Display corruption cause?

[artyboy]

Played on the 2nd card for an hour + at max temps of 65C on CPU and GPU at 76C w/o any crashes. So it seems that the cards are fine. I only played COD-BO2 for this test as no point playing Crysis 3 as its known to always crash even before i upgraded my comp (i.e. using previous CPU/GPU/RAM. I did this test using the top PCI-E slot. Seems like the crash only occurs (with all games rather than just Crysis 3) when im playing games in SLI mode. I find it puzzling since my 1,200 watts PSU should be up to the task of powering 2 cards(?). Heat dont seem to be an issue, unless 75C is too close to Nvidia card's 79C (80C) temperature target? I don't know how that works out. Is there any "Dual GPU" which i have to enable on the mobo or something?

Btw the temperature difference of CPU:58/GPU:67C and CPU:65/GPU:76C between the 2 tests is because the first test was done with the PC case's side panel open for the first test and closed on the second test (not deliberate, i just forgot).

 

[royfrosty]

Revert back older drivers see how?

 

[Hailgod]

1200watts can power 4x r9 290 + overclocked cpu

 

[artyboy]

take a look at the difference in temperatures once i put the 2 cards back in SLI mode(!)



Last resort is to use another driver i guess

 

[artyboy]

and then its reversed...

 

[Hailgod]

Did you try getting a new SLI bridge?

 

[artyboy]

Did you try getting a new SLI bridge?

In fact im using the new ROG one, the old one from my MSI mobo i kept it as spare. But i did change the connection of the SLI tabs on the graphics cards - from the middle one to the one on the extreme left.

 

[artyboy]

The picture gets clearer, checked MSI Afterburner - GPU1 is at 75% and GPU2 is at 13% power usage. Seems like COD-BO2 is not optimized well for SLI as the 1st GPU is handling most of the work. No wonder using one or 2 GPUs the difference is minimal. Perhaps using 2 GPUs in SLI is generating too much heat... but temps seemed normal earlier (perhaps dependent on game). Damn demoralized.......

 

[royfrosty]

In fact im using the new ROG one, the old one from my MSI mobo i kept it as spare. But i did change the connection of the SLI tabs on the graphics cards - from the middle one to the one on the extreme left.

Try the msi one instead.

 

[artyboy]

Try the msi one instead.

K will try tomorrow. I changed the PC case's 2 side fans' orientation from blowing at the 2 cards to drawing the heat from the 2 cards outwards towards the grilled side panel.

 

[royfrosty]

K will try tomorrow. I changed the PC case's 2 side fans' orientation from blowing at the 2 cards to drawing the heat from the 2 cards outwards towards the grilled side panel.

Not heat issues de.

Its operating temp can operate up to 90 for Nvidia. 99 should be a black out for either camps.

 

[artyboy]

Not heat issues de.

Its operating temp can operate up to 90 for Nvidia. 99 should be a black out for either camps.

That's why so peck chek bro. Run on one card no crash but when both cards in SLI it crash. I don't know what's the cause. I already ensured all connections are snug, changed connections to the card's SLI connection tabs & reversed the side fans' airflow direction orientation. Sians...

 

[Hailgod]

That's why so peck chek bro. Run on one card no crash but when both cards in SLI it crash. I don't know what's the cause. I already ensured all connections are snug, changed connections to the card's SLI connection tabs & reversed the side fans' airflow direction orientation. Sians...

You can talk to GS and maybe bring down to his shop test and find out the problem? He should know more than us

 

[spyboy48]

That's why so peck chek bro. Run on one card no crash but when both cards in SLI it crash. I don't know what's the cause. I already ensured all connections are snug, changed connections to the card's SLI connection tabs & reversed the side fans' airflow direction orientation. Sians...

does it crashed suddenly when on load and in black screen?
if yes, check your psu!

 

[artyboy]

Who's GS? The crash is like when i play game halfway, it freezes and a BSOD appears saying "hardware fault detected, windows stopped to prevent damage to computer". The it will do the "crash dump" countdown and reboots the PC until reach the "do you want to bootup in safe mode" black screen, start up Windows and have a temporary hang where you see the arrow-cursor in the centre. I then Ctrl-Alt-Del to boot up to windows desktop.

Instead of playing a game for an hour just to test for stability, i left COD-BO2 running (camp inside a sentry box with a firefight battling outside ), off the monitor and speakers and left it to run overnite. When i wokeup this morning, the game didnt crash. I don't know if that is the same as "playing the game" for (more than) an hour . I HATE this game!! Stupid long-winded cut-scene videos and unnecessary "protect the base/car chase/manipulate the console" sequences. Who cares about the stupid story just give me the damn gun!! Btw this is the temperatures of the overnite run, don't know if it can be considered as "playing the game for (more than) an hour" testing like the first two tests

 

[Hailgod]

http://forums.hardwarezone.com.sg/h...corner-all-diy-hardwarezoner-^-^-3581022.html

 

[artyboy]

Ahh thanks bro. Left PC on with COD-BO2 running since last nite - with the 2 cards running in SLI. Woke up this morning and game was still running, no crash. Then played for 1.5 hrs this morning w/o crash. I don't think temperature is an issue as I see only a max of 76C on the GPU and 61C on the CPU. This morning the GPUs were in the 58-67C range because the game is not that demanding. Only changes I made was using a different SLI connection tab on the cards and changing the side panel's 2 fan's airflow direction to draw rather than blow air from/towards the 2 cards.

 

[Hailgod]

btw, what voltage is your 4770k running on? Have you done stress tests on it?

 

[artyboy]

Voltage is on auto as I used Asus OC software to OC. Yes did Prime95 stress test after saving the OC profile. If CPU temperatures are normal shouldn't be an issue right?

 

[Hailgod]

Voltage is on auto as I used Asus OC software to OC. Yes did Prime95 stress test after saving the OC profile. If CPU temperatures are normal shouldn't be an issue right?

ok. I thought maybe overclock not stable and caused cpu to shut down. It happened to me when i undervolted too much and crashed while playing dota 2