GTX660Ti crashes during default clocking

Toggle sidebar Toggle sidebar
E

ekylam

Honorable
Jun 10, 2013
7
0
10,510
Let me describe my system:

* CPU i5-3570
* RAM Corsair Platinum Dominator 2 x 4GB
* ADATA SX900 128GB
* WD 1TB Black Series
* Asus GTX 660 Ti DirectCUII 2GB
* Asus P8Z77-V LX2
* Cooler Master Siliencio 550
* Windows 7 HP 64-bit
* Cooler Master GX 550W

This is a newly self built system. It's Windows 7 64-bit and I've updated the latest Nvidia drivers 320.18. It works perfectly fine other than gaming (SimCity 5) and benchmarking (3DMark11 and Valley Benchmark by Unigine).

In the GPU Tweak, the default setting of GPU Boost clockspeed is 980 and the memory clockspeed is 6008. I lowered to 830 and 5208 respectively then I can pass through Valley Benchmark and occasionally 3DMark11. SimCity 5, well, a lot stable than the default clockspeed.

I tried to reseat the graphic card to another PC with i3-2100, MB Asus P8H67-M, 8GB RAM and Cooler Master GX 400W. I ran 3DMark11 with the default setting and it was no problem. SimCity 5 ran OK without crash.

Actually, it crashes to a system reboot. Then I look at Event Viewer it said there is a bugcheck with 0x00000116. I searched thru Google and it said it's related to the display drivers, but with many different causes they encountered.

I'm wondering if my PSU has enough power such that a higher speedclock would cause a reboot? I don't really have a clue right now. Pls help!!! Tks.
 
Sort by date Sort by votes
E

ekylam

Honorable
Jun 10, 2013
7
0
10,510


However, when I reseat the card into another PC, as described above. I didn't have such problem. What does that mean?
 
Upvote 0 Downvote
envy14tpe

envy14tpe

Champion
Mar 3, 2012
9,574
14
56,665


If the GPU works on a different system then the issue isn't the GPU. It could be a problem with the motherboard. Is this a new mobo? Have you tried a different GPU on it?
 
Upvote 0 Downvote
E

ekylam

Honorable
Jun 10, 2013
7
0
10,510


Yes it is a new mobo. Unfortunately I don't have another GPU to test out. What I noticed is that if I lower the GPU/memory clockspeed, the game playing or benchmarking would be more stable. Say in 3DMark11, if I stay at default clockspeed, it will run around 5mins to crash. If I move to Gaming profile, it will crash in 2mins. And I would deduce, because i3-2100 has a lower clockspeed than i5-3570 so the problem may not surface up. At least there is a trend, i.e. 15% lower than default clockspeed the game will be a lot smoother. But then I don't know what would be the next step forward...
 
Upvote 0 Downvote
R

razvan_mz

Distinguished
Jun 25, 2011
421
0
18,960


did you do a clean install of windows after you changed the MB? if not, then do it, because you don't really know what driver is in conflict.
 
Upvote 0 Downvote
E

ekylam

Honorable
Jun 10, 2013
7
0
10,510


Yes. I did a clean install and carefully installed all the MS update plus the Nvidia drivers. At first I suspected it could be the display drivers. But when I moved the GPU to my another PC, whatever versions it works out in 3DMark11 and SimCity 5.

One more thing, today when I lowered the clock and ran another benchmark software - Valley by Unigine, the GPU temp is around 70C. It didn't crash. So, temp should not be the factor.
 
Upvote 0 Downvote
Novuake

Novuake

Champion
Aug 27, 2011
14,308
0
56,960


Have you tried what I sugested yet? Increase the power slider to max. Its n glitch with "ti" cards that I have heard of before.
 
Upvote 0 Downvote
E

ekylam

Honorable
Jun 10, 2013
7
0
10,510


Nouvake, I tried. But still it crashed. BTW, I know which drivers are causing the 0x00000116 bugcheck, they are dxgkrnl.sys, dxgmms1.sys and nvlddmkm.sys. But why it worked completely fine with my another i3-2100 PC? What should be said is, the symptom is definitely about the display card drivers with DirectX. But what caused this symptom?

* Power? Some said it's plenty and enough?
* Temperature? It could crash it 2mins when starting 3DMark11. Temp should be risen to the threshold so quick.
* Driver compatibility?
* What else?

It would be great if I had another GPU to test but I don't have such luxury...
 
Upvote 0 Downvote
E

ekylam

Honorable
Jun 10, 2013
7
0
10,510
Guys, finally finally I got it resolved. The problem was due to a cable! Tks to the hints given in another post about loosing cable. I checked against mine and found that although the two 6-pin cables are tightly connected and not worn out, they are from the same vein. I began to think if it was to cause the system reboot - not enough voltage. Then I separated the two 6-pin into different power vein. And the 3DMark and Unigine Valley benchmark tests were passed.

Cuz in my another PC I connected to two separated 6-pin and this was one of the differences.

Thanks all for your kind suggestions. I think my lesson learnt today is check the basics.
 
Upvote 0 Downvote
envy14tpe

envy14tpe

Champion
Mar 3, 2012
9,574
14
56,665


Good news. Sometimes a fix can be something simple, glad this was.
 
Upvote 0 Downvote
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom