis - it will overload GPU with render job. So what is going to be about physics/render? 50/50? 25%/75%? Maybe 10%/90%? It`s definetely not going to be 99/1 like in that Physix test in 3DMark Vantage. Something gonna go slowly. And fps gonna drop fo sure. There is no free cheese fo you and me. Number of shader processors in GPU is strictly fixed, it doesn`t stretch when you have another one task to do - currently physics modelling. And that boost is final score of 3DMark Vantage is total fake, I mean that test doesn't make any sense, if done by GPU. Pyscics should be modelled with what is done for it. For example Ageia physics card, or some free cores of your regular CPU. Cheers.Charlie Demerjian....
An anti-nvidia content from an ATI fanboy idiot..... content dismissed.Pseudo-ethical rant
Who cares about your pseudo-ethical rant over benching?The important article should be, does it actually make games work better?
nope silly
"So, with the latest driver, Forceware 177.39 drivers, Nvidia put its now in-house PhysX APIs into the drivers. Instead of it running on the CPU or on the PhysX chip, it is running it on the GPU. It owns the GPU and it's drivers along with the physics API and all those drivers. This is a dangerous situation."it might be bad for some stupid synthetic benchmarks, but this particular driver enabled nvidia cards to do physics, which is what it was promised to do. offloading physics from the cpu meant it work. you just have to pray Mr. ATI Fan boy that havok does well on the 4800s. its 3dmarks fault they didnt anticipate gpus doing physics anyway and i guess you guys will be seeing a hotfix in the upcoming weeks. lets see what happens..
all in all i dont care about synthetics, if enabling physx will increase performance from my games why not?
PPU?
So how exactly is this any different from PhysX being run on a PPU?Then given Nvidia's instructions for using it with UT, and of course how to see the biggest affect in 3DMark it sounds like they're just showing off their Cuda-PhysX implementation. So hardly anything sinister or dastardly. I even imagine that the drivers will eventually leave beta and get certified by Futuremark, but until then they have to start somewhere.
But really Charlie, how much money is Intel throwing your way? Though you haven't written anything about Larrabee lately. So maybe some Nvidia hack called your mom something nasty. Either way you're credibly as a journalist has been sinking lower and lower as a result.
not cheating imo
I disagree this is cheating. It looks like a flaw on 3DMark's part. 3DMark uses a generic API, which was *meant* to be hardware-accelerated, and calls it a CPU test. They allow arbitrary hardware to run this test (such as PhysX card) with an unreasonable exception of GPU, even though some kind of integration of PhysX and GPU (such as just putting them on one card) was always an option.The "running the same code" theory was always BS because any PhysX-capable hardware will require own separate drivers ("code") which is different from CPU emulation. Simply using PhysX card runs different "code" as well (even if it's distributed in same package).
Uhm
I guess nVidia should kill off the entire idea of using some of the streaming processors to do physics, right? Seriously, this isn't really a problem with nVidia's drivers, it's a problem with how 3DMark Vantage crunches numbers. They included the PhysX code in the first place, and there are no surprises that G80's and upwards are far better at crunching physics in multiple streams at very high frequencies than the sloppy old ageia chip.Anyone with a PhysX card would get better scores than someone without one even if they had the same system (physx card excluded of course), and that's just as unfair. Still, futuremark called the shots on this one, nVidia is doing something good by finally putting down the "GPU as a high performance number cruncher". Whether the numbers are crunched on a separate card or via a select number of the SP's on nVidia cards is completely irrelevant.
he has a point guys!
the 3d mark physics tests are there to test how wel your other system components can handle physics while your GPU is doing graphics.but the GPU isnt needed for graphics while running this test!
that means 100% of the GPU can be used to run the physics test with the nvidia drivers.
which is a situation that will obviously NEVER EVER happen while playing games. and that means doing it this way yields totally unrealistic resultes.
call him a ati fanboy all you want but he does have a valid point.
baaaawwwww
cry moar, ati lover8.6 'official' has been out...
http://game.amd.com/us-en/drivers_catalyst.aspx?p=xp/radeonx-xpThey been out since before the 4850 got released? How much more official do you need?
Physics are changing
It really doesn't make sense to apply physics tests to CPUs when it makes far more sense to apply the physics tests to a GPU. Physics needs to be taken away from CPUs and that burden to be placed onto a GPU. Does it make sense to physics test a CPU with one, two, or four cores when you can place such difficulty upon a GPU which has far more cores?I'm No Apothecary
Curses! Voilated again!3DMARK = POTENTIAL OF SYSTEM
3DMark is only showing potential power of your system. If people are using flux capacitors, then GO ON RIGHT AHEAD. Pfff who cares if your system only can use 500W. IT'S NOT CHEATING IF YOUR COMPUTER CAN USE 1.21 JIGGAWATTS FOR 300 CARDS IN SLI!Biased anybody??
Hope Charlie sees where the wind blows from all above comments. Stop bashing the green, remember...reporters are supposed to be objective...? Nevermind that clearly does not apply to you. This is just one of a string of articles having you whine and trash about Nvidia. Sure the GTX280 is a different approach as compared to ATI, after all it is a big chip. So whats the big deal. Pls. share with us what did NVidia do to you, did they not give you a free sample since you are the ...reporter and they are not respecting you as they damn well should? (what a surprise). Nvidia has the right to use PhysX anyway they want to and if by any chance it improves performance for games using PhysX, bring it on! Its about time, last year everybody was talking about GPU physics and nothing happened. At least Nvidia is bringing the real application to the table and for that I really have to salute them.It isn't even running the same program?
I don't understand what Nvidia's in-house PhysX DLLs are doing that is bad and I still don't understand how they aren't running the same program after re-reading the article a couple times.If the same amount of number crunching is required but Nvidia just sends it to their GPUs then that isn't really cheating...
I find it debatable and maybe it is technically cheating by the rules of 3DMark. Although, Nvidia definitely should have it certified so no one would question the validity of the 3DMark results.
Is too early to say N is cheating,but Charlie found the problem
I don't think Charlie got no points and just Afans.The whole problem is ,this is whose responsibility to make the score right(according to Futuremark's rule) here? The drivers? or the 3Dmark. New drivers always rip off bug in order to run software correct . And software sometimes will correct itself.
So, I found it's difficult to say it's a cheat or not.
You retards
Stop bashing Charlie over this article, we all know hes ati biased, and well so am I (tho i still have my 8800gtx as well as my 3870x2). But there is more too it than that, use your brains for once. The CPU test is designed to stress CPU ONLY, the GPU test is designed to stress the GPU with minimal stress to the CPU, but in this case with these drivers nvidia is using the gpu for both, and everyone goes "Wow look at those scores!". Think about it, in REAL GAMES BOTH ARE STRESSED! So this fanciful boost only results in a LOWER fps in actual games because the GPU is too busy worrying about physics instead of pushing frames. It really is a blatant exploit. Stop the mindless flaming and actually think about it, or does Charlie have to spoon feed it to you?280GTX SLI scores
So i've been running some benchmarks on 3DMark Vantage. My first score was around 15500K with version 177.35. This is on stock Q9450 2.66ghz and 2x280gtx. Now I installed the PhysX drivers 8.06.12 ran it again on stock and got 19800k roughly. OC'ed my CPU to 3.2ghz and hit 20800k roughly. So am I cheating since I didn't install 177.39 ?Charlie sucks AMD C@:k
Well we all no charlie loves AMD , and will bash any competitor of his beloved AMD.One very bitter man.
This is why synthetic benchmarks break down
You just have to test with real games. This is like a pickup truck that can tow 10,000 pounds and can go 0 to 60 in 8 seconds. Both of those may be true, but you can't sum those scores and assume you can tow 10,000 lbs from 0 to 60 in 8 seconds. Since this test DOES sum the scores, we just have to discount the scores as unrealistic results.I'd say the fail here belongs to 3DMark, not nVidia, though. While the nNidia drivers produce bad results in 3DMark, but I doubt they were DESIGNED to produce bad results, they were designed to improve gameplay. Its the unrealistic one-load-at-a-time nature of 3dMark that causes problems, and the tests' assumption that physics processing doesn't take place on the GPU.
ha ha
Readers comments overwhelmingly show what they think of Charlie's 'articles.' Take a clue Charlie, your readers see you as a whiney little girl. The only thing your submissions are good for is some mild entertainment and getting the INQ's page hit counters up. What a sorry excuse for a journalist.The boost works
I hacked the vista version 177.39 drivers to include my card, the 8800gts (G92) and I noticed a 500 point increase in 3dmark06 without the new physics driver (which I think is only for vista?). Then after installing the physics driver the boost is noticeably substantial. (I won't boast a 3dmark score here) and uh, it submits to futuremark without any problems.nVidia cheat? Who does his surprise?
Every launch generation there needs to be some cheat issue. Whether it's failed optimizations as in the Crysis cheat, or the use of a GPU for the PPU test against Futuremark's own license agreement.Since the PPU can't do graphics work but the GPU is basically unburdened in the CPU test, you now have a result that's not even comparable to the previous CPU+GPU+PPU results, and would be no indication of the overall performance either.
Once again 3Dmark proving its worth or lack thereof and once again NV proving they'll cheat to gain point in a worthless benchmark despite their statements to the contrary.
Now intel just needs to write a driver that skips the test and creates it's own score, and then we'll have equality.
Guess Futuremarks terms of use are about as pointless as the benchmark itself.
Now if Nvidia could only run some tests with partial precision or on rails then I could truly feel like I'm back in 2003. It's kind of appropriate that you just round off the 3 and you could interchange the articles and defences for Nvidia cheating in 2003 as if they were written in 2008.
All we need now is the threat of legal action again, and then the circle would be complete.
3D Mark
3D Mark, the gamers penis replacement.Legitimate accleration
It this seems like a legitimate acceleration to me- it demonstrates how nVidia GPUs can accelerate PhysX
- the PhysX test was never intended to be a 'CPU-only' test as it's accelerated by the old PhysX card, so it is truely a PhysX test not a CPU test.
- the only possible problem is that it tests PhysX (nearly) in isolation, and as others have pointed out, is not stressing the graphics at the same time. But it's still pretty obvious what's going on.
- UT3 is also accelerated by this, so it works well in a real game
- so I'm not sure what the problem is here
You retards->And you retard too!!
You say:"The CPU test is designed to stress CPU ONLY, the GPU test is designed to stress the GPU with minimal stress to the CPU, but in this case with these drivers nvidia is using the gpu for both, and everyone goes "Wow look at those scores!". Think about it, in REAL GAMES BOTH ARE STRESSED!"
Oh man, what do you say? This f..king shit?:
"
"The CPU test is designed to stress CPU ONLY"
All right, sure. A machine with a Ageia Card, for example. You don´t say this stupid thing.
Futuremark had created a fu..ing CPU test that doesn´t be at all. Their problem, itsn´t the problem of Nvidia.
Another stupid thing from you:
""Wow look at those scores!". Think about it, in REAL GAMES BOTH ARE STRESSED!""
Yeah, sure too. The actual games have very much process power in physics than in graphics, "sure".
In REAL GAMES the physic represents a little fraction of the power cpu resources. So, what´s the problem with GPU physic?
Nothing, only that this "thing" can slowdown the fps in a game, BUT with better physics that the games have right now.
And with simple physic of current games a better physics (similar to Ageia card) without backdraws. What a problem!!
PD: Charlie, hava a little respect for yourself, and stop writing crap like this.
er_wendigo - the real true retard
You are a blinkered nvidia fanboi fool, let me guess, an average joe sitting at his home playing games and talking smack.For the intellectually trouble goons like yourself:
A CPU test is strictly for the CPU only, there is no talk of "oh its only an ageia card in there"
So if Nvidia is meddling with that through their new feature it is a clear deception.
All you have in your reply is swears and nothing of importance, just shows how some clowns come in here and try to defend Nvidia for everything that is being posted in here critical of their activities.
Sounds like you are a cry baby teen, one of those idiots who support sports teams and would fight over it as if it meant more than your average life.
Get life, grow up.
BS
This is Futuremarks fault for using hardware accelerated PhysX (when it is available) in its "CPU test". When an Nvidia driver offers PhysX hardware acceleration, then you get higher score there _as you should_ because that's how the benchmark is designed to run - To use PhysX. And you get the same result in games that use PhysX too.what an "impartial" newscoverage
I'm a happy radeon 3850 owner, but I'm noones fanboy. I have to say that this article is nothing but objective. It's full of hate and bitterness towards Nvidia, for one reason or the other. Why is that so.. I don't know.No big coorporation is an angel, and there is no side picking between ADM/Nvidia/Intel etc...
The bottom line is this. Nvidia has a software solution that will improve physix performance in games thru existing hardware for no additional time or money investment by the end user. And what in the world is wrong with that!?
Frankly speaking, we should all be glad, someone found a way how to get more out of hardware. They should be cheered, not spit at. Futuremark scores or any benchmark is irrelevant.
Give Charlie a break
OK, we all know Charlie has something of a pet hate for Nvidia but don't instantly write off everything he has to say. This is a good article with a very valid point.Must add
When you tell a thin lady she is fat, she and her mates get upset.When you tell a fat lady she is fat, everyone gets upset.
Just did a vantage run
I installed the latest 177.41's together with the nvidia physx 8.06.12's and ran the vantage benchmark.While the rest of the parts didn't change, the Physics test jumped way past the score of the highest scoring vantage test. AND that supersystem have an Ageia PhysX 100 card as well (which is just as much cheating as using the geforce if we use charlie's logic since neither run the test on the cpu). The fastest system in vantage @ ORB have a physics score around ~41 ops/s while I got ~118 ops/s, more than double.
Is this cheating? No. Instead of sending the data to a dedicated physx-chip it sent them to the gpu, both non-CPU's. The flawed logic of charlies' is that nvidia is cheating because they allow the GPU to run the calculations instead of a dedicated physx-chip or the CPU. The only way it would be cheating is if the test had been named "CPU test" instead of "Physic test".
time to eat crow
While it's listed as physics @ the futuremark ORB, it does say "CPU test" in the vantage result window after you've submitted the result, something I did not notice. Still, futuremark decided to allow hardware accelerated physics in the first place by using physx (ageia).FUTUREMARK CONFIRMS THIS IS NOT CHEATING...
This article below from PC Perspective has an interview with ATI/AMD/Epic/FUTUREMARK/Nvidia and all of them stated for a fact that this what not cheating. the PhysX test is exactly what it is and if a company brings new inovation to push the boundaries of PPU then it's not cheating. Charlie should read this article cause unfortunatly ATI/AMD camp is not on his side :)http://www.pcper.com/article.php?aid=584
Cheating ??
Maybe some people see this as cheating. But if that way the gamer has better performace, why call it wrong? Do you really think that the only reason of inventing things is to get a better score in 3DMark? The way I see it, they only cannot pretend to have a real and genuine 3DMark score, because it violates with their politics and rules. But cheating it ain't. Cheating would be if it shows me a greater 3DMark score but in real performance goes lower than drivers before let's say. Then it would be faking the score, so cheating. Since the preformance the way they do it is obvious greater than before, it is just new way to make things, which gives better results. The only one who can complain is 3DMark if the score is given as genuine. But who can be so "smart" to think that the reason of Nvidia to walk that way is just to fake 3DMark numbers? Come on, wake up! What matters is the result in gaming, not if someone likes the numbers of a benchmark program....