Kernel Power 41 crash?

Garon Booth

Reputable
Apr 1, 2014
224
0
4,690
Hi, first let me start with my problem, lately after upgrading my gpu to a gtx 960 ftw, from a smaller r7 260x I keep getting blue screen errors, machine check exceptions, watchdog timeouts, I have a g3258 clocked at 4.2 , with stock fan, my core volts are 1.282. I never had this problem when I had my r7 260x, same overclock and everything back then. What kind of tests/programs can I do/run to locate the problem? Also note ive never done a stress test in my life. thanks for the help. greatly appreciate it. ( the crashes usually happen after playing games for numerous hours, its not one game in particular) also what types of results should I be getting in the tests. thx again!
Build
g3258 4.2
gtx 960 ftw
8gb ram
evga 500psu 80+ bronze
h81m-e34 mobo
 
Solution
Usually means vcore is starved. Try bumping vcore up to 1.29v. Since you've never run a stress test, its probably your issue. There's a good test over at the Asus ROG website. For temp testing use realtemp, cpu-z and prime95 v26.6 with small fft. For the kicker, that stock cooler isn't really made for OC heat. You are taking a 53w cpu and pretty much doubling it's heat output. This too can cause BSOD after several hours of heavy gaming since the case insides are slowly building heat, which lowers cooler efficiency. This is why you temp test. P95 will simulate 100% cpu load. Run it for an hour. Use the realtemp (designed for Intel cpu's and for use with p95) to monitor temps. If the cpu is reaching 80°+, back off the OC.

Karadjgne

Titan
Ambassador
Usually means vcore is starved. Try bumping vcore up to 1.29v. Since you've never run a stress test, its probably your issue. There's a good test over at the Asus ROG website. For temp testing use realtemp, cpu-z and prime95 v26.6 with small fft. For the kicker, that stock cooler isn't really made for OC heat. You are taking a 53w cpu and pretty much doubling it's heat output. This too can cause BSOD after several hours of heavy gaming since the case insides are slowly building heat, which lowers cooler efficiency. This is why you temp test. P95 will simulate 100% cpu load. Run it for an hour. Use the realtemp (designed for Intel cpu's and for use with p95) to monitor temps. If the cpu is reaching 80°+, back off the OC.
 
Solution

Garon Booth

Reputable
Apr 1, 2014
224
0
4,690


Yes, I agree. I ran prime95 with realtemp. While I stayed at 1.282v and my temps were crazy I was peeking 90 Celsius..? Should I try to increase volts or bring my clock down?
 

Garon Booth

Reputable
Apr 1, 2014
224
0
4,690


Ended up downclocking to 4.0 and lowering my vcore, cleaned fans, and ran p95 for around 15 min, max never went over 81c