The Official OC (OverClocking) Thread!

17m 38.516 for 32m



edit

added img

i didnt turn av/as/fw or msn off so i might get it quicker that way, i have also read that intels chipsets are quicker than nvidia aswell as bigdks tips on os choice for those interested in racing.

 

I did my test with msn, skype and googletalk running (although not in voice chat), along with AVG. However, more effectual I'd say was that the PC had been on for several hours and had played several games by that point.

 

i didnt consider a restart as mine too has been on for a while. was it yesterday i was fiddling with my ram or the day before?

this is on my 24/7 oc and i know 3.3ghz and tweaked ram will improve that time but to be totaly honest im not sure wether i will run it at 32 again. never say never but as a stress test i rate orthos higher as my temps never went above 34c with this.

 

The general rule of any OC to be accepted on forums now, is for Orthos or Prime to be run for at least 8 hours full load on both cores, any errors or stopping of the test is not accepted as valid.

 

The strange thing is, I'm relatively confident my PC will pass that test, but yet I know it isn't 100% stable, because occasionally it will BSOD when windows first loads. However, if I get past that stage (which I do most of the time) then it runs fine for the rest of the day. I mean, yesterday it happened once at initial boot, then i restarted and ran 3dmark06 (for score, not stability test), and then played some games for 5 hours, Counterstrike, Hidden, Supreme Commander, Company of Heroes - all ran fine. The system was on for about 15 hours in total, no glitches, it's bizarre.

 

sam validate cpuz and post the link so i can see if that shows anything that might be doing it.

 

Apologies I thought I did.

"probleme de taille fichier" apparently.

 

sorry you may of done and i have missed it, will recheck now.

edit

yes you did 2 pages back, my bad.

what have you got your ram running 1t or 2t?
make sure its 2 unless you spend time playing with your rams timings and volts.

also could be vdroop under real heavy load, give cpu another notch up on the volts see i that works.

mine will boot with 1.32v but needs 1.34 to combat the vdroop in orthos' ftt test.

 

2T, and 1.35V with the +100mV offset, measures out at 1.42V.

 

does it need the +100? that works out at your giving it 1.45v mine doesnt need that to hit 3.3ghz.

run orthos and watch your vcore to see how much it drops under full load, this is what i did to find out how much i realy needed and how much i needed to give it to keep above that level.


edit

only thing i can think thats making you crash is you nb, i had a few unexplaind crashes due to my ram but i have had no trouble getting to 1640 fsb with my ghetto engineering.

 

All CPUs are different in their need for current.

My E6600 is very power hungry compared to most I've seen, but if I try to drop it down to what others have managed, I just can't keep it cranked up.

As long as you keep it cool enough, it shouldn't matter that you're having to stick more voltage into it to enable it to get its required power.

 

here's Super PI 32M @ 3.6ghz



my cpu voltage in cpu-z keeps fluctuating from 1.175 to 1.408. anybody know why this is happening? in everest ultimate it fluctuates from 1.40 to 1.42 which isn't so bad. I have the vcore set at 1.4250 in the bios.

 

1.56V is unfortunately compulsary for my overclock, since 1.45V limits me to 2.25Ghz.

 

Which versions of these E6600s are you guys finding to not OC that well as stated earlier in this post? What is the product code (stepping number revision?

 

AMD X2 5000+ ; Which of these 4 revisions of the new AM2 socket ar a better core to take over 3+GHz on the ASUS CROSSHAIR? I find 2 in a 65nm & 2 in a 90nm. 2 as 89w & 2 as 65w

ADA5000IAA5CU , ADO5000IAA5DD , ADO5000IAA6CZ , ADA5000IAA5CS

http://www.amdcompare.com/us-en/des...000IAA5DD&opn=ADO5000IAA6CZ&opn=ADA5000IAA5CS

 

I think I'd say the 65W ones are the 65nm product. My money's on the 65s for the higher OC, but I can't say for certain.

 

I'm at 3.6ghz. I've got the vcore set at 1.4250 and fsb termination voltage, mch voltage, ich voltage all still set on auto. I can run Super PI at 32M, everest stability test, and encode with RB/CCE no problem

however, I can not say it is completely stable because it crashes when I try to run OCCT or ORTHOS even if I bump up the vcore to 1.450 and mch to 1.65. when I tried running those proggy's the cpu core temps bumped 66 C before crashing. that's too hot anyway and I'm not going to raise the vcore higher than 1.450. while running RB/CCE the cpu temp bumped 46 C and cpu cores bumped 60 C.

for what I'm going to do with my pc (mainly encoding), I am content with the performance and temps at 3.6ghz and my encode times are blazing fast. I get no BSOD's, lags, or lockups. I will add I could run ORTHOS and OCCT at 3.4ghz without it crashing.

 

cpuz has some trouble reporting the cpu voltage sometimes.

when i was fisrt trying to get get mine as high as i could i stopped at 1.55v as the extra .11 volts didnt help me get to the 1600fsb on x9, cpuz was saying i was giving it 1.2v both speedfan and everest got it right tho.

when im testing settings i tend to use 2 of these 3 and if they dont agree i will bring in the other to see if it agrees with one or the other.

nice scalling on superpi with it at 3.6ghz, have you done any tweaking of your ram yet? i took (just) over a second off my 1m time by tightening my timings from 5-5-5-15 1.8v(auto) to 3-4-3-11 2.1v.


speaking of ram voltages, nvidia have said that anybody who is using 2.4v over a prolonged time on their sticks will melt them. thats nice to know isnt it. they also said that its not just there chipsets that do this but all of them (iv only seen people with nf680i mobos report this so far but who knows?). hold on and i will look for a link and edit this in a mo.

 

OK I found out the BSOD issue, you'll love this. If I turn the system on for the first time in a day, I always get the BSOD, restart and it's absolutely fine. If I restart it, it is too, but if I shut it down, when I next turn it back on again, BSOD.

 

not really, I did up the voltage to 2.2 but I am still running the timings at 4-4-4-12.

just for testing purposes, I did up the vcore to 1.4625 and mch to 1.65 and reran orthos and occt. orthos did eventually crash again but ran much longer. I shut down occt, it was still going strong. I see no reason to stress my components that hard, that long since I will never be stressing them like that in any real world applications.

I'm sure if I uped the vcore one more setting, it would pass orthos. It encodes with RB/CCE rock stable at 3.6ghz with vcore set at 1.4250, and mch, ich, and fsb termination voltages all set on auto.

I'm also more comfortable with the lower temps at the lower voltages. at these settings at idle my mobo temp is 32 C and my cpu temp is 33 C. yesterday while encoding the mobo got to 35 C and the cpu bumped 46 C but the ambient temperature was also 5 degeees F (2.8 C) hotter then. I'm very content where it's at and how it performs. no BSOD, lags, or lockups. here's a few bencmarks.