Overclock.net › Forums › Graphics Cards › NVIDIA › [Official] NVIDIA GTX 780 Ti Owner's Club
New Posts  All Forums:Forum Nav:

[Official] NVIDIA GTX 780 Ti Owner's Club - Page 966

post #9651 of 16490
Play some games like bf4, tomb raider, crysis and far cry 3 to test for stability. Use those programs like furmark or occt if you want to kill your hardware. They put such a load on the card that u will never see in real use situations. Plus when bf4 still crashes after your occt test you will still have to lower clocks lol.
post #9652 of 16490
Quote:
Originally Posted by SDhydro View Post

Play some games like bf4, tomb raider, crysis and far cry 3 to test for stability. Use those programs like furmark or occt if you want to kill your hardware. They put such a load on the card that u will never see in real use situations. Plus when bf4 still crashes after your occt test you will still have to lower clocks lol.

So what you propose is no repeatable or measurable stability test at all. Taking that view to it's obvious conclusion, you might as well just test whether the card will idle without crashing the machine. I'm not saying that something that is stable for 24 hours in OCCT and CudaMiner is necessarily stable under all conditions. There are parts of the card that those tests don't stress at all. But since those are the only tests I'm aware of that return any conclusive stability information, it is at the very least a good starting point.

Hardware is designed to last for years. If software is capable of generating enough load to damage the hardware in hours or days, then the hardware was faulty to begin with.
post #9653 of 16490
Quote:
Originally Posted by gordan View Post

So what you propose is no repeatable or measurable stability test at all. Taking that view to it's obvious conclusion, you might as well just test whether the card will idle without crashing the machine. I'm not saying that something that is stable for 24 hours in OCCT and CudaMiner is necessarily stable under all conditions. There are parts of the card that those tests don't stress at all. But since those are the only tests I'm aware of that return any conclusive stability information, it is at the very least a good starting point.

Hardware is designed to last for years. If software is capable of generating enough load to damage the hardware in hours or days, then the hardware was faulty to begin with.

Well, since I use my computers to game, run benchmarks, edit photos and videos, and do small CAD renderings, I use those same games, benching, editing, and rendering programs to test for MY systems stability, pretty common sense really, because those are what I need them to be stable with.
post #9654 of 16490
Quote:
Originally Posted by Jimhans1 View Post

Well, since I use my computers to game, run benchmarks, edit photos and videos, and do small CAD renderings, I use those same games, benching, editing, and rendering programs to test for MY systems stability, pretty common sense really, because those are what I need them to be stable with.

Except you only have a way of detecting errors that are so big and regular that they result in an outright crash or extreme artifacting. Surely, establishing a base line on running error free at least on tests that do error checking is common sense. It is better to have verifiable evidence that you can run error free on some workloads than having no verifiable evidence of running error free on any workload.
post #9655 of 16490
I'm with Gordan on this one, the fact that furmark / Kombustor etc is very intensive test is precisely why you would want to use this in a range of tests to determine GPU stability - in the same way you would use Prime 95 to validate a successful CPU overclock. While 95% of the time your system won't be loaded anywhere near these levels, it's for the rare times your CPU and/or GPU are very heavily loaded (ray tracing comes to mind for example) that you really want it to be stable.

But it depends what you're after, if you're happy to accept the odd crash then you don't need to test exhaustively, if you're using your machine to earn a living then I would want to ensure absolute stability so I would test with everything you could.
 
Mega Media PC
(21 items)
 
Media Server
(16 items)
 
CPUMotherboardGraphicsGraphics
Intel Core i7 3820 P9X79 PRO Reference EVGA GTX 780 TI Reference EVGA GTX 780 TI 
GraphicsRAMHard DriveOptical Drive
Reference EVGA GTX 780 TI Samsung  Corsair Force 3 Samsung BD-ROM drive 
CoolingCoolingCoolingCooling
XSPC Raystorm Full Copper block Phobia 150mm silver nickel reservior EK-FC780 GTX Ti - Nickel Original CSQ EK-FC780 GTX Ti - Nickel Original CSQ 
CoolingCoolingCoolingCooling
EK-FC780 GTX Ti - Nickel Original CSQ Alphacool 140.3 UT60 Nexxxos Radiator Alphacool 140.3 UT60 Nexxxos Radiator 12x Noctua NF-A14 PWM fans on rads 
CoolingCoolingOSMonitor
EK D5 Vario X-top EK D5 Vario X-top Win 7 Pro x64 Dell U2713H 
KeyboardPowerCaseMouse
Razer Lycosa ECGA SUperNova G2 1300W Caselabs TH10 Razer Lachesis 
Mouse PadAudioAudio
Razer eXactMat NI Komplete Audio NAD C320 Stereo Amplifier 
CPUMotherboardGraphicsRAM
Core i7 920 d0 Asus rampage ii gene Asus GTX 480 Corsair xms3 ddr3 2000 c9 
Hard DriveHard DriveOptical DriveCooling
OCZ Vertex 3 Samsung 2.5" notebook drive Samsung DVD-R drive Noctua 120 
CoolingCoolingCoolingOS
Noctua 120 Noctua 140 Noctua 140 Win7 x64 pro 
OSMonitorKeyboardPower
Mythbuntu 12.04.1 Panasonic 46" Plasma HDTV Apple Aluminium Keyboard  corsair hx650 
CaseMouseAudioAudio
Fractal R1 Apple Mighty Mouse Sony STR-DH800 Amp via HDMI in 5.0 speaker conf... Mordaunt Short 905 Centre, 902 front and 904 re... 
Other
Xbox wireless controller dongle 
CPUMotherboardGraphicsRAM
core 2 quad q6600 Asus P5K-E XFX 9800 GT Geil DDR2 4x1GB @ 1066 MHz  
Hard DriveHard DriveHard DriveHard Drive
Seagate Baraccuda Samsung HDD Seagate Baraccuda WD Caviar Black 
Hard DriveOptical DriveCoolingOS
Seagate Baraccuda Samsung DVD-RW Thermalright CPU cooler Mythbuntu 12.04.1 
PowerCaseOtherOther
Corsair 450W PSU Antec 300 WinTV Nova HD-S2 WinTV Nova-T PCI 
  hide details  
Reply
 
Mega Media PC
(21 items)
 
Media Server
(16 items)
 
CPUMotherboardGraphicsGraphics
Intel Core i7 3820 P9X79 PRO Reference EVGA GTX 780 TI Reference EVGA GTX 780 TI 
GraphicsRAMHard DriveOptical Drive
Reference EVGA GTX 780 TI Samsung  Corsair Force 3 Samsung BD-ROM drive 
CoolingCoolingCoolingCooling
XSPC Raystorm Full Copper block Phobia 150mm silver nickel reservior EK-FC780 GTX Ti - Nickel Original CSQ EK-FC780 GTX Ti - Nickel Original CSQ 
CoolingCoolingCoolingCooling
EK-FC780 GTX Ti - Nickel Original CSQ Alphacool 140.3 UT60 Nexxxos Radiator Alphacool 140.3 UT60 Nexxxos Radiator 12x Noctua NF-A14 PWM fans on rads 
CoolingCoolingOSMonitor
EK D5 Vario X-top EK D5 Vario X-top Win 7 Pro x64 Dell U2713H 
KeyboardPowerCaseMouse
Razer Lycosa ECGA SUperNova G2 1300W Caselabs TH10 Razer Lachesis 
Mouse PadAudioAudio
Razer eXactMat NI Komplete Audio NAD C320 Stereo Amplifier 
CPUMotherboardGraphicsRAM
Core i7 920 d0 Asus rampage ii gene Asus GTX 480 Corsair xms3 ddr3 2000 c9 
Hard DriveHard DriveOptical DriveCooling
OCZ Vertex 3 Samsung 2.5" notebook drive Samsung DVD-R drive Noctua 120 
CoolingCoolingCoolingOS
Noctua 120 Noctua 140 Noctua 140 Win7 x64 pro 
OSMonitorKeyboardPower
Mythbuntu 12.04.1 Panasonic 46" Plasma HDTV Apple Aluminium Keyboard  corsair hx650 
CaseMouseAudioAudio
Fractal R1 Apple Mighty Mouse Sony STR-DH800 Amp via HDMI in 5.0 speaker conf... Mordaunt Short 905 Centre, 902 front and 904 re... 
Other
Xbox wireless controller dongle 
CPUMotherboardGraphicsRAM
core 2 quad q6600 Asus P5K-E XFX 9800 GT Geil DDR2 4x1GB @ 1066 MHz  
Hard DriveHard DriveHard DriveHard Drive
Seagate Baraccuda Samsung HDD Seagate Baraccuda WD Caviar Black 
Hard DriveOptical DriveCoolingOS
Seagate Baraccuda Samsung DVD-RW Thermalright CPU cooler Mythbuntu 12.04.1 
PowerCaseOtherOther
Corsair 450W PSU Antec 300 WinTV Nova HD-S2 WinTV Nova-T PCI 
  hide details  
Reply
post #9656 of 16490
Quote:
Originally Posted by pharcycle View Post

I'm with Gordan on this one, the fact that furmark / Kombustor etc is very intensive test is precisely why you would want to use this in a range of tests to determine GPU stability - in the same way you would use Prime 95 to validate a successful CPU overclock. While 95% of the time your system won't be loaded anywhere near these levels, it's for the rare times your CPU and/or GPU are very heavily loaded (ray tracing comes to mind for example) that you really want it to be stable.

But it depends what you're after, if you're happy to accept the odd crash then you don't need to test exhaustively, if you're using your machine to earn a living then I would want to ensure absolute stability so I would test with everything you could.

The problem is that Furmark only tests GPU really, not GPU memory, and it doesn't use GPU memory like a game would. So all you test with Furmark is the GPU clock and maximum power draw...not GPU memory. It's good if you want to test GPU coolers, otherwise, I have to question it's usefulness, since that spinning furry donut rendering has been around for 10 years..with ATiTool before OCCT/Furmark.

It's unfortunate we don't really have a good modern GPU testing tool.
Edited by cadaveca - 4/6/14 at 10:50am
post #9657 of 16490
Quote:
Originally Posted by cadaveca View Post

The problem is that Furmark only tests GPU really, not GPU memory, and it doesn't use GPU memory like a game would. So all you test with Furmark is the GPU clock and maximum power draw...not GPU memory. It's good if you want to test GPU coolers, otherwise, I have to question it's usefulness, since that spinning furry donut rendering has been around for 10 years..with ATiTool before OCCT/Furmark.

It's unfortunate we don't really have a good modern GPU testing tool.

Try this: http://www.catzilla.com/ its interesting! wink.gif
S.S.
(14 items)
 
  
CPUMotherboardGraphicsRAM
I5-3570K@5.0Ghz@1,4v ASUS MAXIMUS V GENE MSI 1080Ti Gaming X 16GB GSKILL RIPJAW X DDR3 1866MHZ (9-10-9-24) (... 
Hard DriveHard DriveCoolingOS
OCZ Agility 3 120GB + Toshiba A100 250GB Western Digital 2TB  Homemade WC (everything on water) ;) W10x64 
MonitorMonitorMonitorPower
ASUS VG278HE ASUS VG278HE ASUS VG278HE Silverstone Strider 1200W 
CaseAudio
Corsair 500R THX 500W 5.1 
  hide details  
Reply
S.S.
(14 items)
 
  
CPUMotherboardGraphicsRAM
I5-3570K@5.0Ghz@1,4v ASUS MAXIMUS V GENE MSI 1080Ti Gaming X 16GB GSKILL RIPJAW X DDR3 1866MHZ (9-10-9-24) (... 
Hard DriveHard DriveCoolingOS
OCZ Agility 3 120GB + Toshiba A100 250GB Western Digital 2TB  Homemade WC (everything on water) ;) W10x64 
MonitorMonitorMonitorPower
ASUS VG278HE ASUS VG278HE ASUS VG278HE Silverstone Strider 1200W 
CaseAudio
Corsair 500R THX 500W 5.1 
  hide details  
Reply
post #9658 of 16490
Quote:
Originally Posted by cadaveca View Post

The problem is that Furmark only tests GPU really, not GPU memory, and it doesn't use GPU memory like a game would. So all you test with Furmark is the GPU clock and maximum power draw...not GPU memory. It's good if you want to test GPU coolers, otherwise, I have to question it's usefulness, since that spinning furry donut rendering has been around for 10 years..with ATiTool before OCCT/Furmark.

It's unfortunate we don't really have a good modern GPU testing tool.

There are at least two good GPU testers that test both GPU and VRAM.

OCCT let's you select a 2GB test data size which is enough for a decent amount of VRAM I/O workout.

CudaMiner requires a lot of memory I/O to the point where it has an option to use texture memory features for caching.

The fact is that for proper stability testing you need to test both the processing unit and RAM I/O both separately and combined to ensure stability.
post #9659 of 16490
Quote:
Originally Posted by OccamRazor View Post

Try this: http://www.catzilla.com/ its interesting! wink.gif

I have been using CatZilla since it came out. My wife noticed the storyline in the bench before I did. tongue.gif
Quote:
Originally Posted by gordan View Post

The fact is that for proper stability testing you need to test both the processing unit and RAM I/O both separately and combined to ensure stability.

Of course. And no single app does this truly effectively.
post #9660 of 16490
Quote:
Originally Posted by gordan View Post

There are at least two good GPU testers that test both GPU and VRAM.

OCCT let's you select a 2GB test data size which is enough for a decent amount of VRAM I/O workout.

CudaMiner requires a lot of memory I/O to the point where it has an option to use texture memory features for caching.

The fact is that for proper stability testing you need to test both the processing unit and RAM I/O both separately and combined to ensure stability.

In your opinion.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: NVIDIA
Overclock.net › Forums › Graphics Cards › NVIDIA › [Official] NVIDIA GTX 780 Ti Owner's Club