Overclock.net banner
1 - 12 of 12 Posts

·
Registered
Joined
·
1,829 Posts
Discussion Starter · #1 ·
I am starting to experience random PC shutdowns when running my system full tilt benchmarking. It has the hallmarks of a PSU issue but I am running an Enermax 1500w max revo.

Essentially when running 3dmark 11 or alike with both my cards highly overclocked, and my CPU @ 4.8ghz I will get random shutdown and restarts.

I have prime tested this CPU overclock and its fully stable, I have run this exact config with a single card highly overclocked and everything is fine, but when going SLI I have this problem. I have had both cards running fine with slightly lower clocks but I am yet to find instability and this problem is not being caused by the cards becoming unstable (I believe anyway).

What do you guys suggest? Event Viewer just shows Kernel power issue (generic) and as far as I know, when I reach instability with these cards I just get a recoverable driver crash.

Maybe I have both my GPU's being fed from the same PSU rail?
 

·
Premium Member
Joined
·
6,238 Posts
I just can see any way that the PSU can be the problem unless its faulty. I'm guessing you have already checked the temps on the cards and ruled that out. You said the system runs fine with a slightly milder OC and with the current OC if you only use one card.

Have you tried running both cards individually or just the top card? I could be that the second card is the one that gets unstable and black screens.

When you say "shutdown" is it an actually complete off as in no fans spinning and no restart? Unstable GPU can cause restarts and black screens.
 

·
Registered
Joined
·
1,829 Posts
Discussion Starter · #4 ·
It fully shuts off for a second and the restarts to post. I just reset the bios to 100% stock settings and left the gpus to the same clocks and it ran through 3dmark11 fine. Looks as though jts either something to do with my cpu overclock or my psu.
 

·
Registered
Joined
·
1,829 Posts
Discussion Starter · #6 ·
Quote:
Originally Posted by EVGA-JacobF View Post

Usually this type of issue is caused by one of two things:

1. Overheating CPU or other component.
2. Faulty or overloaded PSU
Everything is watercooled and temps are easily in check.

I dont believe the PSU to be faulty but I guess I cant rule it out..
frown.gif
 

·
Premium Member
Joined
·
6,238 Posts
Quote:
Originally Posted by trippinonprozac View Post

Everything is watercooled and temps are easily in check.

I dont believe the PSU to be faulty but I guess I cant rule it out..
frown.gif
Well I guess its also possible that the PSU overheats as unlikely as it may be. Have you put your hand on it when its under load? Perhaps there is lots of dust build up in it.
 

·
Registered
Joined
·
9 Posts
I think this is caused by multiple reasons

the most suspicious thing is the overall-clocks , the gpu main clock or the vram clock or the cpu clock

but cpu is totally ok because we often test cpu by linx0.64 or p95-27.7-b2 and all passed

ram should be down-clock to normal level

the other is the GPU self's reason

do you think this is my 1188 to 1254mhz result ?

but my 1176 boost 1241 is the same result

or I should down to the 1110 stock clock ?

I think this software has problem , new 3d11 is ok , 1254 is ok
but in Maxpayne3 1254 is ok too with 8msaa all ultra

I have 2 680 gpus , the slavery card is slow , stable @ 1241 , I use big psu it can up to 1254 in any game and any benchmark software except the new Valley

What's the problem?

I uninstall this software and everything is ok , I'm satisfied with the 1254 or 1241 , I think this is stable

in the Valleys benchmark , 1241 is unstable and this is god damn unbelievable !!!

you ask my temperature? main card is classified 40% fan speed and slavery card vapor-x 65% fan speed , 70c +/- 2 , with the powerful additional 12025/12015 cooling fan .

thank you , my English is not good please forgive me

 

·
Registered
Joined
·
1,829 Posts
Discussion Starter · #9 ·
I ran p95 blend last night overnight for 8 hours 20 mins successfully and the hottest core reached 75c. This all but rules out an unstable CPU overclock.

I game at these clocks no issues at all, and the weird thing that I have noticed is that the shutdowns actually seem to be straight after a section of the benchmark runs (ie after the load returns to normal).

I cant put my finger on what may be wrong??
 

·
Registered
Joined
·
13 Posts
Hello

sorry to dig that old thread but I have the same issue. How did you manage to solve it?
I have two gtx 780 classified in sli, a Cooler Master v850 psu, core i5 2500k 4.5ghz., 1 ssd, 2 sata drives, 8go 12800.
Running at stock clock i have no problem at all, valley, 3d mark11 run fine. When I overclock my gpus (+120 core, +498 on mem) the pc shuts down like if I pulled the power cable but only during Valley (and yersterday after 1hour on bf4)
The strange thing is that it happens always at the same moment during Valley. The 18 steps of the test complete but when the result is about to appear the pc brutally shuts down.
I suspected the PSU, is 850w not enough ? I doubt it.
Temps are okay (gpu never goes over 80°C, cpu always below 70°C)

Any idea?
thanks!
 

·
Registered
Joined
·
323 Posts
Quote:
Originally Posted by Draugr View Post

Hello

sorry to dig that old thread but I have the same issue. How did you manage to solve it?
I have two gtx 780 classified in sli, a Cooler Master v850 psu, core i5 2500k 4.5ghz., 1 ssd, 2 sata drives, 8go 12800.
Running at stock clock i have no problem at all, valley, 3d mark11 run fine. When I overclock my gpus (+120 core, +498 on mem) the pc shuts down like if I pulled the power cable but only during Valley (and yersterday after 1hour on bf4)
The strange thing is that it happens always at the same moment during Valley. The 18 steps of the test complete but when the result is about to appear the pc brutally shuts down.
I suspected the PSU, is 850w not enough ? I doubt it.
Temps are okay (gpu never goes over 80°C, cpu always below 70°C)

Any idea?
thanks!
Well I did a Quick Run of eXtreme Power Supply Calculator minimum for you with that OC is 975w.. I would Highly suspect your PSU at this point.. you will need 140w just for the CPU at 100% load at that OC.. check for yourself:

http://extreme.outervision.com/PSUEngine

bit I would say running Full out 100% Load on all you will need a minimum of 1000w.. Also make sure you have the Amps needed for both those video cards..
 

·
Registered
Joined
·
13 Posts
thanks for the answer. Indeed something is drawing too much power in my configuration. By lowering the i5 oc to 4.3 the shutdown during valley stopped. I'll probably get a new more power ful psu soon or just settle with that...
 
1 - 12 of 12 Posts
Top