2500K (Sandy) over clock results

SimonMaltby said:
I could not decide between High Frequency / Poor CAS or Lower frequency Better CAS so I brought 2 sets of RAM and will sell one when i'm done testing.

So I have 2x2GB Patriot 2400 CAS 9 and 2x2GB OCZ 1600 CAS 6
So far I have just used the Patriot at 2133 CAS 9




I'm a bet sceptical about this thread. There are no screen shots or confirmation of chip/mobo

I don't think 5GHz is going to very easy to achieve at all. So far I have managed to do superPi 1M at 4.7 but at 4.8 I get an instant blue screen.


There seem to be far more overclocking options than I thought there would be. The Asus board has a great DIMM control and Voltage adjustments. This is where the real fun begins imo finding a way to get the magic 5GHz. For now i'm happy that 4.6GHz almost out of the box was a nice easy overclock - comparable to getting 4.0GHz from an i7 920 in terms of difficulty.
Click to expand...

I was not able to provide screens due to the Intel NDA but since people have taken delivery of the new CPU. Here you go.

Done a few minutes ago - ambient is quite a bit lower than it was when I first did the testing.

101BLK x 50

50x1020.png
 
I have to admit to beeing a bit disapointed with my 2500K
What ever I try I can't get more than a few points above 4.7GHz and that requires 1.35v Vcore and no other tweaks. I was really looking for more.

On the subject of RAM.
I cant really tell the difference between 1600 as CAS 6 or 2133 at Cas 9
The latest Bios for my motherboard has a 2400 Ram speed option, but I can't get it to boot with that at the moment.
 
I think the safer vcore for both i7 2500k and i7 2600k is no more than 1.40v at the bios to be on a safe side for longer life. Anything over 1.40v will become shorter life depend on temperature. I would stick to below 1.40v to be honest and I think 5Ghz isn't worth to risk it. Anything between 4.4 and 4.6 Ghz is more than enough.

I just saw on the website that someone did 4.8GHz overclock (Overclocking - CPU 2nd Generation Intel® Core™ Sandy Bridge CPU professionally overclocked by our engineers to 4.8Ghz stable guarantee on air cooler with max temp 71c with 1.36vcore)
 
Last edited:
bulldog147 said:
........
Click to expand...

To me is a fail to some extent, my 3 years old 45nm C0/C1 I7 920 can do 4.5GHz at 1.45v. This has to do at least 5.0GHz, mainly because there is no hyperthreading holding it off. I said to some extent because this is no high end so I would not expect much. For Lynnfield users this is a great upgrade and even 4.6GHz as you said would do wonders but is still a meh.
 
bit tech said 1.4v is for benchmarks only on thier review/guide on sandybridge.

i would probably use 1.35v 24/7 or 1.36 until someone says otherwise.

so onto the main topic simon are you booting at 47 multi? i dont think i saw any above 47 multi as even bit tech used a 47 with 105 base clock.you could always try that.can the motherboard adjust base clock to 105?


quote

We overclocked the i5-2500K by first setting the Load-Line Calibration of the Asus P8P67 to Extreme. We then set the Base Clock to 104.7MHz and the CPU multiplier to 47x for an overall frequency of 4.92GHz. To make this stable, we set the CPU voltage to 1.35V, the CPU PLL to 1.9V, the VCCSA to 1.1V and the VCCIO to 1.106V. See How to Overclock an LGA1155 CPU for more details on how we overclocked all the CPUs.


so there was 4.9 and the 2600k only did 4.8 as they said they winced it out of it with a 48 multi.it sounds like they failed to be able to adjust the base clock which i find weird? perhaps 2500k is more friendly to base clock adjustment?


keep us posted and i would not worry about 2400mhz to be honest.2133mhz at 1t and the timings u planned to use should be spot on.and also those uefi bios's might not be fully refined yet to allow a higher multi or base clock so make sure u google as to which bios is most stable.latest is never greatest as i remember the trouble i had on this E6300 getting past a wall at 3.2GHZ.every bios but the very first one(f4) would put me down to 2.8 and refuse to boot no matter which settings i changed in the bios i knew like the back of my hand.
 
Last edited:
Metroid said:
To me is a fail to some extent, my 3 years old 45nm C0/C1 I7 920 can do 4.5GHz at 1.45v. This has to do at least 5.0GHz, mainly because there is no hyperthreading holding it off. I said to some extent because this is no high end so I would not expect much. For Lynnfield users this is a great upgrade and even 4.6GHz as you said would do wonders but is still a meh.
Click to expand...

Sandybridge isn't 45nm as it 32nm! Big difference!
 
These Sandybridge CPU's are not replacements for the i7 range socket 1366 CPU's. They are socket 1156 replacements. Think that is being lost in the discussion.
 
I'm still not getting very far. I feel i'm getting closer to 4.8 but it's very hard work. 1Day, I know you are using a different Motherboard (Gigabyte UD7) and I'm starting to wonder how limited I am by the chipset / mofsets.

I wonder if I can compare any of your bios settings to the Asus and see if i can find some common ground. I'm also toying with resetting the Mofset heatsinks on the board.
 
Yes sure no problem

All you need to tell me is what your voltages are.
And what your main settings are, most are just default or auto.

Leave your board alone for now. :D It is fine I am sure.

Oh and just to spin you out. I have installed a fresh copy of Win7 while at 5GHz :) This board is my new 27/7 kit and I needed to see what the SATA6GB/s boot times where like.

Oh man I really am pleased I got this Gigabyte P67-UD7 motherboard.

And it looks like I might have sold my Rampage Extreme II on the MM. So a good day all round.
 
1Day said:
Yes sure no problem

All you need to tell me is what your voltages are.
And what your main settings are, most are just default or auto.

Leave your board alone for now. :D It is fine I am sure.
Click to expand...

I have just stripped everything down and dropped to bios 1003 from 1053 as 1053 is Beta.

My voltages are:
Vcore: 1.35v
VCCSA: 0.925v
VCCIO: 1.056v
CPU PLL: 1.796v
PCH: 1.031v

I have the following set
Spread Spectrum: Auto
Loadline Calibration: Expert

These are really all the Asus P8P67 Pro has. How do they compare to the Gigabyte?
 
About oc 2600k
"For processors with unlocked multiplier, which is 2500K and 2600K, the situation is different. First, processors retail or retail processors similar ESs (ie revision D2) you will need the latest BIOS, which should include Internal vPLL Override option, your replacement could pass 5GHz. Currently available BIOS for Asus, Gigabyte, Biostar and Intel. Because I used Gigabyte in P67A-UD7 test than we deepened the problem in Gigabyte motherboards. Bios for UD7 sites, and UD4 UD5, along with necessary utility BIOS update can be downloaded from here. To move to 5GHz, vPLL Override option should be set to Enabled.

Once this option is set, if overclocking CPU K series is all about setting and multiplier's vCore. Currently, using 1.6v for the CPU, I was able to reach 5400MHz with no problems using a simple cooling NH-U12P. More details about the multiplier or scaling of the process temperature, BIOS version, settings and other little "secret" you can find an article dedicated to this subject."
http://lab501.ro/procesoare-chipseturi/sandy-bridge-partea-a-ii-a-mainstream-with-an-attitude/34
To download the latest bios just click on the green "aici"
 
q6600 said:
Currently, using 1.6v for the CPU, I was able to reach 5400MHz with no problems using a simple cooling NH-U12P.
Click to expand...

Whole bunch of people are going to cry seeing 1.6v going through that chip. Would love to see how long it actually takes before it pops.
 
I have tried the 1053 (Latest Beta Bios) and enabled the Pll Override, but it did not help. I just flashing back to it now and will try again.
 
Please folks do not push 1.6volts through your CPU for any extended length of time.

Monstru the guy who wrote the article is an extreme bencher and reviews hardware for the extreme user not every day user. Oh and he likes dogs and beer. :D
 
You must log in or register to reply here.
Back
Top Bottom