Overclock.net › Forums › Industry News › Hardware News › [[H]ard|OCP] NVIDIA 3-Way SLI and AMD Tri-Fire Redux
New Posts  All Forums:Forum Nav:

[[H]ard|OCP] NVIDIA 3-Way SLI and AMD Tri-Fire Redux - Page 6

post #51 of 168
Quote:
Originally Posted by The-Beast View Post
But running 2 GPU in x4 is, as long as they're AMD's GPU.
The AMD GPU's were never running at 4x, IIRC they were running at 16x 16x in the original review, however, the 3rd 580 WAS running at 4x and that's why we saw such poor scaling on the 3rd card, and of course the obvious CPU bottleneck that so many pointed to.

Anyone building such a high end system on the x58 platform is going to use a 980x or a 990x and will be clocked @ 4.5 at bare minimum, just look at CallsignVega's rig for example.

And oh yea, thanks [H]ardOCP for re-running this review and addressing the CPU bottleneck feedback!
Edited by 2010rig - 5/3/11 at 12:24pm
2010rig
(14 items)
 
Galaxy S3
(8 items)
 
 
CPUMotherboardGraphicsRAM
X5660 @ 4.5  ASUS P6X58D-E 980TI? 12GB OCZ Platinum - 7-7-7-21 
Hard DriveCoolingOSMonitor
1 80GB SSD x25m - 3TB F3 + F4 NH-D14 Windows 7 Ultimate LG 47LH55 
KeyboardPowerCaseMouse
Natural Wireless Keyboard Corsair 750HX CM 690 II Advanced MX 518 
CPUGraphicsRAMHard Drive
Snapdragon S4 Dual core 1500mhz Adreno 225 Samsung 2GB 16GB Onboard Flash 
OSMonitorPowerCase
Android 4.4.2 - CM11 4.8" AMOLED 1280x720 2100 mAh battery Otterbox Defender 
  hide details  
Reply
2010rig
(14 items)
 
Galaxy S3
(8 items)
 
 
CPUMotherboardGraphicsRAM
X5660 @ 4.5  ASUS P6X58D-E 980TI? 12GB OCZ Platinum - 7-7-7-21 
Hard DriveCoolingOSMonitor
1 80GB SSD x25m - 3TB F3 + F4 NH-D14 Windows 7 Ultimate LG 47LH55 
KeyboardPowerCaseMouse
Natural Wireless Keyboard Corsair 750HX CM 690 II Advanced MX 518 
CPUGraphicsRAMHard Drive
Snapdragon S4 Dual core 1500mhz Adreno 225 Samsung 2GB 16GB Onboard Flash 
OSMonitorPowerCase
Android 4.4.2 - CM11 4.8" AMOLED 1280x720 2100 mAh battery Otterbox Defender 
  hide details  
Reply
post #52 of 168
Quote:
Originally Posted by Seeing Red View Post
I just added an extra source of info. I didn't state what part of the bandwidth argument it applied to. You can use this data to come up with your own conclusions
In what way does it relate to the results of this benchmark result then?
post #53 of 168
Quote:
Originally Posted by EastCoast View Post
In what way does it relate to the results of this benchmark result then?
Let me explain it then, it wasn't to nolify the review but certain arguments made against it. It basically puts to rest the conspiracy theories of the NF200 hindering AMD crossfire performance. It also shows that it actually helps AMD cards somewhat. Same P67 NF200 mobo used in both reviews even though the CPUs are clocked differently.

I never said anything about the data addressing any concerns about multiple GPUs on one PCIe slot. If you don't find the data useful, then just move on.
Lil Burninator
(14 items)
 
  
CPUMotherboardGraphicsRAM
i7 4770K ASUS Gryphon CF R9 290Xs Samsung 4x4GB 
Hard DriveCoolingOSMonitor
Samsung EVO 1TB 2x 240s | 360 | MPC35X | Mostly BP W7 Pre x64 XL2720T 
KeyboardPowerCaseMouse
K70 AX1200 CaseLabs S5 + Ped G700/Mamba 
Mouse PadAudio
Vespula WA7 + KRK RP6s + HD650s 
  hide details  
Reply
Lil Burninator
(14 items)
 
  
CPUMotherboardGraphicsRAM
i7 4770K ASUS Gryphon CF R9 290Xs Samsung 4x4GB 
Hard DriveCoolingOSMonitor
Samsung EVO 1TB 2x 240s | 360 | MPC35X | Mostly BP W7 Pre x64 XL2720T 
KeyboardPowerCaseMouse
K70 AX1200 CaseLabs S5 + Ped G700/Mamba 
Mouse PadAudio
Vespula WA7 + KRK RP6s + HD650s 
  hide details  
Reply
post #54 of 168
It kills me sometimes that our computer review articles are done without the scientific method in mind. Reading HOCP articles just grind my gears knowing they could've found out so much more, and laid 6 pages (so far) of debate to rest.

I like the tom's article being supplementary to HOCP's findings. However there are more specific issues that weren't brought to light, such as architecture and speeds.
    
CPUMotherboardGraphicsRAM
i7 920 Doh eVGA 759 Classy Gigabyte GTX 470 6 gb Mushkin Blackline 
Hard DriveOptical DriveOSMonitor
80gb SSD, 2x640 Black, 2x1TB Green, 640 Green Samsung DVD-RW W7 64bit 1920x1200 Asus 25.5 incher 
KeyboardPowerCaseMouse
Realforce 86ub + HHKbp2 Silverstone Zeus 850W Ravenfish Rayzor Naga | Sseries Xai 
Mouse Pad
Rayzor Sphex | PT Talent 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
i7 920 Doh eVGA 759 Classy Gigabyte GTX 470 6 gb Mushkin Blackline 
Hard DriveOptical DriveOSMonitor
80gb SSD, 2x640 Black, 2x1TB Green, 640 Green Samsung DVD-RW W7 64bit 1920x1200 Asus 25.5 incher 
KeyboardPowerCaseMouse
Realforce 86ub + HHKbp2 Silverstone Zeus 850W Ravenfish Rayzor Naga | Sseries Xai 
Mouse Pad
Rayzor Sphex | PT Talent 
  hide details  
Reply
post #55 of 168
Quote:
Originally Posted by OC'ing Noob View Post
In all fairness, an i7-920 is still an extremely fast CPU and considering the long standing trend of games no longer being CPU limited, it is kind of a stretch to assume that the cards are being limited by a high end processor as opposed to the cards simply reaching the limits engine
The numbers are in it WAS limited by the CPU. But this is on tri SLI. One card would not be limited and two GPU's limit would likely be around 4.0Ghz for best performance on any game. You have to realize the GPU uses the CPU to feed it information. Just cause a game isn't at all limited by the CPU the CPU may be limiting the GPU's in multi GPU rigs. But again what was supprising was AMD wasn't as limited by the CPU.
 
4x r290's
(9 items)
 
3x 7970's
(16 items)
 
CPUMotherboardGraphicsGraphics
i7 3930K 4.8Ghz offset ASUS X79 Rampage IV Extreme GTX 780 with EK waterblock GTX 780 with EK waterblock 
GraphicsRAMHard DriveHard Drive
GTX 780 with EK waterblock  Samsung 16GB (4x4gb) 2400Mhz 10-12-12-31-1t @ 1... 256GB Samsung 950 Pro 256GB Samsung 840 Pro 
CoolingCoolingOSMonitor
RX360, EX360, EX360, and 480 Rad's XSPC Raystorm CPU Block Dual boot (other unnamed OS) and Win8 Pro QNIX QX2710 27" 2560x1440 @ 112hz 
KeyboardPowerCase
Corsair K90 Corsair AX1200 Corsair 700D Modded w/ External Radiators 
CPUMotherboardGraphicsRAM
Intel i5 3570K GA-Z77X-UD3H 4x R9 290's 8GB 1600Mhz 
Hard DriveOSPowerCase
1TB Hard Drive Win8 Corsair Dual PSU - AX1200 and HX650 Open Case 
Other
Thee 16x Risers and One 1x Riser 
CPUMotherboardGraphicsRAM
i7 3930K 5Ghz Bench ASUS X79 Rampage IV Extreme 3x AMD 7970's w/ EK water blocks Samsung 16GB (4x4gb) 2400Mhz 10-12-12-31-1t @ 1... 
Hard DriveCoolingCoolingMonitor
128GB Samsung 830  RX360, RX240, EX360 XSPC Raystorm CPU Block QNIX QX2710 27" 2560x1440 @ 114hz 
KeyboardPowerCase
Corsair K90 Corsair AX760i Corsair 700D Modded w/ External Radiators 
  hide details  
Reply
 
4x r290's
(9 items)
 
3x 7970's
(16 items)
 
CPUMotherboardGraphicsGraphics
i7 3930K 4.8Ghz offset ASUS X79 Rampage IV Extreme GTX 780 with EK waterblock GTX 780 with EK waterblock 
GraphicsRAMHard DriveHard Drive
GTX 780 with EK waterblock  Samsung 16GB (4x4gb) 2400Mhz 10-12-12-31-1t @ 1... 256GB Samsung 950 Pro 256GB Samsung 840 Pro 
CoolingCoolingOSMonitor
RX360, EX360, EX360, and 480 Rad's XSPC Raystorm CPU Block Dual boot (other unnamed OS) and Win8 Pro QNIX QX2710 27" 2560x1440 @ 112hz 
KeyboardPowerCase
Corsair K90 Corsair AX1200 Corsair 700D Modded w/ External Radiators 
CPUMotherboardGraphicsRAM
Intel i5 3570K GA-Z77X-UD3H 4x R9 290's 8GB 1600Mhz 
Hard DriveOSPowerCase
1TB Hard Drive Win8 Corsair Dual PSU - AX1200 and HX650 Open Case 
Other
Thee 16x Risers and One 1x Riser 
CPUMotherboardGraphicsRAM
i7 3930K 5Ghz Bench ASUS X79 Rampage IV Extreme 3x AMD 7970's w/ EK water blocks Samsung 16GB (4x4gb) 2400Mhz 10-12-12-31-1t @ 1... 
Hard DriveCoolingCoolingMonitor
128GB Samsung 830  RX360, RX240, EX360 XSPC Raystorm CPU Block QNIX QX2710 27" 2560x1440 @ 114hz 
KeyboardPowerCase
Corsair K90 Corsair AX760i Corsair 700D Modded w/ External Radiators 
  hide details  
Reply
post #56 of 168
Quote:
Originally Posted by Seeing Red View Post
Let me explain it then, it wasn't to nolify the review but certain arguments made against it. It basically puts to rest the conspiracy theories of the NF200 hindering AMD crossfire performance. It also shows that it actually helps AMD cards somewhat. Same P67 NF200 mobo used in both reviews even though the CPUs are clocked differently.

I never said anything about the data addressing any concerns about multiple GPUs on one PCIe slot. If you don't find the data useful, then just move on.
In what way exactly is it helping AMD? Honestly, I don't see it unless you are only referencing the results in the review you link. Which is why I am asking how it relates to the article in the OP? You are talking about 2 different PC configurations and are imply that bandwidth wise they offer similar results which doesn't hinder one set of cards from the other (2-3 6950 vs 6990/6970). So I ask in what way is that so with 2 different PC configurations based on the article you linked to in relation to the review in the OP?
post #57 of 168
Quote:
Originally Posted by EastCoast View Post
In what way exactly is it helping AMD? Honestly, I don't see it unless you are only referencing the results in the review you link. Which is why I am asking how it relates to the article in the OP? You are talking about 2 different PC configurations and are imply that bandwidth wise they offer similar results which doesn't hinder one set of cards from the other (2-3 6950 vs 6990/6970). So I ask in what way is that so with 2 different PC configurations based on the article you linked to in relation to the review in the OP?
I'm refering to the Tomshardware review concerning the NF200 chip performance. There is no way to draw a conclusion about the NF200 chip's affect alone from the [H]ardOCP review without involving different CPU architectures. Tomshardware compares a non-NF200 board to a NF200 board for both CPUs. It also shows a slight gain, pretty small but it's still there, on the AMD GPUs with the NF200 chip.

One of the arguments was that the NF200 chip hinders AMD cards' performance. If it didn't hinder the performance of the 6950s (AMD card) why would the NF200 hinder the performance of the 6970s (another AMD card).

TL;DR : The information doesn't have anything to do with the review.
Lil Burninator
(14 items)
 
  
CPUMotherboardGraphicsRAM
i7 4770K ASUS Gryphon CF R9 290Xs Samsung 4x4GB 
Hard DriveCoolingOSMonitor
Samsung EVO 1TB 2x 240s | 360 | MPC35X | Mostly BP W7 Pre x64 XL2720T 
KeyboardPowerCaseMouse
K70 AX1200 CaseLabs S5 + Ped G700/Mamba 
Mouse PadAudio
Vespula WA7 + KRK RP6s + HD650s 
  hide details  
Reply
Lil Burninator
(14 items)
 
  
CPUMotherboardGraphicsRAM
i7 4770K ASUS Gryphon CF R9 290Xs Samsung 4x4GB 
Hard DriveCoolingOSMonitor
Samsung EVO 1TB 2x 240s | 360 | MPC35X | Mostly BP W7 Pre x64 XL2720T 
KeyboardPowerCaseMouse
K70 AX1200 CaseLabs S5 + Ped G700/Mamba 
Mouse PadAudio
Vespula WA7 + KRK RP6s + HD650s 
  hide details  
Reply
post #58 of 168
Wait, okay someone explain to me how this was a good test when the AMD cards were experiencing negative gains from going from a 3.6ghz 920 to a 4.8ghz 2600k.

Just doesn't make sense to me at least. -14.4% then -1.4%...
Gnomeon
(16 items)
 
GnomeBook Pro
(13 items)
 
 
CPUMotherboardGraphicsRAM
Intel Xeon E3-1245 V2 ASUS P8Z77-V SAPPHIRE TRI-X OC Radeon R9 290X 4GB  Crucial Ballistix Tactical 16GB DDR3 1866 
Hard DriveHard DriveCoolingOS
Crucial M550 256GB HGST Travelstar 7K1000 2.5-Inch 1TB 7200 RPM Cooler Master Hyper 212 Evo Windows 10 Pro 64-Bit 
MonitorKeyboardPowerCase
BenQ XL2730Z Cooler Master Quickfire Pro Unbranded (Tenkeyless) SeaSonic SG12 550W Fractal Design Define S 
MouseAudioAudioOther
Mionix Avior Swan M10 Philips Fidelio X1 Audio-Technica ATR2500-USB 
CPUMotherboardGraphicsRAM
Intel Core i7-2720QM Logic Board Intel HD Graphics 3000 and AMD 6750M 1gb Crucial 16GB 1333MHZ 
Hard DriveOptical DriveOSMonitor
750gb Hitachi 5400rpm 8x double-layer SuperDrive OS x 10.7.4 15.4" LED-backlit glossy widescreen display 144... 
KeyboardPowerCaseMouse
Apple Keyboard Apple 85W MagSafe Power Adapter 2011 Unibody Magic Trackpad 
Audio
Built-in Stereo Speakers with Subwoofer 
  hide details  
Reply
Gnomeon
(16 items)
 
GnomeBook Pro
(13 items)
 
 
CPUMotherboardGraphicsRAM
Intel Xeon E3-1245 V2 ASUS P8Z77-V SAPPHIRE TRI-X OC Radeon R9 290X 4GB  Crucial Ballistix Tactical 16GB DDR3 1866 
Hard DriveHard DriveCoolingOS
Crucial M550 256GB HGST Travelstar 7K1000 2.5-Inch 1TB 7200 RPM Cooler Master Hyper 212 Evo Windows 10 Pro 64-Bit 
MonitorKeyboardPowerCase
BenQ XL2730Z Cooler Master Quickfire Pro Unbranded (Tenkeyless) SeaSonic SG12 550W Fractal Design Define S 
MouseAudioAudioOther
Mionix Avior Swan M10 Philips Fidelio X1 Audio-Technica ATR2500-USB 
CPUMotherboardGraphicsRAM
Intel Core i7-2720QM Logic Board Intel HD Graphics 3000 and AMD 6750M 1gb Crucial 16GB 1333MHZ 
Hard DriveOptical DriveOSMonitor
750gb Hitachi 5400rpm 8x double-layer SuperDrive OS x 10.7.4 15.4" LED-backlit glossy widescreen display 144... 
KeyboardPowerCaseMouse
Apple Keyboard Apple 85W MagSafe Power Adapter 2011 Unibody Magic Trackpad 
Audio
Built-in Stereo Speakers with Subwoofer 
  hide details  
Reply
post #59 of 168
One thing i wonder is how much of this performance gain comes from the new cpu and how much comes from the fact its not in an x4 slot
Server/Gaming Rig
(22 items)
 
  
CPUCPUMotherboardGraphics
Intel E5-2670 intel E5-2670 Asus Z9PE-D8 WS Visiontek R9 390 
RAMHard DriveHard DriveHard Drive
8x4GB DDR3 1333 ECC Registered Toshiba 5TB X300 Seagate 4TB HGST 4TB 
Hard DriveHard DriveHard DriveHard Drive
Sandisk 240GB Sandisk 480GB Samsung 830 128GB Toshiba 6TB X300 
Hard DriveHard DriveOptical DriveOS
Toshiba 5TB X300 Toshiba 5TB X300 WH16NS40 Windows 10 Pro 
MonitorMonitorKeyboardPower
Dell S1715H Dell u2311h blackwidow ultimate Kingwin Lazer platinum 1000w 
CaseMouse
Corsair 800D Deathadder Chroma 
  hide details  
Reply
Server/Gaming Rig
(22 items)
 
  
CPUCPUMotherboardGraphics
Intel E5-2670 intel E5-2670 Asus Z9PE-D8 WS Visiontek R9 390 
RAMHard DriveHard DriveHard Drive
8x4GB DDR3 1333 ECC Registered Toshiba 5TB X300 Seagate 4TB HGST 4TB 
Hard DriveHard DriveHard DriveHard Drive
Sandisk 240GB Sandisk 480GB Samsung 830 128GB Toshiba 6TB X300 
Hard DriveHard DriveOptical DriveOS
Toshiba 5TB X300 Toshiba 5TB X300 WH16NS40 Windows 10 Pro 
MonitorMonitorKeyboardPower
Dell S1715H Dell u2311h blackwidow ultimate Kingwin Lazer platinum 1000w 
CaseMouse
Corsair 800D Deathadder Chroma 
  hide details  
Reply
post #60 of 168
If you step carefully around the fan boys this could be a very interesting read, but as said above the testing was done in such an unscientific way that we cant draw any usefull conclusions.

If the findings had been that there was increases or at the very least no movement from both setups then it may have been an interesting insight into how you can become cpu limited in some extream cases but as it is, with AMD actualy loosing performance it is very hard to take the review as being all that meaningfull outside of you buying those exact setups.
Main
(15 items)
 
  
CPUMotherboardGraphicsRAM
2500K Gigabyte P67 UD5 r7970 BEDD Corsair Vengeance 
Hard DriveCoolingOSMonitor
Crucial C300 256gb EK HF supreme full copper Win 7 Ultimate 120hz 1920x1080 23" LG 
KeyboardPowerCaseMouse
Ducky ocn edition 1000w Corsair hx Corsair 800D G9x 
Mouse PadAudio
Steelseries I-2 AKG K702 
  hide details  
Reply
Main
(15 items)
 
  
CPUMotherboardGraphicsRAM
2500K Gigabyte P67 UD5 r7970 BEDD Corsair Vengeance 
Hard DriveCoolingOSMonitor
Crucial C300 256gb EK HF supreme full copper Win 7 Ultimate 120hz 1920x1080 23" LG 
KeyboardPowerCaseMouse
Ducky ocn edition 1000w Corsair hx Corsair 800D G9x 
Mouse PadAudio
Steelseries I-2 AKG K702 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Hardware News
Overclock.net › Forums › Industry News › Hardware News › [[H]ard|OCP] NVIDIA 3-Way SLI and AMD Tri-Fire Redux