First Overclock - How is this?

RichH

RichH

RichH

Associate
Joined
1 Dec 2011
Posts
434
Hi Guys,

This is my first post here as I have just started overclocking. I have just finished a new build and am OCing it at the mo. Got a few questions though if you dont mind:

How am I doing so far?
Would you guys consider this voltage and temp safe or too high?
Should I push it further or leave it there?
Is this OK for 24/7 or should i down clock it a bit?

OC Spec:

Chip: 2500k
BCLK: 100
Multi: 50
Voltage Offset: +0.015v
Voltage in CPUZ under load: 1.4 - 1.416
LLC: Enabled
PLL: Enabled (Sleep mode disabled in Windows)
Idel Temp: 25-30
Load Temp: 68-75 (max)

Thanks in advance.
 
The OCUK recommended max chip voltage is 1.38(MAXIMUM), but I have seen others on the net saying they go 1.4v or higher, of course they do so at their own risk. I wouldn't push mine that high.
 
OK thanks. I have seen many different recommendations for max vcore so I am not really sure what to go by.

Cant remember where i read it but I though Intel's position was 1.52 +/- 5%, which is 1.444v.
 
If you're comfortable with the voltages (and many people are - even OcUK have bundles clocked at 1.4V) then your temps and clock are fine. What have you got strapped to the chip for cooling?

What's the vCore at ~4.8GHz? If the drop in vCore is significant with ~200Mhz knocked off i, personally, would consider dropping the clock - for energy consumption and heat . But as mentioned before if you're comfortable with the voltages and energy used then it's not a problem.
 
RichH said:
Cant remember where i read it but I though Intel's position was 1.52 +/- 5%, which is 1.444v.
Click to expand...

I think you're getting mixed up, Intel's recommendation for the RAM voltage is 1.5V +/- 5%.

I don't see why lower than 1.425V wouldn't be OK but no more than 1.575V.
 
Last edited:
Plec said:
If you're comfortable with the voltages (and many people are - even OcUK have bundles clocked at 1.4V) then your temps and clock are fine. What have you got strapped to the chip for cooling?

What's the vCore at ~4.8GHz? If the drop in vCore is significant with ~200Mhz knocked off i, personally, would consider dropping the clock - for energy consumption and heat . But as mentioned before if you're comfortable with the voltages and energy used then it's not a problem.
Click to expand...

OK thanks. I have a NH d14 cooling the CPU.

If i went back to 4.8 I could disable PLL, which would mean i could use sleep mode again. I cant remember the exact vcore at 4.8, but i think it was less than 1.35.

I was thinking of doing this for 24/7 but keeping the 5ghz profile saved in bios.
 
No the only info on the Intel spec sheet for vcore says 1.52v, personally wouldn't head north of 1.4v. Your chip sounds similar to mine, but as mentioned dropping down to 4.8ghz may give you a lower vcore. I need around 1.312-1.32v for 4.8 but around 1.42v for 5ghz.
 
pgi947 said:
No the only info on the Intel spec sheet for vcore says 1.52v, personally wouldn't head north of 1.4v. Your chip sounds similar to mine, but as mentioned dropping down to 4.8ghz may give you a lower vcore. I need around 1.312-1.32v for 4.8 but around 1.42v for 5ghz.
Click to expand...

OK, are you using PLL overvoltage?
 
RichH said:
I was thinking of doing this for 24/7 but keeping the 5ghz profile saved in bios.
Click to expand...

If the system is stable at 4.8Ghz at 1.35V i would certainly consider using that profile for 24/7 use. 200Mhz is nothing and the voltage/heat saving plus sleep benefits far outweigh the Kudos of 5Ghz.

And 4.8GHz is still a mighty clock on air and those are good voltages - you have a great chip.

RichH said:
I was thinking of doing this for 24/7 but keeping the 5ghz profile saved in bios.
Click to expand...

Sounds like a plan.
 
Yeah, i think i'll do that, thanks. And good to hear the chip is a good one, I suspected it was.

Slightly off topic, you say that that is a good clock "on air", but I dont see how water would allow for a better clock as the limiting factor here is voltage not temp.

Am i missing something?
 
RichH said:
Slightly off topic, you say that that is a good clock "on air", but I dont see how water would allow for a better clock as the limiting factor here is voltage not temp.

Am i missing something?
Click to expand...

I'm not the best person to explain as i clock to get the maximum oomph from my CPU with as little voltage and heat as possible (different strokes).

Voltgages are the main limiting factor but the by-product, heat, is also a contributing factor to stability (obviously) - even when not hitting or getting close to TJ Max.

However, in your particualr instance this doesn't appear to be the case - had you pushed further then better cooling - i.e. custom water - will/may have helped.

There are lots of instances where forum members have been pushing for higher clocks and temps haven't been a major concern and yet clocks had failed. And yet when they've upgraded their cooling - be that a better air cooler (yours is as good as it gets) or custom water - they sometimes find that they can use lower voltages for the same clock, and then push on for significantly more. And yet temps, on initial inspection, did not appear to be an issue - but clearly they were a contributing factor along with voltage.
 
Last edited:
Plec said:
There are lots of instances where forum members have been pushing for higher clocks and temps haven't been a major concern and yet clocks had failed. And yet when they've upgraded their cooling - be that a better air cooler (yours is as good as it gets) or custom water - they sometimes find that they can use lower voltages for the same clock, and then push on for significantly more. And yet temps, on initial inspection, did not appear to be an issue - but clearly they were a contributing factor along with voltage.
Click to expand...

I see thanks.
 
pgi947 said:
No the only info on the Intel spec sheet for vcore says 1.52v, personally wouldn't head north of 1.4v. Your chip sounds similar to mine, but as mentioned dropping down to 4.8ghz may give you a lower vcore. I need around 1.312-1.32v for 4.8 but around 1.42v for 5ghz.
Click to expand...

Hi
Do you mind sharibg your settings if you havent done so, just so I have a starting point. I am struggling to get even 4.7ghz. Thanks
 
k_poh said:
I am struggling to get even 4.7ghz. Thanks
Click to expand...

You may have hit your chips natural limit - and 4.6GHz is a respectable clock on a 2500K. Some members hit a brick wall at 4.4GHz so don't feel you're missing out at 4.6GHz.

Sometimes upping the the CPU PLL will stabilise stubborn high clocks.

What vCore are you experimenting with?
 
k_poh said:
Hi
Do you mind sharibg your settings if you havent done so, just so I have a starting point. I am struggling to get even 4.7ghz. Thanks
Click to expand...

Is that aimed at me?

My 5Ghz settings are in the first post, I left everything else at optimised defaults, other than the RAM settings obviously. I don't have the option to manually set vcore so had to do it by changing the vcore offset and checking the resulting vcore in cpuz, which is a bit of a PITA.

The key thing for me in getting from 4.8 to 5.0 was enabling LLC to get rid of the vdroop and enabling PLL overvoltage after being hugely unstable at 4.9 and being unable to boot to windows at 5.0.

Bear in mind that if you enable PPL overvoltage you may not be able to wake from sleep, so sleep needs to be disabled in windows.

My plan of attack was to get stable at 4.6 then slowly increase the offset until I got a 1.35ish vcore. I then increased the multiplier as high as i could until it became unstable, which it did at 4.9. I then enabled PPL overvoltage which allowed me to boot to windows at 5.0. This was unstable but enabling LLC got rid of the 0.1v vdroop and made a real difference. I was then able to increase voltage to 1.4 (via v offset) at which point it seemed to get stable.

Hope this helps.
 
You must log in or register to reply here.
Back
Top Bottom