E8600 Prime Stable Overclocking Database

[RCG_Bex]

ROFL - so true - I'd 4.2GHz on my E6750 and it'd play games for like 4hrs before it gave in where as 4GHz was rock solid stable

~Bex

[BeastNotro]

I added a couple more .. come on ppl get priming those OC's.....

jaredpace i will take the highest one of yours (thats minimum 3 hours primed) when i have time to look.

Reason why i stated Minimum of 3 hours primed (as a requiremenet of this database) is that many ppl can do 1 hour-2hours, but that 3 hour is extensive and i have seen many times cpu's crash just after 3 hours. Now of course i would love to see your OC primed more then 3 hours, so do your best and prime for longer if you can.

I usually do a minimum of 12 hrs of Orthos Blend and never have had an issue afterwards, but for this 4.2ghz i did 16 hrs before i shut her down and went to bed . I am sure it would have went further, but i was not staying up till 8am to see and i can't leave my PC run all night as its to loud ... then i can't sleep...

As for what type of cooling i did not ask for that as we all know here that:

A: It will be Air Cooling
B: Water Cooling
C: Chilled Cooling/Air or similar
D: LNC Cooling/Mach GT II or similar
E: Beyond Frozen ...

If wanting to know a person cooling type look in their signature first and if not listed, then i suggest PM them or post in this thread to ask what they are using.

I never asked for Ambient temps either .... again just like cooling PM or ask here what it the Ambient temp was at.

I know 4.0ghz is a Joke for a E8600, but i had to make a minimum as some just can't get over 4.0ghz stable on their setup and there is that some that can't get 4.0ghz stable due to their setup, but yes in the 95% chance class ... most should do 4.0ghz NP's stable if all of the setup is good.

[dinos22]

your database will be useless then because no one will dig through pages of posts to try and figure out ambient temps and type of cooling

the idea is to give ppl enough info in first post to be able to make conclusions on certain batches

right now it makes absolutely no sense imo and it will get worse as the list gets bigger

[BeastNotro]

dinos22

I understand what your saying, but if you see a OC that may want to try out on your RIG. This database shows what ppl are getting stable minimum of 3 hours at what voltage was needed to do so as for temps that will depend on a persons cooling rig setup and that will differ from person to person, so even if we know what a persons temps were it does not mean you will have the same temps.

Again all i suggest is PM them or post in this thread to ask what type of cooling they used and/or temps if you wish to know.

No doubt it will be Air or water or Phaze type cooling and if the person has it listed in their specs then that makes it even easier for those who wish to know what that person was running. I am not concerned about what cooling they used as if i tried for 5.1ghz on my Air setup i am sure to get 80C in prime ...

As for kitfit1 i do believe he is running a Mach GT II setup and is at -20c for his OC.

[zsamz_]

you reminded me of my old 3000+ AMD Venice

i was sooo obsessed at having a high clocked 24/7 rig that i ran it at 2750 or 2800 can't recall now and the rig was stable for my use (some games, normal stuff and some video editing etc.) >> the funniest part...........the rig could not even finish 1M SuperPi hahahaha

but i've run it for months and never had issues lol..........how the hell does a chip encode videos for 7-8 hours in one go and cannot even finish 1M SuperPi

that was just magic

these days i run low clocks...low volts in main machine..........i think processors have reached some incredible efficiency that it's not even needed for me to push it hard to squeeze everything out of it

you really think people need to prime for 8+ hours
lol i test 1M SuperPi @ every speed

i did run @ 5.0 was getting 72c in prime i'll wait till winter when it gets colder

to get this cpu n just do 4.0 is ridiculous i'm sure everyone can do it with stock volts

[dinos22]

i think it's crazy that ppl need some much speed out of thier rig to write posts on XS

i'm sitting on a 1GHz dual core laptop in power saving mode

pretty fast for me

[BeastNotro]

Ok i adjust the ghz minimum (to 4.2ghz) as i agree 99% should do 4.0ghz and if a person can't. Then its an issue with settings or other hardware or a BUM chip...

[BeastNotro]

i think it's crazy that ppl need some much speed out of thier rig to write posts on XS

i'm sitting on a 1GHz dual core laptop in power saving mode

pretty fast for me

LMAO that just cracked me up and true at dat it is...

[HDCHOPPER]

Damn, i cant prime even for 10 minutes without crashing @ 4.7GHz......And i dont want to use more than 1.4v :P :P :P
Would lowering the multiplier and upping the FSB produce significally less heat people ?

that would probbly make more heat at least from the nb after uping the volts to stable the buss and ram

for what it's wourth.....

and believe it or not .. on my board using c1e state enabled and using a higher voltage than needed is like a well the processor will only take out of the "well" what it needs to run what you set .. as opposed to hard setting the same voltage that the processor did decide to use ( under c1e) wouldnt stable out ... go figger maybe it's only the MARS mobo but eye will take it

sorry all not trying to hijack this thread just wanted to answer his post

[zsamz_]

i think it's crazy that ppl need some much speed out of thier rig to write posts on XS

i'm sitting on a 1GHz dual core laptop in power saving mode

pretty fast for me

well thats too slow for my 6 year old

[Dee]

to get this cpu n just do 4.0 is ridiculous i'm sure everyone can do it with stock volts

I wish everyone could do 4.00GHz on stock voltage, but they can't because all CPU's are different. You can only get an approximate idea as to how your week/batch will run from a database like this, but on the other hand I know people with the exact same week/batch as mine who can't get the same kind of OC's, voltage or temps I'm getting. Mine isn't a particularly brilliant chip, it's just average. Some week 20's are great, some are dogs.

Anyway, what IS stock volts? Whatever the motherboard sets it to? 1.25? 1.20? Plus motherboards over-volt, under-volt, vdrop, etc. Then consider cooling, ambient temps, and everything else, too.

[zsamz_]

I wish everyone could do 4.00GHz on stock voltage, but they can't because all CPU's are different. You can only get an approximate idea as to how your week/batch will run from a database like this, but on the other hand I know people with the exact same week/batch as mine who can't get the same kind of OC's, voltage or temps I'm getting. Mine isn't a particularly brilliant chip, it's just average. Some week 20's are great, some are dogs.

Anyway, what IS stock volts? Whatever the motherboard sets it to? 1.25? 1.20? Plus motherboards over-volt, under-volt, vdrop, etc. Then consider cooling, ambient temps, and everything else, too.

on auto @ 4.0 it was 1.16v

[Dee]

I would consider that an excellent example. Wanna sell it?

I had mine at 1.25V for 4.00GHz, and I didn't fully stress test it. Currently it's at 4.20GHz using 1.368V (measured under load with CPU-Z).

It failed Prime small FFT's at 4.33GHz using 1.392V, and I don't want to take it past 1.4V, so there ya go. I think I can cope running at approximately 4.2GHz for 24/7 daily use. I wouldn't consider 1.368V unsafe and I doubt it will degrade over time. It would be nice to have an insane clocking chip but let's be realistic. Only lucky people get them. Personally, when it comes to getting good clocking CPU's, my luck has been terrible for 4 years straight.

[zsamz_]

I would consider that an excellent example. Wanna sell it?

I had mine at 1.25V for 4.00GHz, and I didn't fully stress test it. Currently it's at 4.20GHz using 1.368V (measured under load with CPU-Z).

It failed Prime small FFT's at 4.33GHz using 1.392V, and I don't want to take it past 1.4V, so there ya go. I think I can cope running at approximately 4.2GHz for 24/7 daily use. I wouldn't consider 1.368V unsafe and I doubt it will degrade over time. It would be nice to have an insane clocking chip but let's be realistic. Only lucky people get them. Personally, when it comes to getting good clocking CPU's, my luck has been terrible for 4 years straight.

i've seen better tham mine
but i'll keep mine lol dfi board helps too nice clean volts
try tuning yout gtl settings 61-63%

[Dee]

Tell me more about GTL settings. What will tuning these do?

[zsamz_]

Tell me more about GTL settings. What will tuning these do?

helps to use less volts + higher fsb

[MrHydes]

nice thread beastNotro...

easy to compare oc's with other ppl

cheers

[Nostromo]

Q820a405

[Dee]

helps to use less volts + higher fsb

OK, that's what I thought, but I'm still not sure about settings for the P5Q-E. What do you suggest? I have north Bridge GTL and 2 other GTL settings for the CPU. I'd be happy to tweak those but somehow I doubt it will help my "below average" E8600.

[kitfit1]

helps to use less volts + higher fsb

That would be true, if you were running a quad. With both E8500's and the E8600 that i have run, any VTT over 1.17v was a complete waste of time and caused lots of instability. By that i mean the "Crossclock" settings would change between boots. So sometimes it would be stable, but more often than not it would be unstable. I have spent a lot of time trying to solve this problem with manually setting the "Crossclock" values, but all that can be done is to get them near to where they should be (which of course is no good at all).

Far bigger an aid to stability with the E8500's and E8600's, if you have it in your bios, is to set the "Slew Rates" to strong. This makes an enormous difference and more than makes up for a low VTT.

[Hamilton]

In a few years on XS, I have only seen a handfull of people claiming their CPU had "degraded" over time... Now, be Xtreme and push that chip up to 1,45V

I did run my old E6600 at 1.6V 3.87GHz for a year and it did degrade slightly...
When I sold it and it was put back on air it would only do 3.36GHz Prime95 8h stable when before it did 3.44GHz.

Currently running my QX9770 at 1.504V!!

[Dee]

Q820A670
4.25GHz
1.360V @ Idle
1.376 @ Load

More to come. I'm sure I can get this higher.

[BeastNotro]

That would be true, if you were running a quad. With both E8500's and the E8600 that i have run, any VTT over 1.17v was a complete waste of time and caused lots of instability. By that i mean the "Crossclock" settings would change between boots. So sometimes it would be stable, but more often than not it would be unstable. I have spent a lot of time trying to solve this problem with manually setting the "Crossclock" values, but all that can be done is to get them near to where they should be (which of course is no good at all).

Far bigger an aid to stability with the E8500's and E8600's, if you have it in your bios, is to set the "Slew Rates" to strong. This makes an enormous difference and more than makes up for a low VTT.

I have my FSB termination at 1.28 same as what my quad was at, but i don't think it needs to be that high ... will it kill this dual-core ???

All i know is i did 16hrs prime on it, but never adjusted it lower to try again. I am still in the process of making my template up and when i do i will add to my signature like i did with my Q6600.

By the way i updated list again in main post.

[Dee]

Try FSB Termination at 1.12. That's where mine is and it's fine.

[BeastNotro]

Thks Dee i was figuring 1.14-1.16, but if 1.12 is good i may try that, but i like priming for long hours so when i get the time to try in a day or so,i will ... I hate to have it fail less then 16hrs now...

I want to try that 1402 bios, but i waiting for more guinea pigs to try it out first.