I5 2500k Overclock - will 1.22v fixed decrease lifespan?

Cpt Deadboots

Distinguished
Dec 12, 2007
318
0
18,780
Just got my new hardware! I got right down to business (overclocking, that is) - I have an i5 2500k on an ASRock Z77 Extreme 4 with 8 GB G Skill Ares DDR3 1600.

So far I've hit 4.5 GHz at 1.22v (fixed) and it's been stable for 3 hours on Prime95 Blend. Is a 3 hour test acceptable for determining O/C stability?

Is running on fixed voltage going to have a detrimental effect on the lifetime of this chip? I'd use offset, but I can't figure out how to get it to operate properly - regardless of how low of a setting I use, my vcore shoots up to ~1.45 when I start Prime, and my temps immediately hit 80+.
 

Z1NONLY

Distinguished
1.22v isn't bad at all. But it's not as good as the 0.9xx you would have at idle with an offset voltage overclock.

Have you tried using a negative offset? At 4.5Ghz you probably need an offset of at least -0.075v to get it down in the 1.2xx range.
 

Cpt Deadboots

Distinguished
Dec 12, 2007
318
0
18,780
-.140 gives me 4.5 @ 1.224v load - previously I was misunderstanding how offset settings worked, heh. Stable after 1 hour of blend; seems pretty solid. I might be able to go lower, but is it really worth figuring out the EXACT .05v increment that causes a crash? I'm alright with the 71c max I saw during the test, and I can't imagine how tinkering further would really benefit the chip.
 
Running 4.5ghz at 1.22v fixed is great, I wish mine could have done that good it took 1.325v for me to run 4.5ghz with my 2500K, of course I did have one of the first ones released.

The temperature under P95 load is fine, you will not be running P95 all the time!

You didn't mention the CPU cooler you're running, you can bring down load temperatures with a better CPU cooler especially at that low a fixed CPU voltage, but as far as longevity I don't see a problem.

I've used fixed voltage since the first day of running my 2500K.

Additionally you cannot depend on P95 as a total stability test as it does not test GPU or Sound added to the mix. you'll need to additionally test using benches like 3DM11, or Vantage, to test the added GPU load.



Congratulations! Ryan
 

Z1NONLY

Distinguished



4.5Ghz with only 1.224v and power-saving features enabled is really good. No real need to go lower on the voltage. (unless you just want to experiment.)
 

Cpt Deadboots

Distinguished
Dec 12, 2007
318
0
18,780


Hmm, I must have mounted it wrong or applied the thermal paste incorrectly. I have a Cooler Master Hyper 212+, and, now that you mention it, load temps probably should be a bit lower with this low of a voltage.

I'll test with Vantage later today. Seemed ok in Crysis earlier - I'll run the bench included there as well, and maybe Unigine Heaven when I get around to downloading it.

Thanks! I've been really happy with this chip and board so far. I've never owned hardware this simple to overclock; I'm still pretty awestruck by how easy it was/stable it is.
 

Cpt Deadboots

Distinguished
Dec 12, 2007
318
0
18,780
I believe my room's ambient temp is around 70 F. You're sure I didn't do something wrong? My coldest core hangs out at 61 C running small FFTs in Prime; hottest is anywhere from 68-72. It just seems strange to me that I'm getting temperatures that many people see at 1.3v +. I see temps of 80 C pretty easily when I give the chip 1.36v.

Perhaps the paste hasn't cured yet. I'm using Arctic Cooling MX-2.