2500k weird behaviour

Ivan Erohin

Honorable
May 5, 2013
21
0
10,510
Board: p67a-g45(B3)
Problem: i've been running 2500k at 4.5 with 1.305v for about 3 years, without any issues, but yesterday i've noticed some weird stuff... (maybe it's been like that all the time, i dunno)
Most of the time it's stable, but sometimes one of the 4 cores(randomly) can rise to 80-88 multipler for a sec(or 55), or drop to 2 multiplier for a sec (then it goes normal 45 multiplier), sometimes all 4 cores are rising for 46-48, or dropping for 42-43.
What i tried to do:
1) Disabled green power features in bios (C-States and stuff)
2) Disabled turbo mode in bios
3) Disabled spread spectrum in bios
Well.. no results. What else should i do to prevent this from happening? All i need is 45 multiplier all the time, thanks in advance
I'm going to run some stress tests for now, to see if it's stable.
update: can this be just a MSI Control Center going mad? I never changed anything from there..
 

Ivan Erohin

Honorable
May 5, 2013
21
0
10,510
Here is more..
http://funkyimg.com/u2/4551/696/7764351.bmp - normal, idle behaviour
http://funkyimg.com/u2/4551/697/7586522.jpg - stress-test behaviour
As you can see, voltage and multiplier are dropping.
I have a good cooling (10x CNPS Performa and decent case airflow), i heard two things about it:
1) This happens when CPU hits TDP limit.
2) Multiplier and voltage wont be dropping during usual gaming.
Is that true, or should i increase TDP limit? Oo
And yeah i really think MSI CC is mad.. because min\max multipliers are 16\57
upd: fixed the voltage at 1.34(1.328-1.336) in bios, disabled EIST also.
Multiplier still drops to 38-40 under load, temps are not going above 65C