Futuremark removes GTX 280 scores

[FUGGER]

But I am told they will be back up shortly.

All non WHQL drivers will not be publishable (a given)

177.39 is a beta driver and will not be valid for ORB (a given)

Nvidia 177.35 drivers are being considered for ORB approval.

Catalyst 8.6 drivers are being considered for ORB approval.

Physx 8.06.12 driver is a system driver and does not effect graphic display or performance on GPU tests.

Multipath IO driver (http://en.wikipedia.org/wiki/Multipath_I/O) is a system driver, does not effect graphic display but does effect graphics performance.

Vantage has a "Physics" test, no one expected it work so well.

Futuremark is fixing the non whql drivers and the non FM approved drivers currently and that is fine and the way it should be with exception to 3Dmark 2001 where anything goes. :clap:

The removal of 01 scores was a mistake to the global change in driver support on the ORB and it will be fixed.

Physics is a win for consumers, ATI is freaking out with no answer as they should be, they got caught with their pants down.

In time if ATI does get it together and delivers Physics on their GPGPU it may do better with 800sp and higher clock speeds but as it sits now, ATI wants nothing more than to snuff physics in Vantage.

Futuremark, dont make us cripple our systems to run your benchmark. I would not disable Physx 8.06.12 nor would I disable Multipath IO for gaming.

Separate "physics test" from "CPU test" and avoid the semantics being used against you. We all know and accept it is a physics test.

[TheGoat Eater]

That blows - your hard work removed :down:

I have vantage but haven't run it yet and this is not not making me want to run it at all!
^^^This and 9600 GSOs not being implemented in ORB sucks for me also (so I can feel where you are coming from on a fraction)

[FUGGER]

Its ok, scores on 177.35 will be back up. I ran my stuff with WHQL approved drivers knowing the problems with non fm approved.

My 119K score will be back up too, that was a mistake.

Vince took a hit but he will recover, we undertand WHQL and FM support, we know what should be visible and accept that.

[Baron]

Any idea of a time frame when they say "being considered" ?

[TheGoat Eater]

Its ok, scores on 177.35 will be back up. I ran my stuff with WHQL approved drivers knowing the problems with non fm approved.

My 119K score will be back up too, that was a mistake.

Vince took a hit but he will recover, we undertand WHQL and FM support, we know what should be visible and accept that.

yeah - good to hear that you got your results back:up:! I know this is most likely gonna piss people off though - like those who spent a lot of time working and spent $$$ on LN2,DICE, what have you only to get results removed... with about nothing to show for it in the long run... well those that used the betas :shrug:

you agree as this is big (to me) as far as time / money

I know i am just cutting my teeth in this arena (extreme overclocking) but to me it seems awful to loose monetary and time value put towards a goal... now if I could get 9600GSOs up and recognizd by ORB (FM) for Monday for a certain reason that would be good for me (one can hope right ;) :shrug: ) and my OC lifestyle well being :D:rolleyes::p:

[STaRGaZeR]

Vantage has a "Physics" test

http://img258.imageshack.us/img258/8378/86724835mu7.jpg

FM should have warned about this, but as you see, it says CPU TEST. And CPU means no Ageia PhysX card and no PhysX on GPU, either ATI or NVIDIA. If they want a Physics test, put it alone. Both "cheats" should be disabled.

Also I don't know if you were the one who deleted the two threads in the news section about the TheINQ article regarding this, but that's fanboyism in its pure form.

[Warboy]

http://img258.imageshack.us/img258/8378/86724835mu7.jpg

FM should have warned about this, but as you see, it says CPU TEST. And CPU means no Ageia PhysX card and no PhysX on GPU, either ATI or NVIDIA. If they want a Physics test, put it alone. Both "cheats" should be disabled.

Also I don't know if you were the one who deleted the two threads in the news section about the TheINQ article regarding this, but that's fanboyism in its pure form.

I was wondering myself about the 2 threads, but nevertheless, The PPU was allowed since launch, Futuremark clearly knew it was going to be used. PPU has been out for how many years? wasn't it out for the development of Vantage?

Everybody heard that Nvidia was going to take PhysX and make it run on 8800s or higher GPUs. Your Acting like this is Nvidia fault, Nvidia just kept their word by developing it. It wasn't supposed to be leaked to the public either. That's why it doesn't support all the GPUs it should, Futuremark possibly knew about the work nvidia was doing. Because if you think about it, Neither parties are at fault here. Neither parties really expected a leak.

[STaRGaZeR]

I'm not against PPU or NV PhysX, never! I'm against using them in a CPU test. That was Futuremark's error since the very beginning, and now they have this rough decision to make.

[FUGGER]

To be fully aware, Physics was made to run on Ageia card, the driver for CPU was for development so that shoots the CPU theory down.

Crash and Burn: Physics

The fact is that it is a physics test.

[STaRGaZeR]

You can say what you want, in the final version you need two GPU test and two CPU test to get a score. A CPU is a CPU.

[Warboy]

You can say what you want, in the final version you need two GPU test and two CPU test to get a score. A CPU is a CPU.

Well, Fugger is right. From what I've seen and been told. They pretty much adopted PhysX into the benchmark, So CPU Test 2 is Clearly Physics based.

Clearly written on Futuremarks site about Vantage is this.


CPU Test 2: Physics The Physics Test features a heavy workload of future generation game physics computations. The scene is set at an air race, but with an unfortunately dangerous configuration of gates. Planes trailing smoke collide with various cloth and soft-body obstacles, each other, and the ground. The smoke spreads, and reacts to the planes passing through it. The physics test takes advantage of the AGEIA PhysX physics accelerator, if found on the system.

[STaRGaZeR]

As you know, PhysX effets can be calculated by the CPU just fine, but at much lower speed. 3DMark06 also uses the PhysX libraries and nobody told anything back then. A PPU is not a CPU and it should not be allowed in the test.

But this thing is getting nowhere. They should have put it in a different test since the beginning, now it can't be helped. Ageia cards will still be supported, but not NV cards.

I doubt you'll ever see PhysX on GPU (ATI/NV) allowed in Vantage.

[demonkevy666]

As you know, PhysX effets can be calculated by the CPU just fine, but at much lower speed. 3DMark06 also uses the PhysX libraries and nobody told anything back then. A PPU is not a CPU and it should not be allowed in the test.

But this thing is getting nowhere. They should have put it in a different test since the beginning, now it can't be helped. Ageia cards will still be supported, but not NV cards.

I doubt you'll ever see PhysX on GPU (ATI/NV) allowed in Vantage.

I agree with that.

wouldn't you need a driver for that to work in the first place to make the card look like a PPU instead of a GPU.

I wonder if that will make hem have to remake the whole thing to get ride of the CPU doing PPU stuff.

[bowman]

The second 'CPU' test has always been allowed to run a physics card, and this was widely touted before release.

This is Futuremark's fault for calling it a 'CPU' test because it wasn't one from the start. Only if you didn't have a physics processor. Then, noone had an Ageia PPU, but now, a lot of people have a GeForce PPU (in effect, or PPonGPU if you want to be pedantic :p:)

[Warboy]

As you know, PhysX effets can be calculated by the CPU just fine, but at much lower speed. 3DMark06 also uses the PhysX libraries and nobody told anything back then. A PPU is not a CPU and it should not be allowed in the test.

But this thing is getting nowhere. They should have put it in a different test since the beginning, now it can't be helped. Ageia cards will still be supported, but not NV cards.

I doubt you'll ever see PhysX on GPU (ATI/NV) allowed in Vantage.

Well I guess Nvidia and Futuremark have to work out something. Because I don't think Nvidia will sit back and say "Oh noes, Don't show off our PhysX on GPUs" The best thing really is rename CPU Test 2 to Physics Test 1, And Have it effect final score less then what it does now.

[FUGGER]

Warboy;3088574']Well I guess Nvidia and Futuremark have to work out something. Because I don't think Nvidia will sit back and say "Oh noes, Don't show off our PhysX on GPUs" The best thing really is rename CPU Test 2 to Physics Test 1, And Have it effect final score less then what it does now.

That is the best option, scoring may change some but physics is still a major test that gamers need to know what is stronger. Its sad that ATI is so far behind it makes this a one sided race, but hey its not our fault they had plenty of time to deliver.

ATI had the monopoly on Crossfire with Intel as AMD cpu's suck. Sure lots of people complained but we survived. Now that we have SLI on Intel its a whole new ball game and the winner is...Nvidia.

[Warboy]

That is the best option, scoring may change some but physics is still a major test that gamers need to know what is stronger. Its sad that ATI is so far behind it makes this a one sided race, but hey its not our fault they had plenty of time to deliver.

ATI had the monopoly on Crossfire with Intel as AMD cpu's suck. Sure lots of people complained but we survived. Now that we have SLI on Intel its a whole new ball game and the winner is...Nvidia.

I hope Nvidia gets the license back to make Chipsets for Nehalem. Can't Wait to play with Nehalem+SLi :D

[Enoc]

Warboy;3088574']Well I guess Nvidia and Futuremark have to work out something. Because I don't think Nvidia will sit back and say "Oh noes, Don't show off our PhysX on GPUs" The best thing really is rename CPU Test 2 to Physics Test 1, And Have it effect final score less then what it does now.

lke Fugger said, that's a reasonable option...maybe the best part is what you say about the score...should effect less...

if FM does not correct that part it could either backfire on them or Nvidia...when it comes to reviewers...this is a example...X nvidia card gets 15k points , X Ati card gets 7k points, half the points...when the hit up the games benchmarks..frame wise..they are even...or let say..ati wins...we all know that FM benchmark is a indicative of a card performance...the higher the better...
so the reviewers will end up with 2 conclusions...

1.Nvidia car is a good benchmark card..but when it comes to delivering frames in a game..is bad...
2. FM vantage is not a good indicative of performance anymore cause you got some video cards that score way less(4 -6 k points) and have give better frames in games...so is going to discredit vantage...

and in the end...the one who is going to look good Fugger is Ati...the PR is going to be " more performance with half of the points in FM benchmarks"

look at the bigger picture...;)

[demonkevy666]

lke Fugger said, that's a reasonable option...maybe the best part is what you say about the score...should effect less...

if FM does not correct that part it could either backfire on them or Nvidia...when it comes to reviewers...this is a example...X nvidia card gets 15k points , X Ati card gets 7k points, half the points...when the hit up the games benchmarks..frame wise..they are even...or let say..ati wins...we all know that FM benchmark is a indicative of a card performance...the higher the better...
so the reviewers will end up with 2 conclusions...

1.Nvidia car is a good benchmark card..but when it comes to delivering frames in a game..is bad...
2. FM vantage is not a good indicative of performance anymore cause you got some video cards that score way less(4 -6 k points) and have give better frames in games...so is going to discredit vantage...

and in the end...the one who is going to look good Fugger is Ati...the PR is going to be " more performance with half of the points in FM benchmarks"

look at the bigger picture...;)

that sounds like the 2900 XT descriptions lol

benchamrks where high but FPS in games wheren't as great.

[Warboy]

lke Fugger said, that's a reasonable option...maybe the best part is what you say about the score...should effect less...

if FM does not correct that part it could either backfire on them or Nvidia...when it comes to reviewers...this is a example...X nvidia card gets 15k points , X Ati card gets 7k points, half the points...when the hit up the games benchmarks..frame wise..they are even...or let say..ati wins...we all know that FM benchmark is a indicative of a card performance...the higher the better...
so the reviewers will end up with 2 conclusions...

1.Nvidia car is a good benchmark card..but when it comes to delivering frames in a game..is bad...
2. FM vantage is not a good indicative of performance anymore cause you got some video cards that score way less(4 -6 k points) and have give better frames in games...so is going to discredit vantage...

and in the end...the one who is going to look good Fugger is Ati...the PR is going to be " more performance with half of the points in FM benchmarks"

look at the bigger picture...;)

I do look at the biggest possible picture.

[jimmyz]

If they are going to change the benchmark (highly doubtfull) then they should do something about the other issues like load times.


And while they are at it they can rewrite the graphics test to use DX10.1. after all it is std now with service pack 1.

[FUGGER]

Good points Jimmy.

[STaRGaZeR]

Warboy;3088574']Well I guess Nvidia and Futuremark have to work out something. Because I don't think Nvidia will sit back and say "Oh noes, Don't show off our PhysX on GPUs" The best thing really is rename CPU Test 2 to Physics Test 1, And Have it effect final score less then what it does now.

That's not possible and you know it. Then Vantage would be a completely different benchmark. The same goes to DX10.1 . That can be done for Vantage 2 or whatever. The only thing they can do now is, without changing anyhing in the current scores and tests, add another preset with GPU PhysX or DX10.1 allowed. Or do nothing and just block any submissions done with this driver.

But I want NV there doing pressure, that will force ATI to do something. Now that we are in the beginning of the GPGPU era so to speak they should define an open source language that can be used by everyone. CUDA and CTM are not a viable option IMO.

[Warboy]

That's not possible and you know it. Then Vantage would be a completely different benchmark. The same goes to DX10.1 . That can be done for Vantage 2 or whatever. The only thing they can do now is, without changing anyhing in the current scores and tests, add another preset with GPU PhysX or DX10.1 allowed. Or do nothing and just block any submissions done with this driver.

But I want NV there doing pressure, that will force ATI to do something. Now that we are in the beginning of the GPGPU era so to speak they should define an open source language that can be used by everyone. CUDA and CTM are not a viable option IMO.

It is very possible, all they have to do is change a simple bitmap and how it calculates scores and boom, Hotfix 2 for Vantage.

[FUGGER]

It is possible and the best route.

Will I consider a game now that support Physx? yes.


AMD and Havok announce plans to optimize physics processing on the AMD platform

SUNNYVALE and SAN FRANSISCO, Calif. — June 11, 2008 — AMD (NYSE: AMD) and Havok today announced plans to jointly investigate the optimization of physics effects utilizing AMD’s full line of products.


Havok Physics scales extremely well across the entirely family of AMD processors, including quad-core products such as the AMD Phenom™ X4. As part of the collaboration, Havok and AMD plan to further optimize the full range of Havok technologies on AMD x86 superscalar processors. The two companies will also investigate the use of AMD’s massively parallel GPUs to manage appropriate aspects of physical world simulation in the future.

Reality is that I have had more time to investigate Superpi on CUDA...

http://www.havok.com/content/view/639/53/

If and when this becomes reality then hopefully FM will have an appropriate benchmark ready, cant blame FM when ATI hasnt done anything to date.

[AbelJemka]

ATI had the monopoly on Crossfire with Intel as AMD cpu's suck. Sure lots of people complained but we survived. Now that we have SLI on Intel its a whole new ball game and the winner is...Nvidia.
Very weird to see you write that. I don't understand what this sentences mean :shrug:

[FUGGER]

We wanted SLI on Intel chipsets while Crossfire was the only multi gpu option for Intel. Eventually we got 680i

It was out of our control no matter how much we complained and bugged Nvidia/Intel.

[Movieman]

We wanted SLI on Intel chipsets while Crossfire was the only multi gpu option for Intel. Eventually we got 680i

It was out of our control no matter how much we complained and bugged Nvidia/Intel.

Personally I'd love to see Intel dump some hacked drivers on the net and get this whole issue settled.
it is just wrong not to be able to run 2 nVidia cards on an Intel chipset because two companies are having hissy fits at one another over what is really a small issue overall.

[BenchZowner]

Don't get me wrong... but as somebody said before in the previous page ( I think it was Fugger ) AMD/ATi failed to deliver.
ATi and then "AMD" was promising us and talking about GPU Physics since HD2900XT if I recall correctly ( actually even before the release of the GPU )... nVIDIA was promising the same thing at the same time.
Difference is... nVIDIA delivered, "ATi" is still "trying" or wasn't even working on it "for real" as they didn't expect nVIDIA to finally come out with GPU Physics ( both companies were talking about GPU physics for years and none of them did showcase something "valuable" or "real" in the past, so they didn't expect something like that to finally face the surface and eventually the market )

Edit: @ Dave ( MovieMan ) that's true.
Damn, thinking about P5K3 Deluxe, Gigabyte GA-X48T-DQ6 and Asus P5E64 WS Evolution with SLI & CrossFire capabilities together...this summer's wet dream :p: :D

[Movieman]

Don't get me wrong... but as somebody said before in the previous page ( I think it was Fugger ) AMD/ATi failed to deliver.
ATi and then "AMD" was promising us and talking about GPU Physics since HD2900XT if I recall correctly ( actually even before the release of the GPU )... nVIDIA was promising the same thing at the same time.
Difference is... nVIDIA delivered, "ATi" is still "trying" or wasn't even working on it "for real" as they didn't expect nVIDIA to finally come out with GPU Physics ( both companies were talking about GPU physics for years and none of them did showcase something "valuable" or "real" in the past, so they didn't expect something like that to finally face the surface and eventually the market )

Edit: @ Dave ( MovieMan ) that's true.
Damn, thinking about P5K3 Deluxe, Gigabyte GA-X48T-DQ6 and Asus P5E64 WS Evolution with SLI & CrossFire capabilities together...this summer's wet dream :p: :D

I'm sitting here with a board with Intel chipset, 2-rev 2.0 PCI-e X16 slots that I can get a full 16 signal to each and yet I can't do a SLI with nVidia cards due to this foolishness between them.
Now gaming isn't my thing but I'd like to see what the system would score on HWBot and I can't under the current situation.

[Gautam]

Personally I'd love to see Intel dump some hacked drivers on the net and get this whole issue settled.
it is just wrong not to be able to run 2 nVidia cards on an Intel chipset because two companies are having hissy fits at one another over what is really a small issue overall.

It's not that small of an issue. If we could have SLi on Intel chipsets, there would be virtually no need to buy nVidia chipsets. nVidia knows this well...foolishness or not it's a market they want to protect.

As much as I hate it, it's their product and they can do what they want with it.

[Movieman]

It's not that small of an issue. If we could have SLi on Intel chipsets, there would be virtually no need to buy nVidia chipsets. nVidia knows this well...foolishness or not it's a market they want to protect.

As much as I hate it, it's their product and they can do what they want with it.

I understand the issue and their wanting to protect their "turf" but it can swing both ways. If I was making decisions at Intel I'd refuse them any ability to use anything until this is settled.
If nVidia wants to protect their turf, make a better board than Intel and let the marketplace choose. That's capitalism, not what is happenning now my friend.

[jimmyz]

The real issue with these drivers and Vantage is layed out in plain language in the vantage driver rules. no. 5 states:


5. 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.

I am not posting this to argue against ATI's lack of physics on the gpu just to clarify the rule that Nvidia violated.

I am glad to see something finally being done with physics on gpu's, it has been way to long, with too many announcements and chest beatings with no concrete results until now. (BTW I'm 40yrs old so if your gonna call me names make it FANMAN, I haven't been a boy for a couple of decades...lol)

[Gautam]

If nVidia wants to protect their turf, make a better board than Intel and let the marketplace choose. That's capitalism, not what is happenning now my friend.

Amen to that, but we can't hold our breath for it. :p:

[Movieman]

Amen to that, but we can't hold our breath for it. :p:

That i will agree with.. Like two kids fighting in a schoolyard over who gets the ball. Both companies are losing by this.
I may just go out and buy an ATI card for spite to deny nVidia the money even though they have the superior card.

[Warboy]

The real issue with these drivers and Vantage is layed out in plain language in the vantage driver rules. no. 5 states:


5. 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.

I am not posting this to argue against ATI's lack of physics on the gpu just to clarify the rule that Nvidia violated.

I am glad to see something finally being done with physics on gpu's, it has been way to long, with too many announcements and chest beatings with no concrete results until now. (BTW I'm 40yrs old so if your gonna call me names make it FANMAN, I haven't been a boy for a couple of decades...lol)

But it says nothing about Physic Accelerators. And Technically the driver doesn't supply it. Any 175+ Driver works just fine. The Physics Software is what enabling it. It never said Software, It says drivers. So I still don't really get why people are referring to that.

[Sampsa]

Just got a reply from Futuremark, here is a rough translation from Finnish to English: "In short, we will investigate every official WHQL driver and accept only those which will fully meet our requirements and give comparable test result."

[hecktic]

Just got a reply from Futuremark, here is a rough translation from Finnish to English: "In short, we will investigate every official WHQL driver and accept only those which will fully meet our requirements and give comparable test result."

lol they cant be serious....

I think the easy way out of this is for some smaller brand that develops and maintains gaming related benchmarks emerge and offer true results with cpu, gpu, and ppu.

If that happend, then I think FM would get their hands out of their rear and give us all a quick fix satisfying everyone (unfortunately that might leave ATI in the dust due to lack of PPU + GPU integration but oh well...)

EDIT: oh and I see no one mentioned this... this crazy ban on 280 scores is going to have a dire effect on 280 sales lol...

[fubarswe]

.., simply cant belive this.., but it feels safe to be in the hands of pros.., or..?

[ceevee]

To be fully aware, Physics was made to run on Ageia card, the driver for CPU was for development so that shoots the CPU theory down.

Crash and Burn: Physics

The fact is that it is a physics test.


Agreed lets hope AMD doesn't ruin this for everyone.

[AbelJemka]

Agreed lets hope AMD doesn't ruin this for everyone.
Yes AMd is evil and dictate Futuremark theirs drivers rules :confused: