Overclock.net › Forums › Industry News › Software News › [INQ] Nvidia cheats on 3DMark with 177.39 drivers
New Posts  All Forums:Forum Nav:

[INQ] Nvidia cheats on 3DMark with 177.39 drivers - Page 5

post #41 of 103
The ATI/Nvidia war was cute at first, but now it kinda making me sick. Reviews on websites are slinging their weight to one card (4850, 9800gtx) or the other. Now after reading this I really do not trust 3dmark scores either.

Just gonna wait on the OCN faithfull to buy these dang cards and get some realtime gaming expertease on some of our fav games and then go with it.
post #42 of 103
So now beta drivers = cheating? Or is this just something else to start the ATI side whining again? Ati people your hotfixes = cheating! lol I'm just joking
24/7
(13 items)
 
  
CPUMotherboardGraphicsRAM
2600K @ 4.5 GHz Asus P8P67 Pro Sli gtx470 825/1650/1800 OCZ 2 x 2gig 1866 
Hard DriveOSMonitorKeyboard
2 600gb raid 0 / 120ssd / 2tb 7 Ultimate LG 23 
PowerCaseMouseMouse Pad
Corsair HX1000 Cooler Master HAF-X Razer Mamba Razer Vespula 
  hide details  
Reply
24/7
(13 items)
 
  
CPUMotherboardGraphicsRAM
2600K @ 4.5 GHz Asus P8P67 Pro Sli gtx470 825/1650/1800 OCZ 2 x 2gig 1866 
Hard DriveOSMonitorKeyboard
2 600gb raid 0 / 120ssd / 2tb 7 Ultimate LG 23 
PowerCaseMouseMouse Pad
Corsair HX1000 Cooler Master HAF-X Razer Mamba Razer Vespula 
  hide details  
Reply
post #43 of 103
Quote:
Originally Posted by verbatim81973 View Post
The ATI/Nvidia war was cute at first, but now it kinda making me sick. Reviews on websites are slinging their weight to one card (4850, 9800gtx) or the other. Now after reading this I really do not trust 3dmark scores either.

Just gonna wait on the OCN faithfull to buy these dang cards and get some realtime gaming expertease on some of our fav games and then go with it.
I know what you mean verbatim this happens every time a new card comes out. Its pretty much better to just not read anything until the cards have been out for a month at least so driver issues + people here/xtremesystems get the actual cards and personally do the tests.

Sometimes it is fun arguing with fanboys just to waste some time though.

Edit: Sorry about the double post.
24/7
(13 items)
 
  
CPUMotherboardGraphicsRAM
2600K @ 4.5 GHz Asus P8P67 Pro Sli gtx470 825/1650/1800 OCZ 2 x 2gig 1866 
Hard DriveOSMonitorKeyboard
2 600gb raid 0 / 120ssd / 2tb 7 Ultimate LG 23 
PowerCaseMouseMouse Pad
Corsair HX1000 Cooler Master HAF-X Razer Mamba Razer Vespula 
  hide details  
Reply
24/7
(13 items)
 
  
CPUMotherboardGraphicsRAM
2600K @ 4.5 GHz Asus P8P67 Pro Sli gtx470 825/1650/1800 OCZ 2 x 2gig 1866 
Hard DriveOSMonitorKeyboard
2 600gb raid 0 / 120ssd / 2tb 7 Ultimate LG 23 
PowerCaseMouseMouse Pad
Corsair HX1000 Cooler Master HAF-X Razer Mamba Razer Vespula 
  hide details  
Reply
post #44 of 103
While I can see how one could label this cheating, it really cannot be called such as of yet. First and formost, the drivers and PhysX app are both press related releases meaning that they are both just to show Nvidia's progess, beta's. The way I understand it is that they are placing some of the load on the GPU rather than the CPU, because it's faster. This is what it comes down to in games, one doesn't care about aesthetics when comparing two cards, just which performs best in games. It's true that not alot of games supported Ageia's PhysX, and none support Nvidia's as of yet, but that was when the card was seperate piece. Now that Nvidia is implementing it into their own existing cards it will be alot more widely adopted as it's at no extra cost other than the video card. Basically meaning that after games have adopted PhysX, like we've been told that many future games will, it will increase game performance if you have a compatible card; then it won't be just an increase on a sythetic benchmark like Vantage.
Put simply, it's not cheating when Nvidia has adopted/implemented something that ATi hasn't (I'm petty sure that ATi is working on a similar suit though). In the end, it's just a synthetic benchmark and doesn't reflect real-world performance as of yet. I agree that a seperate test for PhysX would be more appropriate though, not just a jacked up CPU score. Or maybe even a CPU score divided into part PhysX and part CPU. (IE: PhysX 22000 CPU: 7000 Total: 29000)
I'm not looking to get flamed and no I'm not a fanboy. If I'm wrong in any area just kindly point it out and I'll fix it.
Backup
(13 items)
 
  
CPUMotherboardGraphicsRAM
Core 2 Duo E6300 @ 1.86GHz Intel DG965MQ XFX GeForce 8800GT 512MB Alpha Dog Edition mushkin 2GB (2 x 1GB) DDR2 800 @ 5-5-5-12 
Hard DriveOptical DriveOSMonitor
Seagate Barracuda 250GB + Samsung HD501LJ 500GB Toshiba 20X DVD±R Burner Windows XP MCE x86 19" Gateway FPD1975W LCD 
PowerCaseMouse
Corsair VX550 uBTX Wireless Laser 
  hide details  
Reply
Backup
(13 items)
 
  
CPUMotherboardGraphicsRAM
Core 2 Duo E6300 @ 1.86GHz Intel DG965MQ XFX GeForce 8800GT 512MB Alpha Dog Edition mushkin 2GB (2 x 1GB) DDR2 800 @ 5-5-5-12 
Hard DriveOptical DriveOSMonitor
Seagate Barracuda 250GB + Samsung HD501LJ 500GB Toshiba 20X DVD±R Burner Windows XP MCE x86 19" Gateway FPD1975W LCD 
PowerCaseMouse
Corsair VX550 uBTX Wireless Laser 
  hide details  
Reply
post #45 of 103
The CPU tests are designed to stress the CPU/PhysX card, so I'm assuming the GPU is normally more or less idle during this test. Nvidia is capitalizing on this GPU downtime to do other processing and consequently give an enormous performance boost.

In a real-life game, however, you would not see nearly the same results because the GPU would be so pre-occupied with the graphics portion, it wouldn't have any downtime to help with physics processing.

It definitely weights things in Nvidia's favour, but I'd put the blame on Futuremark for creating such disjointed tests that only test one element at time.
RAID0R
(14 items)
 
  
CPUMotherboardGraphicsRAM
i5 750 4.0GHz MSI P55-GD80 GTX 470 | 8800GT PhysX 2x2GB G.Skill Ripjaws 
Hard DriveOptical DriveCoolingOS
60GB Agility 2|1TB RAID0|1.5TB Pioneer DVR-217D XSPC Raystorm | XSPC RX240 Windows 7 Professional x64 
MonitorKeyboardPowerCase
27" Dell 2709W | 17" Samsung Logitech G15 Corsair HX850 Corsair 650D 
Mouse
Microsoft IntelliMouse 
  hide details  
Reply
RAID0R
(14 items)
 
  
CPUMotherboardGraphicsRAM
i5 750 4.0GHz MSI P55-GD80 GTX 470 | 8800GT PhysX 2x2GB G.Skill Ripjaws 
Hard DriveOptical DriveCoolingOS
60GB Agility 2|1TB RAID0|1.5TB Pioneer DVR-217D XSPC Raystorm | XSPC RX240 Windows 7 Professional x64 
MonitorKeyboardPowerCase
27" Dell 2709W | 17" Samsung Logitech G15 Corsair HX850 Corsair 650D 
Mouse
Microsoft IntelliMouse 
  hide details  
Reply
post #46 of 103
Quote:
Originally Posted by Somenoob View Post
All games that use Unreal Engine 3.0. And Age of Empires...for Mac IIRC. Only notable games I know.
Once again, Age of Empires does not use PhysX. It uses Havok engine. I own all of their games, so really, I should know.

Edit: Maybe the Mac port uses PhysX, but I doubt it as this means that the games core physics engine was changed. On Windows it is most certainly Havok and nothing but Havok.

Also, to reiterate again, for the moment the 3DMark scores with PhysX should not be compared to ATI scores. However, the burden here should not be on NVIDIA to "not use PhysX" but rather on Futuremark to patch their application to have three sets of tests, GPU, CPU, and Physics if any processor is present.
Edited by dejanh - 6/23/08 at 3:17pm
Etch-a-Sketch
(16 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i7 3930K Gigabyte GA-X79S-UP5 2x MSI R5850 OC 1.350V/1060/1260 in CrossFireX 64GB Corsair Dominator 1866MHz, 9-10-9-27-180-1T 
Hard DriveCoolingMonitorKeyboard
2x Corsair Force 120GB GT SSD in RAID-0, 2x WD ... EK Supreme HF (EN), HW Labs BI GTX480, HW Labs ... Samsung S27A850 Logitech G19 Gaming Keyboard 
PowerCaseMouseAudio
Corsair Professional Series Gold CMPSU-1200AX Antec 1200 Gamer Case, Custom Full-window Side ... Logitech G9 Laser Mouse Auzentech X-Fi Forte 
Other
Bigfoot Networks Killer Xeno Pro 
  hide details  
Reply
Etch-a-Sketch
(16 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i7 3930K Gigabyte GA-X79S-UP5 2x MSI R5850 OC 1.350V/1060/1260 in CrossFireX 64GB Corsair Dominator 1866MHz, 9-10-9-27-180-1T 
Hard DriveCoolingMonitorKeyboard
2x Corsair Force 120GB GT SSD in RAID-0, 2x WD ... EK Supreme HF (EN), HW Labs BI GTX480, HW Labs ... Samsung S27A850 Logitech G19 Gaming Keyboard 
PowerCaseMouseAudio
Corsair Professional Series Gold CMPSU-1200AX Antec 1200 Gamer Case, Custom Full-window Side ... Logitech G9 Laser Mouse Auzentech X-Fi Forte 
Other
Bigfoot Networks Killer Xeno Pro 
  hide details  
Reply
post #47 of 103
Quote:
Originally Posted by dejanh View Post
Once again, Age of Empires does not use PhysX. It uses Havok engine. I own all of their games, so really, I should know.

Edit: Maybe the Mac port uses PhysX, but I doubt it as this means that the games core physics engine was changed. On Windows it is most certainly Havok and nothing but Havok.

Also, to reiterate again, for the moment the 3DMark scores with PhysX should not be compared to ATI scores. However, the burden here should not be on NVIDIA to "not use PhysX" but rather on Futuremark to patch their application to have three sets of tests, GPU, CPU, and Physics if any processor is present.
I agree, there are a lot of things that should be done:

1) DX10.1 should be a part of the tests, a seperate one just as PhysX should be......I have yet to see any figures for DX10.1 improvements and it would be interesting to see.

2) PhysX should be put on a separate test like stated, it was ok for it to be a CPU test before because no cards actually did physics so much. But now that Nvidia has stepped up it should be corrected. It is early days yet and I reckon this will either be done in a major update or in the next 3DMark....when ever that is

3) People really need to stop comparing GPU performance with 3DMark scores, if you love 3DMark scores that much then buy a 2900XT
Smallville
(16 items)
 
  
CPUMotherboardGraphicsGraphics
Intel i5 3570k Asus Sabertooth Z77 Asus HD7950 DirectCU II Asus HD7950 DirectCU II 
RAMHard DriveOptical DriveCooling
8GB DDR3 Corsair Vengeance 1600Mhz 128GB Corsair Force Pro, 2 x 2TB HDD, 500GB Ext... LG Blu-Ray, DVD/CD RW drive Corsair H60 
OSMonitorKeyboardPower
Windows 8.1 Pro 64-bit Genuine 27" Crossover 27Q S-IPS 1440p Cyborg V7 Corsair HX1000 
CaseMouseMouse PadAudio
NZXT Phantom White Microsoft Sidewinder X8 Wireless Steelseries QcK Sennheiser eH250 
  hide details  
Reply
Smallville
(16 items)
 
  
CPUMotherboardGraphicsGraphics
Intel i5 3570k Asus Sabertooth Z77 Asus HD7950 DirectCU II Asus HD7950 DirectCU II 
RAMHard DriveOptical DriveCooling
8GB DDR3 Corsair Vengeance 1600Mhz 128GB Corsair Force Pro, 2 x 2TB HDD, 500GB Ext... LG Blu-Ray, DVD/CD RW drive Corsair H60 
OSMonitorKeyboardPower
Windows 8.1 Pro 64-bit Genuine 27" Crossover 27Q S-IPS 1440p Cyborg V7 Corsair HX1000 
CaseMouseMouse PadAudio
NZXT Phantom White Microsoft Sidewinder X8 Wireless Steelseries QcK Sennheiser eH250 
  hide details  
Reply
post #48 of 103
A score is a score does it really matter how you go about obtaining the score?
Edited by AMD2600 - 6/23/08 at 3:49pm
My System
(14 items)
 
  
CPUMotherboardGraphicsRAM
1090t ASUS Sabertooth 990fx PNY GTX470 2x4g GSkill 1333 
Hard DriveOptical DriveCoolingOS
2x 500gig WD DVD ±RW Black TRUE Vista Ultimate SP2 
MonitorPowerCase
Gateway FPD2275 Corsair HX1000 Antec 1200 
  hide details  
Reply
My System
(14 items)
 
  
CPUMotherboardGraphicsRAM
1090t ASUS Sabertooth 990fx PNY GTX470 2x4g GSkill 1333 
Hard DriveOptical DriveCoolingOS
2x 500gig WD DVD ±RW Black TRUE Vista Ultimate SP2 
MonitorPowerCase
Gateway FPD2275 Corsair HX1000 Antec 1200 
  hide details  
Reply
post #49 of 103
^ Lol, of course it does. If you are going to use this for comparative benchmarks, how you set up your test matters entirely. Would you trust a popularity poll if they tell you that they polled a total of 10 people in a population of 100 million?

Think about what you said...
Etch-a-Sketch
(16 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i7 3930K Gigabyte GA-X79S-UP5 2x MSI R5850 OC 1.350V/1060/1260 in CrossFireX 64GB Corsair Dominator 1866MHz, 9-10-9-27-180-1T 
Hard DriveCoolingMonitorKeyboard
2x Corsair Force 120GB GT SSD in RAID-0, 2x WD ... EK Supreme HF (EN), HW Labs BI GTX480, HW Labs ... Samsung S27A850 Logitech G19 Gaming Keyboard 
PowerCaseMouseAudio
Corsair Professional Series Gold CMPSU-1200AX Antec 1200 Gamer Case, Custom Full-window Side ... Logitech G9 Laser Mouse Auzentech X-Fi Forte 
Other
Bigfoot Networks Killer Xeno Pro 
  hide details  
Reply
Etch-a-Sketch
(16 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i7 3930K Gigabyte GA-X79S-UP5 2x MSI R5850 OC 1.350V/1060/1260 in CrossFireX 64GB Corsair Dominator 1866MHz, 9-10-9-27-180-1T 
Hard DriveCoolingMonitorKeyboard
2x Corsair Force 120GB GT SSD in RAID-0, 2x WD ... EK Supreme HF (EN), HW Labs BI GTX480, HW Labs ... Samsung S27A850 Logitech G19 Gaming Keyboard 
PowerCaseMouseAudio
Corsair Professional Series Gold CMPSU-1200AX Antec 1200 Gamer Case, Custom Full-window Side ... Logitech G9 Laser Mouse Auzentech X-Fi Forte 
Other
Bigfoot Networks Killer Xeno Pro 
  hide details  
Reply
post #50 of 103
It's clearly a bought and paid for benchmark anyhow... and that's coming from the guy who uses 5 nvidia cards. Raytracing? (NV JUST SO HAPPENED to buy a ray tracing company.) Ageia phyX? (Oops, NV just bought them too!) Lets not even get into the rest of the optimization I personally feel is there for NV and not for ATI.

And the PhysX bench? The most ridiculously misguiding test ever in a benchmark... because in a REAL game your GPU isn't processing 90% of its cycles away on your physics. 90% of your GPU is processing graphics and 10% of that will be in physics. It's horse****, really, bad form from nVidia. However I do have a little MEANINGLESS FUN running up 15k+ vantage scores because of the most skewed set of tests in the history of futuremark (and mad onion for that matter).
Edited by Ihatethedukes - 6/23/08 at 4:00pm
Raptor Raper
(13 items)
 
  
CPUMotherboardGraphicsRAM
Something with transistors FOXCONN Boodrage -> eVGA classified e760 GTX 480 / HD6850@1090/1200 3x2 gig G Skill Trident DDR2000 
Hard DriveOptical DriveOSMonitor
2x128GB Crucial C300 Sumsung SATA DVD-RW 64 bit Windows 7 Samsung 245BW 
KeyboardPowerCaseMouse
Microsoft curve Tagan ITZ 800W DD Torture Rack Diamondback 
  hide details  
Reply
Raptor Raper
(13 items)
 
  
CPUMotherboardGraphicsRAM
Something with transistors FOXCONN Boodrage -> eVGA classified e760 GTX 480 / HD6850@1090/1200 3x2 gig G Skill Trident DDR2000 
Hard DriveOptical DriveOSMonitor
2x128GB Crucial C300 Sumsung SATA DVD-RW 64 bit Windows 7 Samsung 245BW 
KeyboardPowerCaseMouse
Microsoft curve Tagan ITZ 800W DD Torture Rack Diamondback 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Software News
Overclock.net › Forums › Industry News › Software News › [INQ] Nvidia cheats on 3DMark with 177.39 drivers