**Official e8400/e8500 Retail OC Thread

[zfactor]

WOW HUGE immprovement there!!!!!!!!! nice job

[cincyrob]

even 1.36 is really high imo it takes me 1.268 to do 4.0

i agree. but take in consideration my mobo and mem. you have far superiour stuff than i.....

heck in 6 months my 10yo will have this rig at 3.2ghz and i will have my new one running 4.2ghz (E8600)..lol

would that pencil mod help me any with the vcore? temps? wining the lotto????ect.....

[LagunaX]

Today i stop here...THE PENCIL MODE HAS WORK PERFECTLY.
Thank's a lot man, i appreciate, i try to get better performance tommorow, tx again ..and the temps are better...that's awesome

I have done my good deed for the day

[Rickpatbrown]

i dont think jealous is the correct term. maybe CURIOUS, as to why it so high to just reach 4.0ghz..you have a awsome board and should need that much voltage. my POS mobo dont have half the goodies your's have and it only takes 1.36v for me to get 4.0ghz but i cant get any higher

Yeah . . . I was being sarcastic. That's what this guy is for ->

My chip sucks, plain and simple. Unless I am an idiot and don't know how to clock a timex, this is as high as it will go.

I've had it in the 1.5v range for about a month now. I'm gonna leave it here and see what happens. This should shed some light on the life span of 45nm chips.

I thought about eBaying it, but I wouldn't feel right selling something that has been at this voltage.

An interesting thing is that it is more stable with the lowest PLL setting. I thought high PLL would help stability.

[oli_ramsay]

Yeah . . . I was being sarcastic. That's what this guy is for ->

My chip sucks, plain and simple. Unless I am an idiot and don't know how to clock a timex, this is as high as it will go.

I've had it in the 1.5v range for about a month now. I'm gonna leave it here and see what happens. This should shed some light on the life span of 45nm chips.

I thought about eBaying it, but I wouldn't feel right selling something that has been at this voltage.

An interesting thing is that it is more stable with the lowest PLL setting. I thought high PLL would help stability.

I've got a dud chip too, takes 1.5125 to do 4GHz. I'm not really happy running with that voltage 24/7 especially since the VID is only 1.15V.

I've decided to settle for 3.9GHz at 1.45V (about 1.41 under load) and hopefully get an E0 stepping when they come out

[Rickpatbrown]

I've got a dud chip too, takes 1.5125 to do 4GHz

What batch was yours? I bought mine in April when they were crazy hard to find in stock. I think they might have pushed out some crappier ones since demand was so high.

[oli_ramsay]

Mine's 3rd hand and has obviously been punished, i.e. had 1.55 running through it to get 4.4GHz. The batch is Q744 which is a very early one, but for some reason there's none on the 1st page here with that batch number.

[Pisklink]

It does seem like you need a higher vcore than most, and this looks to be a result of your higher VID. What does coretemp v0.99beta report the VID as?

Also becasue you are running DDR3 at 1700+mhz, your vNB also probably needs to be high to support that, but only testing at lower voltages will tell.

Are you able to lower the vcore any and remain stable? If not, you can almost assume that you need that much to hit that frequency. Just test and try to figure out minimum voltages.

I will take a look what VID coretemp reports and report it back to you.

Furthermore I noticed the settings in my first post are nog stable yet. This evening after work I will stress test with orthos on small FTT's and memory settings on auto to see how much vcore I need to get stable (with max. of 1.5 cause thats the highest I'm willing to go). In case that 450x9 is still not stable at 1.5V I could try to lower FSB to 440 and test again. What would you suggest for VTT (@ 1.38V) and CPU PLL (auto) considering my settings?

Moreover I think this board is a bit picky on memory and memory speed. My memory should be able to handle 1760Mhz speeds, I guess. I have to stick around ~430 - 450 FSB to make sure I get a clock of ~4Ghz but this has a negative impact om my memory. If only 1800Mhz DDR3 would nog be that expensive. I could also try to use a 7.5 MP with ~530 - 540 FSB. But when my memory won't handle speeds of 1700+ what would you say about getting vNB to 1.6V as the highest setting?

Do you think that pencil mode would work for me as it did for So3oL4Nu?

[gurusan]

The OP disappeared

[Copenhagen69]

The OP disappeared

LOL .. good call

wonder where he ran off to

[Calmatory]

What's the maximum safe Vcore you guys would recommend for safe 24/7 usage? My target is 4GHz, I can do 3.9 @ 1.45 in BIOS (1.42 real). I think it needs over 1.5v to go 4Ghz stable. What do you guys think?

Doesn't matter that much, as long as you have temps under control and the CPU is stable.

...or is this Wolfdale somehow different from any other CPU manufactured yet? People seem to be hella careful about vCore here, "omg I passed 1.3635 V does my CPU die soon???"

...or are they just ignorant about the past? When people didn't care what manufacturer said, they jsut wanted the max OC possible with insane vCores.

[So3oL4Nu]

I have done my good deed for the day

You have done a lot more dude ...after 6-7 hours the Vcore idle is rocksolid...i have 46 degres idle, before the pencil mode, i got 49-50 degres idle...
I look today to improve better the Vdrop...is still 0,3 bios / idle, and to little from idle to load...

[Pisklink]

Thanks for your reply and I will give your advise a shot. I hope it works out and will report back the results.

As for FSB testing, you are most likely to hit the cpu highest clock on a lower FSB. So 9 x 445 will give you the 4 GHz, and keep your ram in the high performance zone.

That sucks a bit because my mem is rated for 1600Mhz so 1:2 and 2:3 dividers are not since they would bring a overclock that is too high to be stable.

Unfortunately I have no experience with DDR3 so I'm not much use to you there but you seem to know it quite well.

I will turn to the OCZ forums for this. I already read some things and it seems that many people have troubles getting their OCZ 1600 modules run stable on P5E3 board.

Personally I wouldn't want to be running 1.6v vNB 24/7. Also I'd be trying to keep the cpu under 1.45v for 24/7 and again, whatch your temps.

Me neither but I'm willing to set these values only for testing and see if it's stable. When I found a stable and reasonable overclock (with regard to voltages) I will start finetuning the voltages to get the lowest possible.

[cincyrob]

well i couldnt find (more like i didnt save them) numbers from before.(vdrop)
so i desided to do the OC again...bumped the vcore up in bios to 1.44375v before that the vcore was 1.204v in bios now after the bump up im at 1.392v

but check this out. im stable. i got some high temps under this stress load. so its not gonna stay here unless i get that vendetta2 this weekend. i knew i could get 4.1ghz if i just fiddled with it some more. i think this CPU is a middle of the road chip. it might handle more on a better board and cooler and mem.speaking of my Mem. i took it out of manual(cause i dont know what to set it at for a OC like this)put it back in auto. its running fine at 5-6-6-21- 2t the pick im about to show is of orthos at 38 mins or so. it failed the last 3 times at 17mins and 33mins. it is now at 63mins and fine.

[So3oL4Nu]

I pencil mode it again, and again cu gain the best Vcore...after 1h, i manage this !
Vcore 1.26 BIOS
Vcore 1.248 IDLE
Vcore 1.248 UNDER LOAD, no drop from idle to load ! INCREDIBLE !

I think if i stay 1hour i can drop it more, up to 0,010 drop...but i think i stay with those settings...
A question....how i seal it now?...it's necesery to seal it or can i stay with it like that ?...

[zfactor]

just drop a bit of clear nail polish over the spot and let dry dont brush it on just let it drip over it

[gurusan]

yeah and if you want to remove it use a cotton swap dipped in acetone to carefully brush away the nail polish

[HamTard]

Does anyone know if 9.5 multiplier with an E8400 works? Nobody seems to be using it, but my BIOS gives me the option. I'm guessing it doesn't work anyways, since the E8400 is supposed to be locked at 9, but it seems weird the option is there.

Anyways, How much do temps go up on an increase of 0.05V vcore? I haven't touched vcore because I'm on stock cooler at the moment, and I'm just wondering if anyone has overvolted on a stock cooler with success.

[aussie-revhead]

I am testing an E8500 atm and it gets high max clocks on 9x multi , it doesnt like 9.5x .

[mskvorc1]

Well, here's mine E8500 so far. Done pencil mod. 1.3125V BIOS, 1.296V Idle and 1.304V Load. Although I have bad batch Q816A678, so far I'm happy because I had pretty low expectations with this one:



This is after SmallFTT's, now I have to run Blend...

[So3oL4Nu]

just drop a bit of clear nail polish over the spot and let dry dont brush it on just let it drip over it

Red polish it's a problem? ...and tell me it's a problem if i don't drop a bit of polish?

[zfactor]

color shouldnt matter imo as long as it doesnt have any mettalic flake in it.. at least i would not think so..

[So3oL4Nu]

color shouldnt matter imo as long as it doesnt have any mettalic flake in it.. at least i would not think so..

...it's ok, but u don't answer to the prevision question...if i let it witout the nail polish it's ok ?

[Leeghoofd]

Doesn't matter that much, as long as you have temps under control and the CPU is stable.

...or is this Wolfdale somehow different from any other CPU manufactured yet? People seem to be hella careful about vCore here, "omg I passed 1.3635 V does my CPU die soon???"

...or are they just ignorant about the past? When people didn't care what manufacturer said, they jsut wanted the max OC possible with insane vCores.

So you will refund my Wolfdale if I smoke it ? most 45nm CPU's are plagued by bad sensors so plz do not talk about temps, not every user here has got high end air cooling and some even use the stock cooler as the coretemps seem to be okay.. some CPU's don't even go over 45°C at 4ghz under air... what's the temp really , do you know it ? I have faith in my watercooling to keep it under 65°C at all costs...

45nm seems more fragile to high voltage abuse (not talking 10 min benches) :

General warning for 24/7 use is to stay below 1.4volt Vcore, 1.6 PLL and below 1.4 CPU VTT... dunno why you doubt it... CPU will prolly not die, but could degrade making the end user not happy at all...

Really no idea why you posted this, I understand ya sense of humour or sarcasm but Xs is becoming more and more a general help forum...plz be carefull what you post for the crowd out there...

[Infa]

It's not named Xtreme couse some scary peps think the cpu will explode with some vcore , (mind me not meaning 2v).
i agree with calmatory actually, making this big hype of this overrated vcore problem, the cpu WILL not die of 1.40-45v.
if you are unsure of this thing called OC, then you shouldnt even bother... sure im not saying we shouldnt tell them that they are running things out of spec, and that you cant return stuff if they break, but this paranaoi in a xtreme OC forum is just absurd. i havent killed any cpu yet , and i overvolted many cpus through time, no one is even putting out a sign of degradation, and no one is absolutely not killed. and now i mean some middle xtreme OC , and not Ln2 or dice. who have killed their cpu here? by getting out of spec, and NOT meaning 2v or 1.80, around 1.40-60, maybe theres 2 unlucky bastards out of 10000 .

ahh well ... just felt to say this. nothing personal.

cheers.