Overclock.net › Forums › Industry News › Hardware News › [HardOCP] Crossfire R9 Fury vs SLI GTX 980 vs Crossfire 390X @ 4K Resolution
New Posts  All Forums:Forum Nav:

[HardOCP] Crossfire R9 Fury vs SLI GTX 980 vs Crossfire 390X @ 4K Resolution

post #1 of 168
Thread Starter 
Quote:













We wanted to fully bring you Far Cry 4 performance out of ASUS STRIX Radeon R9 Fury DC3 CrossFire and AMD Radeon R9 390X CrossFire at 4K. Unfortunately AMD CrossFire is completely broken and not working correctly in this game.



Conclusion:

Warning: Spoiler! (Click to show)

It was clear in most games that ASUS STRIX Radeon R9 Fury DC3 CrossFire has the potential to be the fastest configuration. Without a shadow of a doubt ASUS STRIX Radeon R9 Fury DC3 CrossFire was faster than GeForce GTX 980 SLI in Dying Light, Grand Theft Auto V, and Battlefield 4. ASUS STRIX Radeon R9 Fury DC3 CrossFire was potentially faster in The Witcher 3 but it is harder to tell with that game due to bughy performance. ASUS STRIX Radeon R9 Fury DC3 CrossFire was not faster in Project Cars and it was broken in FC4. That means in 3/6 games in our suite ASUS STRIX Radeon R9 Fury DC3 CrossFire provided better performance. However, also in 3/6 games AMD CrossFire had some kind of issue that NVIDIA SLI did not.

It is a really mixed bag of performance. We think that ASUS STRIX Radeon R9 Fury DC3 CrossFire most certainly has the potential to be faster than GeForce GTX 980 SLI at 4K. The main issue we see is that it is software, AMD CrossFire software, that is holding back that potential. ASUS STRIX Radeon R9 Fury DC3 CrossFire performance is all over the map, it is hard to predict and find it consistent. NVIDIA SLI is more consistent in terms of what works and the efficiency of performance improvement provided.

This is really the saddening part of the whole ordeal, how many issues we encountered with AMD CrossFire versus NVIDIA SLI. We had no issues at all with NVIDIA SLI in these six games, no stutter, everything worked and it was a smooth gaming experience. With AMD CrossFire we had issues in 3 out of 6 games played.

We strongly feel that ASUS STRIX Radeon R9 Fury DC3 CrossFire has the potential to offer better performance than GeForce GTX 980 SLI. The issue lies in software support for AMD CrossFire versus NVIDIA SLI. In one of the games we tested AMD CrossFire doesn't even work at all, in two others there was extreme stuttering and odd framerate behavior. This is half of our game suite that exhibited an issue with AMD CrossFire.

Therein lies the main issue, there is a distinct discrepancy between AMD CrossFire and NVIDIA SLI in today's games. It is an issue of software support, and efficient use of the hardware feature through software in games. AMD has more problems with this than NVIDIA does. NVIDIA is on the ball at each new game release releasing "Game Ready" GeForce drivers with SLI profiles, AMD, isn't.

When it works, ASUS STRIX Radeon R9 Fury DC3 CrossFire can be a top performer. In the games it does work in, ASUS STRIX Radeon R9 Fury DC3 CrossFire was a lot of fun playing games on at 4K and provided an advantage compared to the competition.

ASUS STRIX Radeon R9 Fury DC3 CrossFire is not inexpensive, at near $1100 it is big investment on GPU gaming performance. Gamers spending that much money for gaming performance should demand nothing but the best. This solution is more expensive than GeForce GTX 980 SLI but can offer more performance when CrossFire is working. However, if your particular game you are playing has issues with CrossFire, a better overall solution will be GeForce GTX 980 SLI.

If it comes between the Radeon R9 390X CrossFire and GeForce GTX 980 SLI we'd lean toward Radeon R9 390X CrossFire as offering a better experience, again if CrossFire works in your game, and that looks to be a big "if." AMD Radeon R9 390X CrossFire is an excellent value, has the VRAM needed for 4K and is able to keep up with GeForce GTX 980 SLI if not beating it in some games...if CrossFire is working.

Source: http://www.hardocp.com/article/2015/09/28/asus_strix_radeon_r9_fury_dc3_crossfire_at_4k_review/3
Edited by BiG StroOnZ - 9/28/15 at 7:16pm
post #2 of 168
First of all: why were the NVIDIA Project Cars run and 390X GTA5 run not redone? There's no way those drops are normal, I mean GTX 980 SLI had higher mins with DS2X than MSAAx4. Similarly 390X CF did way better on the second GTA5 run.

Also, it's weird that GTX 980 SLI lagged behind in Witcher 3 considering that it's an NVIDIA title and runs well on NVIDIA hardware. Maybe it's the high resolution plus lack of Gameworks. At least it had the minimums, but I think a 40 FPS frame rate cap would have given the AMD setups much less stuttering while still maintaining higher averages than the NVIDIA setup.

Lastly, nice FC4 graph.
Edited by B!0HaZard - 9/28/15 at 7:24pm
M1XN
(18 items)
 
Study Zenbook
(5 items)
 
 
CPUMotherboardGraphicsRAM
i7 4770K @ 4.2 GHz ASUS Maximus VI Impact MSI GTX 970 Gaming 4G 2x8 GB Kingston HyperX DDR3 @ 2400 MHz 
Hard DriveHard DriveHard DriveOptical Drive
Crucial M4 64 GB Crucial M4 128 GB SAMSUNG Spinpoint M9T 2 TB LiteOn DL-8ATSH 
CoolingOSMonitorKeyboard
Noctua NH-C14 Windows 10 64-bit ViewSonic VP2770 Vortex Pok3r 
PowerCaseMouseMouse Pad
Corsair SF600 NCASE M1 V2.5 Logitech G502 SteelSeries QcK+ 
AudioOther
HiFiMan HE-400 Logitech C920 
CPUGraphicsRAMHard Drive
i5-3317 GT 620M 2 GB soldered + 8 GB Corsair OCZ Vertex 3 120 
OS
Linux Mint 
  hide details  
Reply
M1XN
(18 items)
 
Study Zenbook
(5 items)
 
 
CPUMotherboardGraphicsRAM
i7 4770K @ 4.2 GHz ASUS Maximus VI Impact MSI GTX 970 Gaming 4G 2x8 GB Kingston HyperX DDR3 @ 2400 MHz 
Hard DriveHard DriveHard DriveOptical Drive
Crucial M4 64 GB Crucial M4 128 GB SAMSUNG Spinpoint M9T 2 TB LiteOn DL-8ATSH 
CoolingOSMonitorKeyboard
Noctua NH-C14 Windows 10 64-bit ViewSonic VP2770 Vortex Pok3r 
PowerCaseMouseMouse Pad
Corsair SF600 NCASE M1 V2.5 Logitech G502 SteelSeries QcK+ 
AudioOther
HiFiMan HE-400 Logitech C920 
CPUGraphicsRAMHard Drive
i5-3317 GT 620M 2 GB soldered + 8 GB Corsair OCZ Vertex 3 120 
OS
Linux Mint 
  hide details  
Reply
post #3 of 168
You'd almost be crazy to spend all the extra money for two Fury's over a couple of 390X's it seems to me. Sure the Fiji cards are more powerful but with the driver improvements Hawaii has enjoyed lately, the 390X's come close to matching the Fury's at a $300 discount. Of course, the 980 is probably the best overall choice simply because SLI tends to present less issues than CF overall but when CF works the Fury's and 390X's seem to have the raw power to out-muscle the GM204's...
post #4 of 168
If you had a 120+ Hz monitor, I bet vsynced CFX could rival SLI, it's really just a frame timing issue.
M1XN
(18 items)
 
Study Zenbook
(5 items)
 
 
CPUMotherboardGraphicsRAM
i7 4770K @ 4.2 GHz ASUS Maximus VI Impact MSI GTX 970 Gaming 4G 2x8 GB Kingston HyperX DDR3 @ 2400 MHz 
Hard DriveHard DriveHard DriveOptical Drive
Crucial M4 64 GB Crucial M4 128 GB SAMSUNG Spinpoint M9T 2 TB LiteOn DL-8ATSH 
CoolingOSMonitorKeyboard
Noctua NH-C14 Windows 10 64-bit ViewSonic VP2770 Vortex Pok3r 
PowerCaseMouseMouse Pad
Corsair SF600 NCASE M1 V2.5 Logitech G502 SteelSeries QcK+ 
AudioOther
HiFiMan HE-400 Logitech C920 
CPUGraphicsRAMHard Drive
i5-3317 GT 620M 2 GB soldered + 8 GB Corsair OCZ Vertex 3 120 
OS
Linux Mint 
  hide details  
Reply
M1XN
(18 items)
 
Study Zenbook
(5 items)
 
 
CPUMotherboardGraphicsRAM
i7 4770K @ 4.2 GHz ASUS Maximus VI Impact MSI GTX 970 Gaming 4G 2x8 GB Kingston HyperX DDR3 @ 2400 MHz 
Hard DriveHard DriveHard DriveOptical Drive
Crucial M4 64 GB Crucial M4 128 GB SAMSUNG Spinpoint M9T 2 TB LiteOn DL-8ATSH 
CoolingOSMonitorKeyboard
Noctua NH-C14 Windows 10 64-bit ViewSonic VP2770 Vortex Pok3r 
PowerCaseMouseMouse Pad
Corsair SF600 NCASE M1 V2.5 Logitech G502 SteelSeries QcK+ 
AudioOther
HiFiMan HE-400 Logitech C920 
CPUGraphicsRAMHard Drive
i5-3317 GT 620M 2 GB soldered + 8 GB Corsair OCZ Vertex 3 120 
OS
Linux Mint 
  hide details  
Reply
post #5 of 168
Quote:
Originally Posted by Majin SSJ Eric View Post

You'd almost be crazy to spend all the extra money for two Fury's over a couple of 390X's it seems to me. Sure the Fiji cards are more powerful but with the driver improvements Hawaii has enjoyed lately, the 390X's come close to matching the Fury's at a $300 discount. Of course, the 980 is probably the best overall choice simply because SLI tends to present less issues than CF overall but when CF works the Fury's and 390X's seem to have the raw power to out-muscle the GM204's...
Defaulting SLI as superior is nonsense.

CFX works as often as SLI, and I'm quite certain you'd have trouble producing anything other than blind bias to prove otherwise. You make it sound as if CFX is an unstable pile of crap next to SLI, while CFX has been the leader for quite some time. Even before ditching the bridge, AMD had better scaling and better frame times than NV. Better support too, in many cases.

And let's not forget some very expensive Titans that shipped with SLI barely working on anything at all. Shortly after Hawaii was released we saw Hawaii scaling better than Titan in virtually all games, with support for 3 and 4 GPU's working wonderfully while SLI was getting negative scaling above 2.

I'd be perfectly content to say they're equal, but the constant belligerence gets to me. Do some research and you'll find just as many SLI issues as you can CFX issues.
Trash Box
(15 items)
 
 
G73JH-A1
(13 items)
 
CPUMotherboardGraphicsRAM
i7 980x @ 4.4GHz Asus P6X58D-E Gigabyte 8GB GTX 1070 G1 Gaming 24GB Corsair Dominators 1600MHz 
Hard DriveHard DriveMonitorKeyboard
2x256GB C300's R0 1TB Samsung 850 Pro ASUS MG279Q, Catleap Q270 G19 
PowerCaseMouse
Corsair AX1200 Silverstone RV-02 G9x 
CPUMotherboardGraphicsRAM
AMD 965BE 125W Asus M4A77D Asus 5770 Cucore 4GB Kingston 800MHz 
Hard DriveOptical DriveOSKeyboard
2x750GB Seagates 22x LG DVD-RW w7 Ultimate 64-bit  G15 
PowerCase
Antec Earthwatts 430W 80+ Antec New Solution NSK6580B 
CPUMotherboardGraphicsRAM
720QM AsusTek Mobility HD5870 8GB Kingston 1600MHz 
Hard DriveOptical DriveOSMonitor
256GB C300/500GB Seagate Bluray Combo w7 Ultimate 64-bit 17'' 1080P 
Case
G73JH-A1 
  hide details  
Reply
Trash Box
(15 items)
 
 
G73JH-A1
(13 items)
 
CPUMotherboardGraphicsRAM
i7 980x @ 4.4GHz Asus P6X58D-E Gigabyte 8GB GTX 1070 G1 Gaming 24GB Corsair Dominators 1600MHz 
Hard DriveHard DriveMonitorKeyboard
2x256GB C300's R0 1TB Samsung 850 Pro ASUS MG279Q, Catleap Q270 G19 
PowerCaseMouse
Corsair AX1200 Silverstone RV-02 G9x 
CPUMotherboardGraphicsRAM
AMD 965BE 125W Asus M4A77D Asus 5770 Cucore 4GB Kingston 800MHz 
Hard DriveOptical DriveOSKeyboard
2x750GB Seagates 22x LG DVD-RW w7 Ultimate 64-bit  G15 
PowerCase
Antec Earthwatts 430W 80+ Antec New Solution NSK6580B 
CPUMotherboardGraphicsRAM
720QM AsusTek Mobility HD5870 8GB Kingston 1600MHz 
Hard DriveOptical DriveOSMonitor
256GB C300/500GB Seagate Bluray Combo w7 Ultimate 64-bit 17'' 1080P 
Case
G73JH-A1 
  hide details  
Reply
post #6 of 168
Thread Starter 
Quote:
Originally Posted by Mookster View Post

Defaulting SLI as superior is nonsense.

CFX works as often as SLI, and I'm quite certain you'd have trouble producing anything other than blind bias to prove otherwise. You make it sound as if CFX is an unstable pile of crap next to SLI, while CFX has been the leader for quite some time. Even before ditching the bridge, AMD had better scaling and better frame times than NV. Better support too, in many cases.

And let's not forget some very expensive Titans that shipped with SLI barely working on anything at all. Shortly after Hawaii was released we saw Hawaii scaling better than Titan in virtually all games, with support for 3 and 4 GPU's working wonderfully while SLI was getting negative scaling above 2.

I'd be perfectly content to say they're equal, but the constant belligerence gets to me. Do some research and you'll find just as many SLI issues as you can CFX issues.

Click the show spoiler and read the conclusion in the OP. Eric is basing his opinion off the conclusion that the writers of the article wrote based on their experience in the games tested with the Crossfire configurations used.
post #7 of 168
Why is the OP only highlighting the positives for the 980 and the negatives of the fury and 390X? Why not highlight this "If it comes between the Radeon R9 390X CrossFire and GeForce GTX 980 SLI we'd lean toward Radeon R9 390X CrossFire as offering a better experience."
Giga Buster
(15 items)
 
Mega Buster
(15 items)
 
 
CPUMotherboardGraphicsRAM
Intel i5-2500K Asrock P67 Extreme4 Gen3 MSI RX 470 G.SKILL Ripjaws Series 16GB DDR3 1600 
Hard DriveOptical DriveCoolingOS
Western Digital  LG DVD Burner Cooler Master Hyper 212 EVO Windows 7 Ultimate 64 Bit 
MonitorKeyboardPowerCase
COMPAQ WF1907 CM Storm Devastator EVGA 700W 80+ Bronze RAIDMAX VORTEX 
MouseMouse PadAudio
CM Storm Devastator AULA Onboard Realtek Audio 
CPUMotherboardGraphicsRAM
AMD Phenom X6 1600T @3.8ghz 1.325v M5A99X EVO EVGA GTX 460 G.SKILL Ripjaws Series 8GB DDR3 1600 
Hard DriveOptical DriveCoolingOS
Western Digital  LG DVD Burner Cooler Master Hyper TX3  Windows 7 Ultimate 64 Bit 
MonitorKeyboardPowerCase
ACER 17 inch Agiler Antec 520-HCG Super T 
MouseMouse PadAudio
EUROCASE None Onboard VIA HD 
  hide details  
Reply
Giga Buster
(15 items)
 
Mega Buster
(15 items)
 
 
CPUMotherboardGraphicsRAM
Intel i5-2500K Asrock P67 Extreme4 Gen3 MSI RX 470 G.SKILL Ripjaws Series 16GB DDR3 1600 
Hard DriveOptical DriveCoolingOS
Western Digital  LG DVD Burner Cooler Master Hyper 212 EVO Windows 7 Ultimate 64 Bit 
MonitorKeyboardPowerCase
COMPAQ WF1907 CM Storm Devastator EVGA 700W 80+ Bronze RAIDMAX VORTEX 
MouseMouse PadAudio
CM Storm Devastator AULA Onboard Realtek Audio 
CPUMotherboardGraphicsRAM
AMD Phenom X6 1600T @3.8ghz 1.325v M5A99X EVO EVGA GTX 460 G.SKILL Ripjaws Series 8GB DDR3 1600 
Hard DriveOptical DriveCoolingOS
Western Digital  LG DVD Burner Cooler Master Hyper TX3  Windows 7 Ultimate 64 Bit 
MonitorKeyboardPowerCase
ACER 17 inch Agiler Antec 520-HCG Super T 
MouseMouse PadAudio
EUROCASE None Onboard VIA HD 
  hide details  
Reply
post #8 of 168
Thread Starter 
Quote:
Originally Posted by sage101 View Post

Why is the OP only highlighting the positives for the 980 and the negatives of the fury and 390X? Why not highlight this "If it comes between the Radeon R9 390X CrossFire and GeForce GTX 980 SLI we'd lean toward Radeon R9 390X CrossFire as offering a better experience."

Well, why don't you post the entire quote first off, which is, "If it comes between the Radeon R9 390X CrossFire and GeForce GTX 980 SLI we'd lean toward Radeon R9 390X CrossFire as offering a better experience, again if CrossFire works in your game, and that looks to be a big "if." AMD Radeon R9 390X CrossFire is an excellent value, has the VRAM needed for 4K and is able to keep up with GeForce GTX 980 SLI if not beating it in some games...if CrossFire is working."

And did you miss:

It was clear in most games that ASUS STRIX Radeon R9 Fury DC3 CrossFire has the potential to be the fastest configuration.

That means in 3/6 games in our suite ASUS STRIX Radeon R9 Fury DC3 CrossFire provided better performance.

We think that ASUS STRIX Radeon R9 Fury DC3 CrossFire most certainly has the potential to be faster than GeForce GTX 980 SLI at 4K.

We strongly feel that ASUS STRIX Radeon R9 Fury DC3 CrossFire has the potential to offer better performance than GeForce GTX 980 SLI.

When it works, ASUS STRIX Radeon R9 Fury DC3 CrossFire can be a top performer.
post #9 of 168
What's up with GTA V?
post #10 of 168
Thread Starter 
Quote:
Originally Posted by Forceman View Post

What's up with GTA V?

"It seems that ASUS STRIX Radeon R9 Fury DC3 CrossFire tanks when the view distance and shadow view distance settings are maximized in this game. It is either a VRAM issue, or it is an issue with the AMD CrossFire software for this game at these settings.

GeForce GTX 980 SLI also has 4GB of VRAM per GPU and it does not tank like ASUS STRIX Radeon R9 Fury DC3 CrossFire does. Therefore, it might just be a software issue for why ASUS STRIX Radeon R9 Fury DC3 CrossFire tanks like this. It is possible data is not being swapped correctly out of local framebuffer as efficiently as GTX 980 SLI is being done in this very VRAM capacity limited test. It goes to show how software can affect the potential of performance in VRAM limited situations a great deal."

http://www.hardocp.com/article/2015/09/28/asus_strix_radeon_r9_fury_dc3_crossfire_at_4k_review/6
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Hardware News
Overclock.net › Forums › Industry News › Hardware News › [HardOCP] Crossfire R9 Fury vs SLI GTX 980 vs Crossfire 390X @ 4K Resolution