• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

voltage for mild overclock on 2600k

jjj807

jjj807

Senior member
Hey all,

My rig is in my signature below, 2600k on asus p8z68-v pro.In the ASUS UEFI profile, I set the Base clock to 100 and multiplier to 42 to get a 4.2 ghz OC. I put voltage at 1.4v and load-line and phase control to Extreme

C-states are disabled, as is speedstep
virtualization is off
turbo mode is on
cpu capability is 100%-default

ran intel burn test maximum and i am stable, speedfan said core temps were around 65-70C during testing. 29 at idle. But I am wondering if this voltage (1.4) is too high for my 2600k? Should i stay at Intel's recommended 1.35?


Thanks!
 
Last edited:
read around and lowered my vcore to 1.3

Stable so far 1.3 @ 4.2ghz 30c idle


Can anyone give a little advice? This is my first SandyBridge OC.
 
Typically you don't want to adjust voltage *unless* it is unstable. That you just now noticed 1.3 was stable makes me assume that you never even tried it lower than 1.4 until now. Try stock voltage before raising it. At only 4.2 Ghz, it would surprise me if it didn't work.
 
So 4.2 is achievable at stock vcore, 4.4 not so much, trouble restarting the computer without bios error saying failed overclock- voltage is now at 1.3 for 4.4ghz OC. Will keep you posted
 
FWIW here's the minimum voltage I found needed to be stable (passes 5 cycles of IBT) at every multiplier:
Intel&


(obviously not all 2600K's are the same)
 
IIRC Idoncare raised an interesting question not about minimun stable voltage but minimun "non-eccdroppingperformance" stable voltage.

The experiment was something like testing until you reach the lowest stable then increase the voltage. If the gflops of LinX increase then test again raising the voltage until the performance doesn't increase.

There's no point in having minimun voltage at maximun frequency if you're hindering the performance, increase the voltage or lower the frequency.

BTW this is from the top of my head, Idoncare can corroborate this or invalidate me. (God that sounded weird, excuse my poor english XD).
 
You also want to use the offset instead on manual on the voltage in the bios. If you set it to manual and 1.3v, it will stay at 1.3v even when your cpu clocks down during idle (assuming you have the power features enabled). If you dont have the power features enabled, do so.

Try setting it to offset and +.05. Set the LLC to medium and go from there.
 
remyat said:
IIRC Idoncare raised an interesting question not about minimun stable voltage but minimun "non-eccdroppingperformance" stable voltage.

The experiment was something like testing until you reach the lowest stable then increase the voltage. If the gflops of LinX increase then test again raising the voltage until the performance doesn't increase.

There's no point in having minimun voltage at maximun frequency if you're hindering the performance, increase the voltage or lower the frequency.

BTW this is from the top of my head, Idoncare can corroborate this or invalidate me. (God that sounded weird, excuse my poor english XD).
Click to expand...

^ you got the nut of it.

But it's easy to test. When you find your minimum stable voltage, get a good baseline number for GFlops with IBT. Then bump up your voltage a good amount (but not silly)...say your stable Vmin is 1.25V, bump it up to 1.30V, and retest for GFlops.

The numbers should be the same, if they aren't then that says your current Vmin value (1.25V in this example) is too low.
 
The asus OC Tuner is pretty nice, 103 base clock, 4.6ghz, 1.4v. Considering all you do is hit a button to get a really nice OC that is stable is impressive. way to go asus
 
mine runs 4.2 at stock voltage all day. haven't tried to go any higher.

i just changed the multiplers to 42 across the cores. did not touch any other settings.
 
Last edited:
I just used the Asus tool to do an auto overclock and IIRC, I'm at 4.4 Ghz at roughly 1.3 V. I think I can tweak it to go a little higher but I don't think I want to use 1.4 V.
 
You don't need to change the votlage and you don't need to input a voltage. Your best bet for power management is to set the voltage to normal and that way your cpu will still manage the voltage depending on the load on the cpu.

To increase the voltage just up the dvid.

I would leave cie, c3/c6, eist on.

I run my 2600k at 4.4 at 1.26 all day.
 
Idontcare said:
FWIW here's the minimum voltage I found needed to be stable (passes 5 cycles of IBT) at every multiplier:
i7-2600KonMIVE-Z.png


(obviously not all 2600K's are the same)
Click to expand...

IDC - have you had the chance to run your tests on an additional 20-30 2600K's? That would really give a nice boost to the sample-size and help the curve be more representative of the overall 2600K population.

I'm kidding!!

Your graph has been awesome. Since SB came out, all we usually got was the same data points with anecdotal evidence (4.2ghz @ x.xxv, etc.) It has been great to see them plotted along on a graph to really get a good picture. I don't remember if I thanked you in the original thread, but I thought I should here, just in case.

Thanks, that probably took a fair amount of time and effort, and the whole population here is better for it. Great stuff.
 
IDC's chart is about my experience as well - my 2500k does 4.0ghz (40x100) at 1.18v BIOS (1.20v HWMonitor) PrimeL/inX stable, 4.5ghz (45x100) at 1.31v BIOS (1.32-1.33v HWMonitor) Prime/LinX stable. I haven't tried higher than that because my Z68X-UD3H and Hyper 212+ are not really designed for higher voltages/heat/noise level.
 
ExarKun333 said:
Thanks, that probably took a fair amount of time and effort, and the whole population here is better for it. Great stuff.
Click to expand...

Finding the minimum stable Vcc for every multiplier between 16 and 50, stable enough enough to pass 5 cycles of IBT as a minimum...take a long time? Nah 😀 Just a week or so 😛

Many thanks for your kind words 🙂 I've actually got another installment coming up where we explore the impact of temperature on power-consumption and overclocking 😉
 
Think I found a good medium- Asus TPU switch on, which basically turns the oc tuner on for you:

4430mhz
x43 multiplier
Bus 103 mhz
1.3v
DDR3 1684
speedstepping is on allowing for the vcore and mhz to fluctuate. Seems to be working fine
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top