GTX 780 power supply issue

NightWolfix

Honorable
Dec 10, 2012
16
0
10,510
Hey guys,

I am testing my new GTX780 and the pc restarts whenever I run heavy duty games (3dmark etc). Looks like the PSU is not enough but it is as follows:
https://www.dropbox.com/s/rvm9250m3nwsnyg/20130618_194648.jpg

It has more than 42A on the +12 line, and it should me more than enough to run the card.

I have a core i7 D0 running at default clock and no fancy overclock around.

Do you think that it is time to get a new PSU?
 

DjDafiDak

Honorable
Aug 31, 2012
733
0
11,360
Iam runing a gtx 780 Overclocked with a 620Watt antec.

just comes to show you how PSU is all about the brand, thermaltake is decent, but i would switch to another brand.

maybe its something else though, make sure u have latest drivers + not overheating( even though i doubt that, since its brand new.)
 

NightWolfix

Honorable
Dec 10, 2012
16
0
10,510
Its very strange, the reload happens only when I run 3dmark, the card starts making a screetching sound and the system restarts just after few seconds. I tried the 15 mins furmark gpu burn in test and card is doing fine (no sound).

I tried to stop the card during 3dmark just before a potential restart and the temperatures do not seem bad:
https://www.dropbox.com/s/353zbf234lv3s86/Capture.PNG

A little bit puzzled.

 

NightWolfix

Honorable
Dec 10, 2012
16
0
10,510
I tried to put my old 5850 in and the noise is still there (only during 3dmark) although less intense than with the 780. The system doesn't crash with the 5850. I put the 780 back in, played some crysis 3 and the system seems ok, probably because crysis 3 is not demanding enough to trigger the issue. I cannot think about anything else than the poor/faulty thermal take.
 
well, though it still sounds like overheating to me... it wouldn't be unheard of for a thermaltake psu to be faulty. they're lower middle of the road psus... and tend to have a spotty dependability record. About the only middle of the road PSUs i'd advise someone to use would be a Corsair or Antec...
 

NightWolfix

Honorable
Dec 10, 2012
16
0
10,510
I tried to test the card with the case completely open, the psu outside of the case so there was plenty of space for the card to breathe. I run the entire 15 minutes gpu burn in test and it went ok. Temperatures were around 75 or so and no whining sound . I switched off the pc, let it cool it down, restarted and run the test while it was still relatively cool. After 30 seconds or so of loud whine it just shuts down and restarts. If this happens with the case completely open I cannot imagine what could happen with it closed.
 


so then you have a massive airflow/temp issue inside the case.

what case are you rolling with? the airflow issues might be fixable... you know i gotta say you're like the 3rd or 4th person i've chatted with in the last week with a 700 series gpu and a massive overheating issue. It makes me wonder if something is going on with that series.

one guy had a GTX 580 space heater he replaced with a 780, and the 780 is overheating badly in the same case his 580 worked flawlessly. Sorta makes me wonder.
 

NightWolfix

Honorable
Dec 10, 2012
16
0
10,510


I got a fractal design r4. The card does send out a huge amount of heat, but since the temp is around 72-75 it looks within the range of this test (I have the OC version):
http://www.techpowerup.com/reviews/Gigabyte/GeForce_GTX_780_OC/30.html

Also it should do that both with 3dmark and furmark burn in. I tried also to stop 3d mark few seconds before the expected restart (which usually happens after 40 seconds of 3dmark) and the temp were lower than or equal to the furmark test. No whine issues/restarts with games, just with the damn 3dmark. I have not even tried the advanced test which stresses the cpu even more.

Any tips to rule out the heating problem?

Thanks indeed guys, u are helping me out a lot here.

 


Take the side of the case off, turn the case on it's side. take a room fan, and position it at the bottom of the case, direct the fan directly at your gpu, turn it on high. then turn on your pc... and see if the "crashes" stop.

if they do, then your problem is temps not your psu.
 


great news!

Sorry i wasn't any help, and i'm glad it wasn't temps. I sorta gravitated that way cause i kept running into people with 7xx series nvidia gpus overheating this past week.