Overclock.net › Forums › Benchmarks › Benchmarking Software and Discussion › [OFFICIAL]--- Top 30 --- Unigine 'Valley' Benchmark 1.0
New Posts  All Forums:Forum Nav:

[OFFICIAL]--- Top 30 --- Unigine 'Valley' Benchmark 1.0 - Page 349

post #3481 of 13364
Quote:
Originally Posted by KaRLiToS View Post

In case some of you wonder how come the score are not spot on with the average. I have falsified some data.

  • First, the HD 7990 CrossFire has better score than the HD 7970 Crossfire, so in order to fix this, I decided to remove -5fps to the HD 7990 and to be fair, I did the same thing to the GTX 690 SLI
  • Second, the HD 7990 Quadfire was equal to HD 7970 Quadfire which again I think is non-sense, so I removed -4fps to it along with the GTX 690 Quad SLI to be fair on each side.
  • Third, I removed -7fps to the GTX 670 Quad-SLI (which has only one entry) in order to balance it with the GTX 680 Quad SLI, without that modification, the GTX 670 Quad would have been 12 fps higher than the GTX 680 Quad which again, I find it weird


Hi Karlitos

I'm the one with the quad-SLI 670ies...while you have not altered the main table (fortunately), I respectfully suggest that altering the data underlying the graphs is not really 'o.k.' and raises all kinds of questions...670ies have been known to beat 680/690ies on several occasions (depending on the task, particular setups etc). Here is a the HWBot World Record table for Unigine's Heaven Basic, which shows my 670ies...nothing too weird' about it. I hope you can reverse the 'artificial' fix above...

post #3482 of 13364
Quote:
Originally Posted by KaRLiToS View Post

I didn't alter any data.

You can check for yourself in the main OP, no data has been modified, all the graphs are perfectly scaled with the average .

I altered the average to make it more realistic in this graph above. (Its the only graph that have been modified. I still think what I did is fair.

There have been no mistake in the spreasheets and all scores are entered correctly.

Ok I then think I understood you wrong, my apologies. If only the graph is altered im fine with that, good work. smile.gif
Room Heater
(19 items)
 
  
CPUMotherboardGraphicsRAM
i5 3570K 4.4 Ghz AsRock Z77 Extreme4  Sapphire R9 290x Tri-X Corsair XMS3 DDR3 4x2GB 1600MHz 
Hard DriveHard DriveHard DriveCooling
Western Digital 500 GB Crucial M4 128GB Seagate Barracuda 1TB Coolermaster Hyper 212 Evo 
OSMonitorKeyboardPower
Windows 7 Professional ASUS VG248QE  Corsair Vengeance K90 Coolermaster Silent Pro 1000W 
CaseMouseMouse PadAudio
Corsair Graphite T600 SE Zowie AM Steelseries QcK Mass AKG Q701 
AudioAudioAudio
Homebuilt Tube Amplifier Homebuilt Headphone Amp SPDIF to ANALOG converter + Philips TDA 1541A DAC 
  hide details  
Reply
Room Heater
(19 items)
 
  
CPUMotherboardGraphicsRAM
i5 3570K 4.4 Ghz AsRock Z77 Extreme4  Sapphire R9 290x Tri-X Corsair XMS3 DDR3 4x2GB 1600MHz 
Hard DriveHard DriveHard DriveCooling
Western Digital 500 GB Crucial M4 128GB Seagate Barracuda 1TB Coolermaster Hyper 212 Evo 
OSMonitorKeyboardPower
Windows 7 Professional ASUS VG248QE  Corsair Vengeance K90 Coolermaster Silent Pro 1000W 
CaseMouseMouse PadAudio
Corsair Graphite T600 SE Zowie AM Steelseries QcK Mass AKG Q701 
AudioAudioAudio
Homebuilt Tube Amplifier Homebuilt Headphone Amp SPDIF to ANALOG converter + Philips TDA 1541A DAC 
  hide details  
Reply
post #3483 of 13364
Quote:
Originally Posted by JuliusCivilis View Post

Ok I then think I understood you wrong, my apologies. If only the graph is altered im fine with that, good work. smile.gif

Yes , all the interactive graphs reflects what there is in the speadsheet.

The weirdest thing is the GTX 670 Quad SLI being 12 fps higher than the GTX 680 Quad.
Quad Damage
(24 items)
 
  
CPUMotherboardGraphicsRAM
Intel i7 4930k @ 4.6Ghz Asus Rampage IV Extreme (Koolance Full WB) 2 x GTX 1080ti FE in SLI (EK WATERBLOCK) Corsair Dominator GT 2133 MHz (16GB) 
Hard DriveHard DriveHard DriveHard Drive
3 TB Seagate OCZ Vertex 4 256GB Crucial M4 128GB ADATA SX900 128GB 
CoolingCoolingCoolingCooling
RIVE Koolance Waterblock (VRM and PCH) Feser X-Changer 360 Feser X-Changer 240 Feser X-Changer 480 
CoolingCoolingCoolingOS
WaterCool MoRA3 Pro with Gentle Typhoon 3 x MCP 655 pumps (With EK top...all serial) 2 x EK FC 1080ti Blocks Windows 7 Ultimate x64 
MonitorMonitorKeyboardPower
Asus ROG Swift PG348Q (3440x1440 @100hz) CrossOver 27Q Led-P Logitech G15 EVGA SuperNova 1600T2 (Titanium) 
CaseMouseMouse PadAudio
Corsair Obsidian Heavily Modded Logitech G700 RocketFish Speed/Control M-audio AV-40 (Logitech G-35) 
  hide details  
Reply
Quad Damage
(24 items)
 
  
CPUMotherboardGraphicsRAM
Intel i7 4930k @ 4.6Ghz Asus Rampage IV Extreme (Koolance Full WB) 2 x GTX 1080ti FE in SLI (EK WATERBLOCK) Corsair Dominator GT 2133 MHz (16GB) 
Hard DriveHard DriveHard DriveHard Drive
3 TB Seagate OCZ Vertex 4 256GB Crucial M4 128GB ADATA SX900 128GB 
CoolingCoolingCoolingCooling
RIVE Koolance Waterblock (VRM and PCH) Feser X-Changer 360 Feser X-Changer 240 Feser X-Changer 480 
CoolingCoolingCoolingOS
WaterCool MoRA3 Pro with Gentle Typhoon 3 x MCP 655 pumps (With EK top...all serial) 2 x EK FC 1080ti Blocks Windows 7 Ultimate x64 
MonitorMonitorKeyboardPower
Asus ROG Swift PG348Q (3440x1440 @100hz) CrossOver 27Q Led-P Logitech G15 EVGA SuperNova 1600T2 (Titanium) 
CaseMouseMouse PadAudio
Corsair Obsidian Heavily Modded Logitech G700 RocketFish Speed/Control M-audio AV-40 (Logitech G-35) 
  hide details  
Reply
post #3484 of 13364
Yeah that's an outlier for sure. There just aren't enough samples of quad-SLI 680's to even out the numbers. Quad 670's are not 10% faster than quad 680's IMO.
post #3485 of 13364
my 2cents.gif
the tri sli 670's that are 1-2 fps higher then my dual 670's should not be incorperated in the graphs
Daily Rig
(21 items)
 
Bench Setup
(19 items)
 
 
CPUMotherboardGraphicsRAM
5820k 4.7Ghz @ 1.28vcore / 1.88vccin Gigabyte x99 SOC Champion EVGA GTX 1080Ti G.SKILL Trident Z 32GB 3200Mhz CL14 
Hard DriveHard DriveHard DriveHard Drive
2x Intel 520 120Gb SSD in RAID 0  Intel 180Gb 530 SSD OS Drive 2Tb WD HDD  2x Intel 730 SSD 480GB in RAID 0 
CoolingCoolingCoolingCooling
2x HWLabs GTX360 Black Ice in Push Pull EK blocks for CPU / GPU's 6x corsair sp120 2350RPM 6x Ek Vardar 120 F4's 2200rpm  
OSMonitorKeyboardPower
windows 10 XB270HU 2560x1440 - 144Hz - G-sync Razor blackwidow cooler master V1000 
CaseMouseMouse PadAudio
Corsair 780T EVGA TORQ x10 evga mouse pad 2.1 audio / Hyper X cloud II / Sennheiser Gam... 
Other
EK pump / res 
CPUMotherboardGraphicsGraphics
6600k Z170M OC Formula GTX780Ti AMD HD6450 
RAMHard DriveHard DriveHard Drive
Trident Z 3600Mhz cl16 2x8GB 120Gb intel 520 120Gb intel 520 64Gb adata ssd  
Hard DriveOptical DriveCoolingCooling
750Gb WD Black Sony  F1 Gemini CPU Pot Tek - 9 Fat 
CoolingOSMonitorPower
custom loop two 240rad's xp,W7,W8,W10 x-star 2560x1440  Rosewill Quark-650 Platinum / Antec HCP-850 Pla... 
CaseMouseMouse Pad
Dimastech Mini V1.0 black razor abyssus Razor 
  hide details  
Reply
Daily Rig
(21 items)
 
Bench Setup
(19 items)
 
 
CPUMotherboardGraphicsRAM
5820k 4.7Ghz @ 1.28vcore / 1.88vccin Gigabyte x99 SOC Champion EVGA GTX 1080Ti G.SKILL Trident Z 32GB 3200Mhz CL14 
Hard DriveHard DriveHard DriveHard Drive
2x Intel 520 120Gb SSD in RAID 0  Intel 180Gb 530 SSD OS Drive 2Tb WD HDD  2x Intel 730 SSD 480GB in RAID 0 
CoolingCoolingCoolingCooling
2x HWLabs GTX360 Black Ice in Push Pull EK blocks for CPU / GPU's 6x corsair sp120 2350RPM 6x Ek Vardar 120 F4's 2200rpm  
OSMonitorKeyboardPower
windows 10 XB270HU 2560x1440 - 144Hz - G-sync Razor blackwidow cooler master V1000 
CaseMouseMouse PadAudio
Corsair 780T EVGA TORQ x10 evga mouse pad 2.1 audio / Hyper X cloud II / Sennheiser Gam... 
Other
EK pump / res 
CPUMotherboardGraphicsGraphics
6600k Z170M OC Formula GTX780Ti AMD HD6450 
RAMHard DriveHard DriveHard Drive
Trident Z 3600Mhz cl16 2x8GB 120Gb intel 520 120Gb intel 520 64Gb adata ssd  
Hard DriveOptical DriveCoolingCooling
750Gb WD Black Sony  F1 Gemini CPU Pot Tek - 9 Fat 
CoolingOSMonitorPower
custom loop two 240rad's xp,W7,W8,W10 x-star 2560x1440  Rosewill Quark-650 Platinum / Antec HCP-850 Pla... 
CaseMouseMouse Pad
Dimastech Mini V1.0 black razor abyssus Razor 
  hide details  
Reply
post #3486 of 13364
Quote:
Originally Posted by lilchronic View Post

my 2cents.gif
the tri sli 670's that are 1-2 fps higher then my dual 670's should not be incorperated in the graphs


...my point exactly - This is an incredibly slippery slope this is on - you cannot change real world observations to fit expectations and then report it as data, because once you do, you open the flood gates, not to mention leave your otherwise excellent thread open to hard critique

...but let me propose s.th. which might help: Why don't you mark what are in your opinion 'unusual results' in your graphs with a little ' * ' / footnote that reads: small sample size, or alternatively list how many observations you have per GPU / config 'bar' in your graphs
post #3487 of 13364
Quote:
Originally Posted by Joa3d43 View Post

...my point exactly - This is an incredibly slippery slope this is on - you cannot change real world observations to fit expectations and then report it as data, because once you do, you open the flood gates, not to mention leave your otherwise excellent thread open to hard critique

...but let me propose s.th. which might help: Why don't you mark what are in your opinion 'unusual results' in your graphs with a little ' * ' / footnote that reads: small sample size, or alternatively list how many observations you have per GPU / config 'bar' in your graphs

Its a great idea.

Believe me, its a lot of work to do this thread and its even more harder to take decision alone.

There you go then guys. smile.gif



ahh, that HD 7990 Crossfire rolleyes.gif
Edited by KaRLiToS - 4/10/13 at 3:13pm
Quad Damage
(24 items)
 
  
CPUMotherboardGraphicsRAM
Intel i7 4930k @ 4.6Ghz Asus Rampage IV Extreme (Koolance Full WB) 2 x GTX 1080ti FE in SLI (EK WATERBLOCK) Corsair Dominator GT 2133 MHz (16GB) 
Hard DriveHard DriveHard DriveHard Drive
3 TB Seagate OCZ Vertex 4 256GB Crucial M4 128GB ADATA SX900 128GB 
CoolingCoolingCoolingCooling
RIVE Koolance Waterblock (VRM and PCH) Feser X-Changer 360 Feser X-Changer 240 Feser X-Changer 480 
CoolingCoolingCoolingOS
WaterCool MoRA3 Pro with Gentle Typhoon 3 x MCP 655 pumps (With EK top...all serial) 2 x EK FC 1080ti Blocks Windows 7 Ultimate x64 
MonitorMonitorKeyboardPower
Asus ROG Swift PG348Q (3440x1440 @100hz) CrossOver 27Q Led-P Logitech G15 EVGA SuperNova 1600T2 (Titanium) 
CaseMouseMouse PadAudio
Corsair Obsidian Heavily Modded Logitech G700 RocketFish Speed/Control M-audio AV-40 (Logitech G-35) 
  hide details  
Reply
Quad Damage
(24 items)
 
  
CPUMotherboardGraphicsRAM
Intel i7 4930k @ 4.6Ghz Asus Rampage IV Extreme (Koolance Full WB) 2 x GTX 1080ti FE in SLI (EK WATERBLOCK) Corsair Dominator GT 2133 MHz (16GB) 
Hard DriveHard DriveHard DriveHard Drive
3 TB Seagate OCZ Vertex 4 256GB Crucial M4 128GB ADATA SX900 128GB 
CoolingCoolingCoolingCooling
RIVE Koolance Waterblock (VRM and PCH) Feser X-Changer 360 Feser X-Changer 240 Feser X-Changer 480 
CoolingCoolingCoolingOS
WaterCool MoRA3 Pro with Gentle Typhoon 3 x MCP 655 pumps (With EK top...all serial) 2 x EK FC 1080ti Blocks Windows 7 Ultimate x64 
MonitorMonitorKeyboardPower
Asus ROG Swift PG348Q (3440x1440 @100hz) CrossOver 27Q Led-P Logitech G15 EVGA SuperNova 1600T2 (Titanium) 
CaseMouseMouse PadAudio
Corsair Obsidian Heavily Modded Logitech G700 RocketFish Speed/Control M-audio AV-40 (Logitech G-35) 
  hide details  
Reply
post #3488 of 13364
Quote:
Originally Posted by Majin SSJ Eric View Post

Yeah that's an outlier for sure. There just aren't enough samples of quad-SLI 680's to even out the numbers. Quad 670's are not 10% faster than quad 680's IMO.

This is true, but the graphs are showing the averages of the scores in the thread, & if Joa3d43 has the only quad sli 670s his score is the average.
Changing his results to match the way things should be seems pretty odd, I'd leave it & shame the 680 quad sli users into trying harder.

The 680 quads are perfectly capable of beating Joa3d43's 670s, grow a pair & get the average up guys!
    
CPUMotherboardGraphicsRAM
i7 990x/3930k/3770k/4770k GB x58a-OC/Asus RIVE/x79 gd65/z77 mpower/z77 MV... gtx480s/gtx580s/7970s/680s/Titan Hyper/PSC/BBSE/Hynix/Samsung 
Hard DriveOptical DriveCoolingOS
HDDs/SSDs DVD Air/water/phase/copper pot Windowz 
MonitorKeyboardPowerCase
samsung 2320 Razer Lycosa ax1200 carboard box 
Mouse
Trackball 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
i7 990x/3930k/3770k/4770k GB x58a-OC/Asus RIVE/x79 gd65/z77 mpower/z77 MV... gtx480s/gtx580s/7970s/680s/Titan Hyper/PSC/BBSE/Hynix/Samsung 
Hard DriveOptical DriveCoolingOS
HDDs/SSDs DVD Air/water/phase/copper pot Windowz 
MonitorKeyboardPowerCase
samsung 2320 Razer Lycosa ax1200 carboard box 
Mouse
Trackball 
  hide details  
Reply
post #3489 of 13364
Personally, I do not like sliding averages merely because the raw data doesn't fit into someone predetermined idea of how something 'should' perform.

I just had to defend myself, and have other defend me, to convince some people that my 7970 tri-fire score of 6054/144.7fps was legit.

Remember, we are benching ONE benchmark. There is no doubt my 3 7970s "shouldn't" beat 3 titans, but it did, and that is not my or anyone else's problem. We all know scaling with valley 1.0 isn't exactly great in either camp, but more so for titans/Kepler. But you can't go and change averages because "quad 670s shouldnt beat quad 680s", its just ridiculous and makes me (and a few others I'm sure) very skeptical of the whole process.

If 7990 cfx is equal to 7970 x4 for the data we have, but you think it shouldn't so you change the overall average to reflect what you believed is correct, then I have no faith of trust in the data presented anywhere. Regardless that you say only that graph has these arbitrary averages, how is anyone suppose to trust any other graph?
Edited by jomama22 - 4/10/13 at 3:25pm
     
CPUMotherboardGraphicsRAM
intel 3960x @4.7 ghz asrock extreme9 x79 3x msi 7970 lightnings g.skill ripjawx 1600 cl8 
Hard DriveOptical DriveCoolingMonitor
ssd: kingston hyperx 120gb hdd: seagate barracu... bd-rwh water for cpu benq 2420t 24" 120hz 
Case
mountain mods extended asension 
CPUMotherboardGraphicsRAM
3960x Asrock Extreme9 2x Sapphire 6950 unlocked G.skill CL8 12800 x4 (4gb) 
Hard DriveHard DriveOptical DriveCooling
Kingston hyperx  Seagate barracuda xt 7200/ sata3/ 64 mb cache LiteOn BD-R x12 Watercooling 
OSMonitorPowerCase
win 7 x64 asus 23" 1080p 1000W Thermaltake toughpower HAX 932 Advanced (modified for Radiators on ins... 
MouseAudio
Razer Deathhadder 3.5g sennheiser PC 330, Game blaster with Extreme9 
  hide details  
Reply
     
CPUMotherboardGraphicsRAM
intel 3960x @4.7 ghz asrock extreme9 x79 3x msi 7970 lightnings g.skill ripjawx 1600 cl8 
Hard DriveOptical DriveCoolingMonitor
ssd: kingston hyperx 120gb hdd: seagate barracu... bd-rwh water for cpu benq 2420t 24" 120hz 
Case
mountain mods extended asension 
CPUMotherboardGraphicsRAM
3960x Asrock Extreme9 2x Sapphire 6950 unlocked G.skill CL8 12800 x4 (4gb) 
Hard DriveHard DriveOptical DriveCooling
Kingston hyperx  Seagate barracuda xt 7200/ sata3/ 64 mb cache LiteOn BD-R x12 Watercooling 
OSMonitorPowerCase
win 7 x64 asus 23" 1080p 1000W Thermaltake toughpower HAX 932 Advanced (modified for Radiators on ins... 
MouseAudio
Razer Deathhadder 3.5g sennheiser PC 330, Game blaster with Extreme9 
  hide details  
Reply
post #3490 of 13364
Quote:
Originally Posted by KaRLiToS View Post

Its a great idea.

Believe me, its a lot of work to do this thread and its even more harder to take decision alone.

There you go then guys. smile.gif



ahh, that HD 7990 Crossfire rolleyes.gif

..tx for considering the concerns....your underlying reasoning was perfectly 'ok', it just needed an extra few steps to achieve what you wanted...and this is a really great, popular thread, all the more reason NOT to leave it open to 'yeah but they fix the data' critiques (even if that was not true in the first place)

...for what it is worth, there are some occasions when 670ies beat 680ies - I don't really know why, though I suspect it may relate to voltage 'per CUDA core', 'internal' heat etc...on the other hand, I am in the process of locating some 4x 680 Lightnings...once I have, I fix that graph for you permanently by restoring 670 / 680 balance biggrin.gif
New Posts  All Forums:Forum Nav:
  Return Home
Overclock.net › Forums › Benchmarks › Benchmarking Software and Discussion › [OFFICIAL]--- Top 30 --- Unigine 'Valley' Benchmark 1.0