Overclock.net › Forums › Intel › Intel - General › [Official] - Xeon X5660-X58 Review & Discussion [and Xeon L5639 benchmarks inside]
New Posts  All Forums:Forum Nav:

[Official] - Xeon X5660-X58 Review & Discussion [and Xeon L5639 benchmarks inside] - Page 727

Poll Results: Was my review helpful?

 
  • 95% (244)
    Yes
  • 4% (12)
    No
256 Total Votes  
post #7261 of 7924
Thread Starter 
Quote:
Originally Posted by 99belle99 View Post

How does the X58 fair with AMD in crossfire as there is no SLI link on the cards and all done through the PCIe socket. As you all know we have the aging PCIe 2.0 and does it cause problems with crossfire?

I'm thinking about getting another Fury X for SLI since the prices have dropped for a few brands recently. I might be able to tell you more soon.

Quote:
The biggest issues i have run into using x58 and crossfire is CPU overhead in dx11 and possibly in dx12 if timespy is anything to go by for dx12 future. You will find overhead more a threat than pcie bandwidth. When it works crossfire seems fine on x58 for me.

I'm running a single Fury X DX11 overhead isn't that much of a problem for me at 1440p and 4K. I get roughly 50fps average @ 4K in Batman: Arkham Knight 100% maxed with Nvidia GW. I get 40fps average in Ryse: Son of Rome 4K 100% maxed and those are two great looking titles. I believe I got around 35 or 38fps average in the Witcher 3 @ 4K 100% maxed [no Nvidia gameworks].

DX12\Vulkan removes those limitations if programmed correctly judging by Hitman DX12 [4K: 44fps] and Doom Vulkan [4K: approx. 60fps+].

Time Spy shouldn't' be taken seriously until they get their stuff together with the rendering path.

Quote:
but at 1080p AMD fiji in crossfire can have poor scaling due to cpu overhead.

Why in the world would you run a $650 GPU @ 1080p. There's a better way to spend money for 1080p lol. The card was marketed as a 1440p\4K GPU.

According to a benchmark I viewed the Fury X CFX scaled 94.50% in DX11-FireStrike 1080p Graphics Score.
At 1440p the Fury X scaled 97.74% DX11-FireStrike Graphics Score
At 4K the Fury X scaled 94.66 DX11-FireStrike Graphics Score

Ummm that's some really good scaling across the board for a synthetic benchmark.
    
CPUMotherboardGraphicsRAM
Xeon 5660 @ 4.8Ghz [Highest OC 5.4Ghz] ASUS Sabertooth X58 AMD Fury X 24GB - 1600Mhz Triple Channel 
Hard DriveHard DriveHard DriveHard Drive
Seagate Barracuda 7200 1TB RAID 0 - B Seagate Barracuda 7200 1TB RAID 0 - B Seagate Barracuda 7200 1TB RAID 0 - C Seagate Barracuda 7200 1TB RAID 0 - C 
Hard DriveHard DriveHard DriveCooling
SSD 128GB RAID - A SSD 128GB RAID - A SSD 256GB  Antec Kuhler H2O 620 [Pull] 
OSMonitorPowerOther
Windows 10 Professional  Dual 24-inch Monitors EVGA SuperNOVA G2 1300W x2 Delta FFB1212EH-F00 Fan 4,000rpm  
Other
x4 Scythe Gentle Typhoon D1225C12BBAP-31 Fan 54... 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
Xeon 5660 @ 4.8Ghz [Highest OC 5.4Ghz] ASUS Sabertooth X58 AMD Fury X 24GB - 1600Mhz Triple Channel 
Hard DriveHard DriveHard DriveHard Drive
Seagate Barracuda 7200 1TB RAID 0 - B Seagate Barracuda 7200 1TB RAID 0 - B Seagate Barracuda 7200 1TB RAID 0 - C Seagate Barracuda 7200 1TB RAID 0 - C 
Hard DriveHard DriveHard DriveCooling
SSD 128GB RAID - A SSD 128GB RAID - A SSD 256GB  Antec Kuhler H2O 620 [Pull] 
OSMonitorPowerOther
Windows 10 Professional  Dual 24-inch Monitors EVGA SuperNOVA G2 1300W x2 Delta FFB1212EH-F00 Fan 4,000rpm  
Other
x4 Scythe Gentle Typhoon D1225C12BBAP-31 Fan 54... 
  hide details  
Reply
post #7262 of 7924
I'm sitting pascal out. Its a bit of a trainwreck. Waiting on Volta now. My 780ti is fending off decently.

Crossfire using the PCI-E bus isn't super limited, PCI-E 2.0 x16 is 8GB/s vs PCI-E 3.0 x16 at 16GB/s(granted its WAY slower than the memory bus) but its still limited at some degree
post #7263 of 7924
Yeah, same here but I am completely bypassing Nvidia for Vega. Interested to see how it does
     
CPUMotherboardGraphicsRAM
i7 970 between 4 - 4.4GHz Gigabyte UD7 Rev 2.0 EVGA GTX 1070 FE 2000MHz CL8.8.8.24 1T tRFC 88 
Hard DriveHard DriveCoolingOS
Samsung Evo 128Gb OCZ VertexLE 100Gb Megahalems  Windows 10 Pro 64 
MonitorKeyboardPowerCase
Asus VG248QE Corsair K70  EVGA Supernova T2 1000W Lian Li PC-A20 
MouseAudio
Sidewinder MX5 Creative Zx 
  hide details  
Reply
     
CPUMotherboardGraphicsRAM
i7 970 between 4 - 4.4GHz Gigabyte UD7 Rev 2.0 EVGA GTX 1070 FE 2000MHz CL8.8.8.24 1T tRFC 88 
Hard DriveHard DriveCoolingOS
Samsung Evo 128Gb OCZ VertexLE 100Gb Megahalems  Windows 10 Pro 64 
MonitorKeyboardPowerCase
Asus VG248QE Corsair K70  EVGA Supernova T2 1000W Lian Li PC-A20 
MouseAudio
Sidewinder MX5 Creative Zx 
  hide details  
Reply
post #7264 of 7924
Don't get me wrong, if Vega comes out decent, I'm jumping trains, for the first time in 10 years
post #7265 of 7924
Last time I ran AMD was when the 5870 came out. I bought 2. Then out of curiosity I picked up a GTX 480 space heater, lol
     
CPUMotherboardGraphicsRAM
i7 970 between 4 - 4.4GHz Gigabyte UD7 Rev 2.0 EVGA GTX 1070 FE 2000MHz CL8.8.8.24 1T tRFC 88 
Hard DriveHard DriveCoolingOS
Samsung Evo 128Gb OCZ VertexLE 100Gb Megahalems  Windows 10 Pro 64 
MonitorKeyboardPowerCase
Asus VG248QE Corsair K70  EVGA Supernova T2 1000W Lian Li PC-A20 
MouseAudio
Sidewinder MX5 Creative Zx 
  hide details  
Reply
     
CPUMotherboardGraphicsRAM
i7 970 between 4 - 4.4GHz Gigabyte UD7 Rev 2.0 EVGA GTX 1070 FE 2000MHz CL8.8.8.24 1T tRFC 88 
Hard DriveHard DriveCoolingOS
Samsung Evo 128Gb OCZ VertexLE 100Gb Megahalems  Windows 10 Pro 64 
MonitorKeyboardPowerCase
Asus VG248QE Corsair K70  EVGA Supernova T2 1000W Lian Li PC-A20 
MouseAudio
Sidewinder MX5 Creative Zx 
  hide details  
Reply
post #7266 of 7924
Quote:
Originally Posted by OCmember View Post

Last time I ran AMD was when the 5870 came out. I bought 2. Then out of curiosity I picked up a GTX 480 space heater, lol
My 480gtxs taught me what psu ocp is lol. I also melted my first atx cables with them. They also heated up my garage while i benched in winter lol
post #7267 of 7924
Quote:
Originally Posted by Kana-Maru View Post

I'm thinking about getting another Fury X for SLI since the prices have dropped for a few brands recently. I might be able to tell you more soon.
I'm running a single Fury X DX11 overhead isn't that much of a problem for me at 1440p and 4K. I get roughly 50fps average @ 4K in Batman: Arkham Knight 100% maxed with Nvidia GW. I get 40fps average in Ryse: Son of Rome 4K 100% maxed and those are two great looking titles. I believe I got around 35 or 38fps average in the Witcher 3 @ 4K 100% maxed [no Nvidia gameworks].

DX12\Vulkan removes those limitations if programmed correctly judging by Hitman DX12 [4K: 44fps] and Doom Vulkan [4K: approx. 60fps+].

Time Spy shouldn't' be taken seriously until they get their stuff together with the rendering path.
Why in the world would you run a $650 GPU @ 1080p. There's a better way to spend money for 1080p lol. The card was marketed as a 1440p\4K GPU.

According to a benchmark I viewed the Fury X CFX scaled 94.50% in DX11-FireStrike 1080p Graphics Score.
At 1440p the Fury X scaled 97.74% DX11-FireStrike Graphics Score
At 4K the Fury X scaled 94.66 DX11-FireStrike Graphics Score

Ummm that's some really good scaling across the board for a synthetic benchmark.

To be fair the person who asked the question didn't specify what AMD gpus to run in crossfire. As for 650$ for 1080p gpus was never my intent. I was aiming at 4k but i do benchmark at 1080 because thats what most nvidia benches are at. If it were up to me id never bench below 4k. Unfortunately hwbot and top30 threads here on ocn are all about low resolution mostly. I also refer to overhead with cpu when my crossfire cannot be fully pushed equally. Maybe i am not using cpu overhead correctly thefe. I do know that after testing furyx cf up to quadfire each platform had different results. Worst being my 9590 on 990fx gpu load bounced all over but never staying loaded across all gpus. Then my sr2 had to go up to 1440p

Maybe its related to ipc i dunno, but i go from a bench at 1080p that cannot fully load all gpus to loading all 4 maxxed on different newer cpu.
post #7268 of 7924
Thread Starter 
Quote:
Originally Posted by dagget3450 View Post

My 480gtxs taught me what psu ocp is lol. I also melted my first atx cables with them. They also heated up my garage while i benched in winter lol

Ah those heater editions were awesome.


Quote:
Originally Posted by dagget3450 View Post

To be fair the person who asked the question didn't specify what AMD gpus to run in crossfire. As for 650$ for 1080p gpus was never my intent. I was aiming at 4k but i do benchmark at 1080 because thats what most nvidia benches are at. If it were up to me id never bench below 4k. Unfortunately hwbot and top30 threads here on ocn are all about low resolution mostly..

Yeah it's sad to see that 1080p is still "a thing" so many years after mainstream cards have had no issues running games at 1080p. Heck my old 550 Ti has no issue running Crysis 2 @ 1440p. Well it COULD run it at decent frames with 1GB. So 1080p was not a big issue way back then. Yeah I can understand the low resolutions for the latest e-peen purchase justification . I mean unless you have a decent monitor to support the fps, I don't see the point. It's like getting 150fps+ @ 1080p on a 85hz monitor.
Quote:
I also refer to overhead with cpu when my crossfire cannot be fully pushed equally. Maybe i am not using cpu overhead correctly thefe. I do know that after testing furyx cf up to quadfire each platform had different results. Worst being my 9590 on 990fx gpu load bounced all over but never staying loaded across all gpus. Then my sr2 had to go up to 1440p

I don't think all of the issues are with the CPUs or the platforms. DX11 was a limited API. You probably just hitting the limits which isn't difficultly to do with your setup. The GPU are probably starving for more work and one core just isn't feeding enough data. DX12 and Vulkan should offer more and are widely being adopted [especially Vulkan]. I do remember you saying that different platforms gave different results in the past. I'm sure newer platforms will give you the best performance over the much older platforms.

Quote:
Maybe its related to ipc i dunno, but i go from a bench at 1080p that cannot fully load all gpus to loading all 4 maxxed on different newer cpu.

It could be IPC related since DX11 doesn't really utilize all of the CPU cores. The faster you can make the IPC will definitely allow more data to be processed. Thankfully modern APIs are going to take advantage of our modern hardware. Although IPC matters, using more cores efficiently will even the odds.
    
CPUMotherboardGraphicsRAM
Xeon 5660 @ 4.8Ghz [Highest OC 5.4Ghz] ASUS Sabertooth X58 AMD Fury X 24GB - 1600Mhz Triple Channel 
Hard DriveHard DriveHard DriveHard Drive
Seagate Barracuda 7200 1TB RAID 0 - B Seagate Barracuda 7200 1TB RAID 0 - B Seagate Barracuda 7200 1TB RAID 0 - C Seagate Barracuda 7200 1TB RAID 0 - C 
Hard DriveHard DriveHard DriveCooling
SSD 128GB RAID - A SSD 128GB RAID - A SSD 256GB  Antec Kuhler H2O 620 [Pull] 
OSMonitorPowerOther
Windows 10 Professional  Dual 24-inch Monitors EVGA SuperNOVA G2 1300W x2 Delta FFB1212EH-F00 Fan 4,000rpm  
Other
x4 Scythe Gentle Typhoon D1225C12BBAP-31 Fan 54... 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
Xeon 5660 @ 4.8Ghz [Highest OC 5.4Ghz] ASUS Sabertooth X58 AMD Fury X 24GB - 1600Mhz Triple Channel 
Hard DriveHard DriveHard DriveHard Drive
Seagate Barracuda 7200 1TB RAID 0 - B Seagate Barracuda 7200 1TB RAID 0 - B Seagate Barracuda 7200 1TB RAID 0 - C Seagate Barracuda 7200 1TB RAID 0 - C 
Hard DriveHard DriveHard DriveCooling
SSD 128GB RAID - A SSD 128GB RAID - A SSD 256GB  Antec Kuhler H2O 620 [Pull] 
OSMonitorPowerOther
Windows 10 Professional  Dual 24-inch Monitors EVGA SuperNOVA G2 1300W x2 Delta FFB1212EH-F00 Fan 4,000rpm  
Other
x4 Scythe Gentle Typhoon D1225C12BBAP-31 Fan 54... 
  hide details  
Reply
post #7269 of 7924
I dread to think how many watts I hit my 470 with. Several, and then some....
post #7270 of 7924
Honestly probably somewhere around 255 to 325. I know. My 480's were probably getting my entire house three of them clocked over 800 will do that. Hence my previous power supply being 1600 Watts
ROGBeast RIP
(16 items)
 
- X58 Xeon Club -
(22 photos)
 
CPUMotherboardGraphicsRAM
Xeon W3680 asus ramage iii extreme Galaxy GTX 780 Hall of Fame Crucial Ballistix Tactical LP 
Hard DriveHard DriveOptical DriveCooling
2x Samsung 850 evo raid 0 raptor 10k blueray something EK Supremecy Evo 
CoolingMonitorMonitorKeyboard
EK 780-FC HOF asus pg279 Asus pg279 Corsair Strafe RGB 
PowerCaseMouseAudio
Cooler Master V1200 Platinum  Corsair Carbide 760T Corsair Scimitar Sound Blaster 3D recon 
  hide details  
Reply
ROGBeast RIP
(16 items)
 
- X58 Xeon Club -
(22 photos)
 
CPUMotherboardGraphicsRAM
Xeon W3680 asus ramage iii extreme Galaxy GTX 780 Hall of Fame Crucial Ballistix Tactical LP 
Hard DriveHard DriveOptical DriveCooling
2x Samsung 850 evo raid 0 raptor 10k blueray something EK Supremecy Evo 
CoolingMonitorMonitorKeyboard
EK 780-FC HOF asus pg279 Asus pg279 Corsair Strafe RGB 
PowerCaseMouseAudio
Cooler Master V1200 Platinum  Corsair Carbide 760T Corsair Scimitar Sound Blaster 3D recon 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Intel - General
Overclock.net › Forums › Intel › Intel - General › [Official] - Xeon X5660-X58 Review & Discussion [and Xeon L5639 benchmarks inside]