Gigabyte GTX 660TI Random Reboots?

Shenahal

Honorable
Sep 20, 2012
7
0
10,510
Hello,

I recently purchased a Gigabyte GTX 660 TI and noticed that my computer now randomly reboots with no hint of a issue before hand (no freezing, stuttering, nothing).

The reason I do not think it is a temperature issue is because it has never rebooted while in a game, it happens usually when I am browsing or if i am just on my desktop doing nothing.

Motherboard: asus p8z68-v pro
RAM: 4GB x 2 ( corsair?)
CPU: Intel 2600k
Storage: 1 SSD + 2 HDD
PSU: OCZ 700W

This all started when I first installed the video card, my only concern is how random the reboots are.

For example: It will not reboot for 2 days, but than all of a sudden it will reboot 3 times in a row every time i get to the desktop then stop for a full day or more. It does not happen while i am doing anything specific it just happens.

It does not look like a defective video card because everything runs so smooth and great! My current guess is it is a fairly new card and the drivers are buggy?

Any help on this very frustrating issue would be much appreciated.


EDIT: Is it possible that the video card is pressed to close to a mother board heat since which causes random restarts?
 
I updated the BIOS the other night with the latest from Asus through EZ flash, it fixed my double boot issue but not the video card issue. However Asus has a whole bunch of other updates for my mother board, should i be considering any other updates? I updated with nvidia's latest drivers including the one released a couple weeks ago. When i initially installed the video card i made sure to do a clean driver instal.

I haven't tried the following:

--new PSU
--different PCI slot
-additional motherboard updates (apart from BIOS)
-clean OS install

Would any of these be worth trying?
 
No, there is no issue between adding a new GPU and the memory. They are separate and distinct from one another. It is possible that you bumped one of the DIMMs and it isn't seated properly now. Also, things do become defective and flaky memory can cause the sorts of random shutdowns you describe.

You can get memtest from http://www.memtest.org . Worth checking out and it is free.
 
Sounds good, I will check it out first thing when I get home.


I also forgot to mention Microsoft event viewer shows all the shutdowns as Kernal-Power. Not sure if that means anything, event viewer has provided me with no help in the past so I wont hold my breath.
 
I have the same problem but different mobo: Asus p8p67 (b2 revision) latest bios for it (2xxx) can not flash newer it is only for B3 chipset and i just bought MSI GTX660
PC was 100% stable with GTX460
Now with GTX660 i have restarts while PC is at idle (internet surfing or totaly idle) it happens once a day or so

NO BSOD just kernel power so can`t know more 🙁

It looks like power cut PC shut down for 2sec and wakeup from it, thats all ....
 


Hey reb0rn,

Just a heads up here is what I tried so far that has not worked:

--Updated BIOS
--Changed PCI Slots
--Reset my CMOS battery
--There was one missing screw on my MOBO so I fixed that
--Flushed the PSU
--Updated Video card driver
-------------------------------------------------


Here is what seems to be working so far:

You know the two power cables that connect to the video card? Swap them, so far I have been almost 48 hours without a restart since I did that, this is the longest I lasted lol...on my PSU they are virtually the same cable except one has a modular connection and the other goes directly into my PSU, swapping them for some reason works so far. I will keep you posted. My last resort option is to try an new PSU, even though I have a 750w!
 
Yeah my plan was also to try to swap PSU GPU cable, GTX460 had 2x and GTX660 has just one connector and power drown is almost the same for both (GTX660 uses 10% less)

will do asap and report in a few days

 
It is possible that there might still be a bug in the drivers even though they are recent then again I can't rule out there being a stability issue. You might have to wait until the next release or you could consider contacting the company that branded the card to see if there is anything else that can be done.
 
I duno if software/driver can produce kernel power event id: 41 with no additional log (no blue screen, ect)
I even disabled automatic restart on BSOD and no help, it looks like power cut when it happens

Also i forgot that i have two monitors (hdmi and over dvi/rgb) so its not so common setup
 



Yea I am also running two monitors and the way you describe it was exactly the same for me, the power just cuts. No warning, sometimes it will do it 4 times in a row and stop for a good 20 hours..anyways so far it has not happened again after i swapped my two GPU cables around, crossing my fingers.
 
In the mean time can you report back any temp readings from gpu-z and listen in for any odd noises when gaming and on desktop. Also see if gpu-z and or msi afterburner reports the volts during idle and load. I doubt that it is the power supplies in all these machines being the problem. Could be the boards if the cards are going over the 75w limit for the slot. Beyond that the only thing I can think of besides software is a hardware issue. Maybe it could be a dip in the voltage on the card resulting in the crash.
 
After 30h again had restart while i was sleeping 🙁
Now testing with one monitor only

Some ppl sad maybe PSU is incompatible, my have four rails of 18A on 12v, but still duno how PSU could be a problem if GPU use just 10-15w at idle

@nforce4max the restart only happen at desktop and the card is at ~300Mhz with 0.88v for the GPU for me even the monitor was on sleep few last time it happened
 
Think i found the problematic setup, 72h and so far no restarts, but analog monitor over DVI analog adapter is unplaged (Viewsonic VA2216w @1680x1050)

atm only connected 32inc LCD over HDMI (1920x1080 resolution)

So:
1. driver problem with low GPU voltage on idle to supply hdmi and analog ramdac?
2. HW problem of integrated ramdac not liking my Viewsonic....?
3. defect card?

Also the voltages on GPU at idle are the same for 1 or 2 monitor used 324Mhz for gpu and 162Mhz for memory @0.887v at idle, maybe Nvidia need to tweak it a bit for 2 or 3 monitors?
 
Same problem exactly here with a the Gigabyte GTX 660 (not TI) and a Gigabyte MB with the random restarts at idle. PS is 850W and no issues with the GTX 470 used previously. I am leaning toward faulty driver or design. My setup is single monitor.
 


Do you have two monitor setup?
I fixed my restarts with disabling analog monitor (the one over DVI adapter..) i think the ramdack is fault which is used only with analog monitors
 


Bad design then, Ramdac design is more or less standardized for each major generation when it comes to Nvidia but there have been changes. This is another reason that I won't have anything to do with Kepler.
 
Ok this is becomes pointless 🙁
full 3 days on desktop idle/surfing no problems with just one hdmi monitor

Then i test some PS2 games over pcsx2 emulator and bang restart while gaming 🙁
Then i check gpu log the card was at 20% usage with fps limiter from pcsx2 and frequency was 320-380Mhz with 0.887v and now i am almost 100% sure that card on idle has no enough power or???????

I need a FREE kepler bios editor this is getting pointless!
 


only P0 state can change voltage... so far i am forcing P5 state which is same clock as P8 but a bit more voltage. first i need to be sure the cause of the restart

But it is hard because restart can happen after 48h
 



This is Just a single monitor, HP w2408h
 
I updated the driver to 306.38 also got the new bios from MSI will see had no restart for a week but duno if new driver helped or the bios....

bios is: 80.06.10.00.5e
 

TRENDING THREADS