**Official e8400/e8500 Retail OC Thread

[YMAA]

Stable as a bird. It was running FAH all last night without a hiccup. The chip is getting pretty hot though. I may remount the waterblock later today, I might've shifted it while doing the tubing.

[misho]

OK SO I tried 9x multipliers higher FSB voltage, higher pll voltage, 8.5x multi still very unstable at 4.3GHz, fails orthos in 2 seconds, games crash real bad...
I guess there is NO way to get this cpu stable above 4.2Ghz without increasing the CPU voltage from 1.38V to something higher...........................

I read some ppl had their CPU for over 8 months on 1.5V running fine.. what do you guys think?

[jaredpace]

OMFG SEE LOOK AT THIS GUYS ! 1.312V at 4.5Ghz E8400!

Why can't I have that :

cause your doing something wrong.

[YMAA]

What are your system specs, misho?

[unclewebb]

I think once you hit 1.40 volts, you should call it a day for 24/7 use. The difference between 4300 MHz and 4200 MHz is 2.4%. There isn't an application in the world where you will ever notice the difference so why risk stability for such a tiny gain. For benching there's some fun to be had beyond 1.40 volts but for day to day use you don't gain a hell of a lot.

[jaredpace]

1.40v should be "loaded cpu vcore read by cpu-z or DMM"

[subaruwrc]

Here's my E8400 E0, batch Q826A399, week 08/13/08.



Folded all night at this speed. And here's a validation:

http://valid.x86-secret.com/show_oc.php?id=439932

Might try for a higher clock later today.

nice clock with low voltage other settings?
I like your wallpaper

[YMAA]

These are the settings I can remember for that speed. I'm testing 4.6GHz now.

All skews set to Normal
Pull-ins Disabled
All GTLref set to Auto

vNB - 1.2v
vFSB - 1.2v
vPLL - 1.5v
DRAM - 2.1v (still testing...getting used to the new Mushkins)

EDIT: I'll have to find out how to record BIOS settings directly from the BIOS, I think there's a way to do it.

[misho]

What are your system specs, misho?

CPU IS E8500 C0, batch q818a102, date 06/27/08
MOBO is BIOstar Tpower I45,
memory: OCZ REAPER HPC 2 x 1GB, DDR2 8500
Corsair HX 520watts
COOLING: Thermalright Ultra 120 Extreme paired with 2 x Scythe KAZE ultra 120mm 2000rpm.
Temps at 1.38V is 30 idle, 41C MAX
But look at yours, you can EASILY hit 4.7Ghz with something close to 1.38V, and that is crazy.. Soo lucky u got that cpu LOL :jealous:

[LagunaX]

Most C0's cannot compare to E0's for higher ghz at lower vcore - a 4.2ghz C0 is a rare gem that unfortunately is not worth as much nowadays...I have a e8400 C0 that takes 1.37v for 4.2ghz...

[YMAA]

Try tweaking your GTLref. Start at .63 and work your way up. Some 45nm chips seem to like .63, others like .67. Usually if you're running Prime95 and one thread fails you can maintain that vCore and tweak your GTL.

[hktony]

Most C0's cannot compare to E0's for higher ghz at lower vcore - a 4.2ghz C0 is a rare gem that unfortunately is not worth as much nowadays...I have a e8400 C0 that takes 1.37v for 4.2ghz...

i need to stablize my cpu at 1.45 v at bios to have 4.25ghz ....
poor me c0 e84

[hktony]

Try tweaking your GTLref. Start at .63 and work your way up. Some 45nm chips seem to like .63, others like .67. Usually if you're running Prime95 and one thread fails you can maintain that vCore and tweak your GTL.

thats mean i can change the cpu gtl voltage and lower my vcore then check if it is stable or not ?

[misho]

K guys I changed GTL voltage, pll etc, memory too but nothing

still VERY unstable

[YMAA]

You have to tweak to find it. EVERY CPU likes a different value. Generally what you do is start with a common value, like .63 or .67, then experiment to find what your CPU likes. GTL takes a LONG time to figure out. It helps to change one thing at a time, it may take longer but makes it much easier to isolate problems (things you need to change).

Are you sure it isn't your RAM? Take your RAM well below stock speeds, run at stock voltage, and do a Prime95 small FFTs test to stress just the CPU. Once you've found a speed your CPU likes, clock the RAM back up and see what it likes in terms of clock speed, timings, and voltages.

[PytonOrm]

update today =) all this is on Air with tuniq tower 120

http://valid.x86-secret.com/show_oc.php?id=440390

5130mhz done * even try 5161mhz but when i save at my hdd was it corrupt

tested benched at 5026mhz with 1,76v and superpi 1M gave me 9,3 sec with timings 4-4-4-4@vmem 2,2v

[jaredpace]

cpu at 1.45 v at bios

this usually means nothing as it is not the actual volts the cpu is getting.

[jaredpace]

update today =) all this is on Air with tuniq tower 120

http://valid.x86-secret.com/show_oc.php?id=440390

5130mhz done * even try 5161mhz but when i save at my hdd was it corrupt

hi, what is the maximum CPU speed that you can pass 1 hour prime/orhtos with? ~4800mhz?

[PytonOrm]

update again

http://valid.x86-secret.com/show_oc.php?id=440530

5192MHZ done *

what is Wr on E8500 and E0 on air ?

[Nilos]

My 24/7 settings




I sold my E8600 (Q822a435) and kept this E8500 E0. It's almost the same to me. The E8600 needed 1.3V to be stable at prime. This a little bit more...

[JCGhz]

Does anyone use the Orthos_Stress_Cpu w/ Gromacs Core feature to determine overclock stability? I can use considerably less voltage to be 3d stable and Gromacs core stable.

I only Game,Surf and do large rar file unpacking. I would like to see how many of you use different methods to determine stability.. i think prime and orthos small fft stressing is excessive.

Your opinions?

[LagunaX]

@Nilos,
my 24/7 settings @ 1.3v on air:


Lol, tables are turned now...just kidding buddy - great chip you have there =)

[subaruwrc]

4500mhz 1.328v
E0 rulez
http://images.netbag.hu/20081102/16005560495.jpg

[YMAA]

Setting CPU Margin Enhancement to Performance on Ket's 1406 BIOS for the P5Q-Deluxe seems to have fixed some instability issues for me. May try for higher later, or a lower vCore. Yeah, higher sounds good.

[jaredpace]

wow 3 beautiful 4500mhz OC's on this page. 1.344v, 1.304v, and 1.328v