Kernel - Power event ID 41 (63) System less then 12 months old. Hardware Failure?

justone

Reputable
May 20, 2015
1
0
4,510
Kernel - Power event ID 41 (63) System less then 12 months old. Hardware Failure?
Bugcheckcode 0

HELLO :))))
Hope another pair of eyes looking over this may help me be sure Ive not missed anything, and double check my conclusion that I have a hardware fault that I will talk to my shop about replacing.

I dont have any other PSU or GPU items to switch with, PC shop is 4hrs away ( if it was local I wold have taken the parts in already ) that i bought parts from, I'd like to find the problem part and post it for warranty. The less items I post the cheaper it will cost, so ive been attempting to find the troubled part so I dont have to ship a full box back for testing / warranty !



System
All parts were bought and delivered brand new (besides 1 hdd) and I built in 11/2014.
Has run without issue till recently 5/2015.
So ive been playing the same games / applications, trouble free, on high settings until now ( a good 6-7 months ) without issue.


Component Estimated Wattage
Intel Core i5-4690K 3.5GHz Quad-Core Processor 11W - 88W
Noctua NH-D15 82.5 CFM CPU Cooler 5W - 10W
Gigabyte GA-Z97-D3H ATX LGA1150 Motherboard 17W - 70W
G.Skill Ares Series 16GB (4 x 4GB) DDR3-2133 Memory 20W - 20W
http://www.gskill.com/en/product/f3-2133c9q-16gab
Seagate Barracuda 2TB 3.5" 7200RPM Internal Hard Drive 4W - 20W
Old hhd 3.5" 7200RPM Internal Hard Drive 4W - 20W
Intel 530 Series 240GB 2.5" Solid State Drive 1W - 5W
Gigabyte GeForce GTX 970 4GB WINDFORCE Video Card 36W - 145W
http://www.gigabyte.com/products/product-page.aspx?pid=5209&kw=GV-N970G1%20GAMING-4GD#ov
Noctua NF-F12 PWM 55.0 CFM 120mm Fan 1W - 5W
Noctua NF-F12 PWM 55.0 CFM 120mm Fan 1W - 5W
Noctua NF-F12 PWM 55.0 CFM 120mm Fan 1W - 5W

Total: 101W - 393W
PSU Seasonic semi Modular G 550W Gold Plus
http://www.seasonicusa.com/G-series-450-550-650.htm
added estimated power usage, thought at one stage it may be a lack of power but I think it should be plenty, if everything is working correctly. ( remembering it has worked fine for 7+ months and the 970 suggests 550W )

During game play, seems to be at random, sometimes 1 minute other times 45 minutes of game play before an instant system restart occurs, showing event 41 in the windows event viewer.

No blue screens, no errors, just an instant black screen, and system begins restarting.

Over the last 4+ weeks trying to find the problem, originally thinking it was a software issue, I have done the following;
updated windows
checked drivers
Bios restore defaults
Flash update bios

Problem still exists,
Reinstalled windows
updated windows
re installed drivers / software

Problem still exists, now believe it is a hardware issue,
checked psu connections
changed power point / power board

pulled ram and tested sticks individually 1 at a time out of 4,
- gaming - still crashed randomly on each
- memtest ( 4 passes 4-5hrs ) no errors on each



Temps are all pretty low, I am sure heat isn't an issue.
GPU 25'C idle
CPU average over all cores is 24'C idle
Nothing has been manually over clocked,
CPU does auto "boost" to 3.9 Mhz
GPU is showing 1178/1329 BASE/BOOST and 7010 memory clock via gigabyte OC GURU II
^ I have just noticed GPU showing 1404 clock during game play, this is above its advertised 1329?

Ram is set to auto XMP profile in bios
in bios I see my ram sitting on 1.668V using XMP profile. ( ram is 1.65V )
http://www.gskill.com/en/product/f3-2133c9q-16gab

No hardware changes for the PC since it was built.




I have ran furmark stress test on the 970 to 11mins @ 65 Celsius - no crash
Trying to repeat this test ^ I can only get around 1-3minutes before crash.. At times I have got past 3minutes to press the shortcut key to turn "fur" on or off and it instantly crashes when i press the key...spike in power? spike in gpu / unstable ?

I have ran furmark stress test with 970 NOT plugged in ( using on board gpu from cpu ) for 15mins - no crash
I was able to turn "fur" on and off during this onboard GPU test randomly, fps changed but had no crash problems..

I recently discovered a temporary fix setting power "target" in OC GURU II to 70% ( kept lowering until it stopped crashing ) for the GPU, stopped crashes 100% during games
I have played for several weeks without any crash, its like magic to have it stop. Then, today I saw the first crash again, and I continued launching my game to try keep playing to have it crash within 1-10mins, so the problem has returned. I attempted lowering the power limit further, all the way to 50% and the crash is still happening.. My temporary fix is no longer working.

Seems to happen during a spike or sudden explosion etc
IE grenade/ bomb / shot close range with effects of smoke / sparks in game
or turning "fur" on or off quickly in furmark stress test and it almost on command immediately restarts

Always seems to happen under load, never at idle. I can watch movies / tv shows for 6hrs straight and not see 1 crash.

At this stage I figure its the PSU or GPU and will have to possibly send them both to shop to test them..
Thoughts?
 

tprollisawesome

Honorable
Dec 4, 2013
50
0
10,630
Ive been getting this same problem too! One thing ive noticed is that i get these crashes when my system is idle and sometimes after boot up. One thing i will mention is ever since the batman arkham knight nvidia drivers came out ive never had this problem so frequently. Ive also noticed that sli usage in my gtx 760s is so uneven in all tittles, ill get 70% on one card and 97% on the other card in almost all games. Im gona try rolling back my nvidia drivers
 

tprollisawesome

Honorable
Dec 4, 2013
50
0
10,630

Yea i understand that it just that i should have enough power and my friend with a 1080wat psu is getting the same error code and crashes. We both started crashing and rebooting after this gpu update. I dont know how frequently or when thise crashes started but the post date or your original post is after the drivers were released
 

tprollisawesome

Honorable
Dec 4, 2013
50
0
10,630


I dont think that cards are faulty, after the the batman drivers came out i tested each card individualy and both were able to run at full load