Overclock.net › Forums › Industry News › Hardware News › [GameGPU.ru] The Fastest and Most Optimal GPU and CPU Solutions for Gaming in 2015
New Posts  All Forums:Forum Nav:

[GameGPU.ru] The Fastest and Most Optimal GPU and CPU Solutions for Gaming in 2015 - Page 6

post #51 of 65

How'd you know? :D guess it's been a slow 4th QTR and we have no real facts yet about what's about to drop. VR isn't in our hands just yet(almost)... And I'm taking it all out on you lol. 

 

And all i see is postcount padding about something trivial. Oops slipped again.

post #52 of 65
Quote:
Originally Posted by rainzor View Post

How is 280x matching gtx780 when it's 15% slower, just like it was when it launched.

Ah yeah, as for GK104 cards, well only 2GB VRAM is your answer, there is no gimping involved. Look at the big difference between GTX960 2GB vs 4GB for example..
VRAM can't explain why an R7 370 (Pitcairn Pro) equals a GTX 760 and an R9 270X (Pitcairn XT) is quickly closing in on a GTX 770. Charts (Click to show)
GTX 760 review:
760 is 1.5x as fast as 7850


Now:

760 equals R7 370

Then:

760 was 1.5x the 7850
770 was 1.5x the 7870
780 was 1.21x the 7970 GHz

Now:

760 is 1x the R7 370
770 is 1.07x the R9 270X
780 is 1.13x the 280X

Relatively GK110 hasn't suffered much (although 290X is equaling or exceeding the 780 Ti at 1080p), but GK104 is getting ravaged by Pitcairn and Tahiti.
Edited by Pyrotagonist - 1/24/16 at 8:39pm
post #53 of 65
Quote:
Originally Posted by Pyrotagonist View Post

VRAM can't explain why an R7 370 (Pitcairn Pro) equals a GTX 760 and an R9 270X (Pitcairn XT) is quickly closing in on a GTX 770.

I think it is because AMD had terrible drivers at launch not that Nvidia is nerfing performance now. Pitcairn and Tahiti always had relatively good hardware but the drivers couldn't utilize it effectively. Also some newer games are using more compute which AMD's GCN is pretty good at.

GCN has aged well compared to Kepler (especially GK104) but this is due to changing graphics workloads and improving drivers on AMD's part, not nefarious nerfing by Nvidia.
Desktop
(18 items)
 
RAID
(17 items)
 
 
CPUMotherboardGraphicsRAM
Intel i9-7900X @ 4.7GHz Asus ROG Rampage VI Apex Titan X (Pascal) @ 2.05GHz 32GB G.SKILL DDR4 4000-17-18-18-41-1T 
Hard DriveCoolingOSMonitor
Samsung 960 Pro 2TB EK Custom Water Loop Windows 10 Pro x64 Acer XB270HU bprz 
MonitorMonitorKeyboardPower
HTC Vive LG OLED55C7P Logitech G810 Seasonic PRIME 1200 Platinum 
CaseMouseMouse PadAudio
Old Marble Slab Logitech G900 Logitech G440 Sennheiser HD 600 
AudioOther
Creative SoundBlasterX AE-5 Mellanox ConnectX-3 MCX312A-XCBT 10 GbE Adapter 
CPUMotherboardGraphicsRAM
i7-5960X @ 4.2GHz Asus Rampage 5 Extreme Nvidia GeForce GT 545 32GB DDR4 (2400-12-12-12-28-1T) 
Hard DriveHard DriveHard DriveOptical Drive
Samsung 950 Pro M.2 512GB HGST NAS 4TB x8 - 21.8TB RAID6 Western Digital Black 4TB Samsung SH-S183L 
CoolingOSMonitorKeyboard
Noctua NH-D15 Windows 10 Pro Asus VG278H WASD "CODE" Keyboard 
PowerCaseMouseOther
SeaSonic Platinum-1000 DIYPC Alpha-GT3 Logitech G700s Mellanox ConnectX-3 MCX312A-XCBT 10 GbE Adapter 
Other
Adaptec RAID 71605 
  hide details  
Reply
Desktop
(18 items)
 
RAID
(17 items)
 
 
CPUMotherboardGraphicsRAM
Intel i9-7900X @ 4.7GHz Asus ROG Rampage VI Apex Titan X (Pascal) @ 2.05GHz 32GB G.SKILL DDR4 4000-17-18-18-41-1T 
Hard DriveCoolingOSMonitor
Samsung 960 Pro 2TB EK Custom Water Loop Windows 10 Pro x64 Acer XB270HU bprz 
MonitorMonitorKeyboardPower
HTC Vive LG OLED55C7P Logitech G810 Seasonic PRIME 1200 Platinum 
CaseMouseMouse PadAudio
Old Marble Slab Logitech G900 Logitech G440 Sennheiser HD 600 
AudioOther
Creative SoundBlasterX AE-5 Mellanox ConnectX-3 MCX312A-XCBT 10 GbE Adapter 
CPUMotherboardGraphicsRAM
i7-5960X @ 4.2GHz Asus Rampage 5 Extreme Nvidia GeForce GT 545 32GB DDR4 (2400-12-12-12-28-1T) 
Hard DriveHard DriveHard DriveOptical Drive
Samsung 950 Pro M.2 512GB HGST NAS 4TB x8 - 21.8TB RAID6 Western Digital Black 4TB Samsung SH-S183L 
CoolingOSMonitorKeyboard
Noctua NH-D15 Windows 10 Pro Asus VG278H WASD "CODE" Keyboard 
PowerCaseMouseOther
SeaSonic Platinum-1000 DIYPC Alpha-GT3 Logitech G700s Mellanox ConnectX-3 MCX312A-XCBT 10 GbE Adapter 
Other
Adaptec RAID 71605 
  hide details  
Reply
post #54 of 65
Thread Starter 
Quote:
Originally Posted by Serandur View Post

I am aware, but as far as judging relative CPU performance, it doesn't matter. I'm not judging their ability or any personal aspects, but 2015 results with almost entirely 2013 or older only CPUs isn't the most informative regardless of reason. Simply put, it's self-evident that missing a tick-tock cycle and the most popular enthusiast CPUs of 2015 doesn't make for a comprehensive result.

While it is not ideal, it is still the most comprehensive test currently offered by any reviewing site. As far as I'm aware, there is no other benchmarking reviewing site, that takes every single review of a game that they did previously, that came out in 2015 and then averages overall performance for both GPUs and CPUs.

If you know of another website that does this, I would gladly include their results in the OP. However, I believe only GameGPU.ru takes the time to revisit these older reviews and average the overall performance for these numerous Triple A titles released the previous year. Surely, it would be much more comprehensive to have more variety in their CPU tests, but their GPU tests aren't all that bad (even though having more cards would only improve the results, as would having more CPUs). However, again, I believe this is the only website that does this and hopefully it might inspire other websites to do the same at the beginning of each new year.
Edited by BiG StroOnZ - 1/24/16 at 10:41pm
post #55 of 65
Quote:
Originally Posted by Pyrotagonist View Post

VRAM can't explain why an R7 370 (Pitcairn Pro) equals a GTX 760 and an R9 270X (Pitcairn XT) is quickly closing in on a GTX 770. Charts (Click to show)
GTX 760 review:
760 is 1.5x as fast as 7850


Now:

760 equals R7 370

Then:

760 was 1.5x the 7850
770 was 1.5x the 7870
780 was 1.21x the 7970 GHz

Now:

760 is 1x the R7 370
770 is 1.07x the R9 270X
780 is 1.13x the 280X

Relatively GK110 hasn't suffered much (although 290X is equaling or exceeding the 780 Ti at 1080p), but GK104 is getting ravaged by Pitcairn and Tahiti.
I love this quote smile.gif
Quote:
Originally Posted by tkenietz View Post

It battled the 6 series, it came back and battled the 7 series, now it's returned again after a year in the shadows to wage war with the 9 series! Pitcairn the gladiator!
That could be one of the aib special editions.

Next year 470 Ghz edition feat pitcairn shrink lol
Main
(17 items)
 
Nintendo DS
(8 items)
 
 
CPUMotherboardGraphicsRAM
6700K Gigabyte Z170X-Gaming 3 MSI GTX 1080 Gaming X G.Skill Ripjaws V 16GB 3000 
Hard DriveHard DriveCoolingCooling
Samsung 850 Evo 500GB WD Blue WD30EZRZ 3TB Noctua NH-D15S Nanoxia Deep Silence 140mm, 1100 RPM, 68.5 CFM 
CoolingOSMonitorKeyboard
Nanoxia Deep Silence 120mm, 1300 RPM, 60.1 CFM Windows 10 Enterprise Zalcom ZM27Q1 27" 1440p Glossy IPS Leopold FC750 (MX Brown) 
PowerCaseMouseAudio
EVGA 650 G2 Corsair 400Q Logitech Performance Mouse MX Mayflower Objective2 + ODAC Rev. B Combo 
Audio
Audio-Technica ATH-A990Z 
CPUCPURAMHard Drive
ARM946E-S 67.028 MHz  ARM7TDMI 33.514 MHz  4 MB 256 kB 
OSMonitorMonitorPower
DS OS 3" 256×192 18-bit 3" 256×192 18-bit 850 mAh 
  hide details  
Reply
Main
(17 items)
 
Nintendo DS
(8 items)
 
 
CPUMotherboardGraphicsRAM
6700K Gigabyte Z170X-Gaming 3 MSI GTX 1080 Gaming X G.Skill Ripjaws V 16GB 3000 
Hard DriveHard DriveCoolingCooling
Samsung 850 Evo 500GB WD Blue WD30EZRZ 3TB Noctua NH-D15S Nanoxia Deep Silence 140mm, 1100 RPM, 68.5 CFM 
CoolingOSMonitorKeyboard
Nanoxia Deep Silence 120mm, 1300 RPM, 60.1 CFM Windows 10 Enterprise Zalcom ZM27Q1 27" 1440p Glossy IPS Leopold FC750 (MX Brown) 
PowerCaseMouseAudio
EVGA 650 G2 Corsair 400Q Logitech Performance Mouse MX Mayflower Objective2 + ODAC Rev. B Combo 
Audio
Audio-Technica ATH-A990Z 
CPUCPURAMHard Drive
ARM946E-S 67.028 MHz  ARM7TDMI 33.514 MHz  4 MB 256 kB 
OSMonitorMonitorPower
DS OS 3" 256×192 18-bit 3" 256×192 18-bit 850 mAh 
  hide details  
Reply
post #56 of 65
Quote:
Originally Posted by Pyrotagonist View Post

VRAM can't explain why an R7 370 (Pitcairn Pro) equals a GTX 760 and an R9 270X (Pitcairn XT) is quickly closing in on a GTX 770. Charts (Click to show)
GTX 760 review:
760 is 1.5x as fast as 7850


Now:

760 equals R7 370

Then:

760 was 1.5x the 7850
770 was 1.5x the 7870
780 was 1.21x the 7970 GHz

Now:

760 is 1x the R7 370
770 is 1.07x the R9 270X
780 is 1.13x the 280X

Relatively GK110 hasn't suffered much (although 290X is equaling or exceeding the 780 Ti at 1080p), but GK104 is getting ravaged by Pitcairn and Tahiti.

I'd say quite a bit of it could be explained away by the much lower CPU overhead incurred under WDDM 2.0 in Win10 vs WDDM 1.1 in Win7.

Regardless, that does bring up an important point which Asmodian had already alluded to, which is the terrible launch drivers. Now imagine what the competitive landscape would've looked like had AMD got all their stuff in order on launch day.
post #57 of 65
Quote:
Originally Posted by Pyrotagonist View Post

VRAM can't explain why an R7 370 (Pitcairn Pro) equals a GTX 760 and an R9 270X (Pitcairn XT) is quickly closing in on a GTX 770. Charts (Click to show)
GTX 760 review:
760 is 1.5x as fast as 7850.

First of all, there is a 10% difference between r9 370 and gtx760, they are not equal, just like 7850 and r9 370 are not equal.
760 was 40% faster than 7850, now that is down to 30%. Let's not skew numbers to fit our story. Lack of VRAM will hit faster GPU more, it is quite obvious. And as somebody already mentioned AMD as usual needs more time to write a decent drivers and get most out of their cards. But hey, gimping is teh word
post #58 of 65
Quote:
Originally Posted by Pyrotagonist View Post

VRAM can't explain why an R7 370 (Pitcairn Pro) equals a GTX 760 and an R9 270X (Pitcairn XT) is quickly closing in on a GTX 770. Charts (Click to show)
GTX 760 review:
760 is 1.5x as fast as 7850


Now:

760 equals R7 370

Then:

760 was 1.5x the 7850
770 was 1.5x the 7870
780 was 1.21x the 7970 GHz

Now:

760 is 1x the R7 370
770 is 1.07x the R9 270X
780 is 1.13x the 280X

I'm pretty sure the biggest differences are that TPU has drastically changed the games they used since GTX 760 as well as the card suite. That will definitely change the average relative performance of each card.

760|Now

AC3| AC:U
BF3| BF3
Bio Inf.| BF4
BL2| Civ. Beyond Earth
COD:BO2| COD: AW
Call of Juarez| Crysis 3
Crysis| FC4
Crysis 3| GTA V
Diablo III| Mad Max
FC3| MGS V
Grid 2| Ryse
Hitman: Absolution| Shadow of Mordor
Metro:LL| Witcher 3
Sleeping Dogs| WoW: WoD
SC2|
TESV
Tomb Raider|
WoW|

So, the only games that have remained constant is BF3 and Crysis 3.

Only arranged how TPU arranged them in the graphs:
6970| 380
650 Ti| 380x
660| 770
580| 780
7870| 290
7950| 970
660 Ti| 290x
760| 390
7970| 390x
670| 780 Ti
7970G| 980
680| Fury
770|
780|
Titan|

Only the 780 and 760 remained constant.
Edited by airisom2 - 1/25/16 at 5:25am
post #59 of 65
since the 760 are present on both graphs, and both HD7970 and R9-280X have the same die, theres nothing wrong in comparing it indirectly.
and even if the games were switched to newer games, that just means Nvidia 700 series aren't getting any driver optimizations for the newer games.
post #60 of 65
Indirect comparisons never work. I've seen a good amount of threads turn into pointless arguments because of people throwing up five benchmarks that shows one card performing a certain way, and someone else throwing up another 3 to prove that it doesn't, and that it actually performs worse. Then, they get into how one review website is biased, or they're using different drivers, or the 290X is throttling so the results aren't representative of the card's potential, or they use a different testing rig, or you can't compare the same games because they can test in different areas, or they don't show frametimes so that review is pointless, etc. It just doesn't work tongue.gif
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Hardware News
Overclock.net › Forums › Industry News › Hardware News › [GameGPU.ru] The Fastest and Most Optimal GPU and CPU Solutions for Gaming in 2015