E8400 @ 4.4GHz on air and still going!!!

cornelious0_0

[H]F Junkie
Joined
Apr 6, 2003
Messages
12,783
Just wanted to share some of my latest success with my fellow [H] members.

After having my Shuttle die on me I was left to rebuild my system which in the end turned out to be a blessing in disguise.

The system is listed in my sig but clock speeds are always changing and being updated. Today after a few hours of tinkering I got the system fully benchmark stable @ 4.4GHz and booting into windows at 4.5GHz on air using my Ultima-90 and a single 40cfm 92mm fan!!!

I plan to swap out the GTX260 for a GTX280 in the next few months and I'll be building a 120.2 watercooling loop for the CPU in the next couple months as well. I'm really hoping to be able to push the E8400 a bit further then 4.5GHz for benchmarks and hopefully get 4.4GHz or 4.5GHz stable for 24/7 use...and at this rate I don't have any doubts in my mind that I'll have many issues.

I'll have pics posted of some benchmark results at 4.4GHz /w the GTX260 overclocked before the end of the night, just wanted to let you guys know that all is well and that I'm just getting started with this beast of a CPU. :cool:
 

fromeo85

Guest
Joined
Mar 8, 2008
Messages
2,364
If you are currently running 4.25 @ 1.42v and you got it stable at 4.4, what was the volts for that? You may want to be careful, if you start pushing 1.5v with that chip its going to end up killing it.
 

Zero82z

Fully [H]
Joined
Jan 20, 2004
Messages
27,897
Exceeding 1.4V is not a good idea with that CPU. It may end up dying a lot faster than it normally would.
 

cornelious0_0

[H]F Junkie
Joined
Apr 6, 2003
Messages
12,783
To benchmark at 4.4GHz and boot at 4.5GHz does require just under 1.5v, nothing to plan on running on air for months or years on end, but also not nearly enough to actually kill an E8400 in a few weeks or months of use.

I know its not the safest thing to do, but I've seen other ppl run higher voltages on their 45nm chips for long periods of time so I'm not TO worried for now.

Gotta live a little right? :p
 

GonzoP

2[H]4U
Joined
Jul 3, 2006
Messages
3,660
To benchmark at 4.4GHz and boot at 4.5GHz does require just under 1.5v, nothing to plan on running on air for months or years on end, but also not nearly enough to actually kill an E8400 in a few weeks or months of use.

I know its not the safest thing to do, but I've seen other ppl run higher voltages on their 45nm chips for long periods of time so I'm not TO worried for now.

Gotta live a little right? :p

That's good stuff. I just made the choice myself to take my own E7200 which I had at 3.6Ghz and 1.32v up to 3.9Ghz and 1.41v. That's as high as I can or wanna run it. I can post at 4Ghz but vista does not quite make it. 1.4v is also where I consider there to be a yellow zone, 1.5v is certainly the redline zone. All in all if you end up at 4.3Ghz at 1.4v that's decent
 

cornelious0_0

[H]F Junkie
Joined
Apr 6, 2003
Messages
12,783
Yep, 4.2GHz on air takes a smidge over 1.4v, and that is where I leave it for 24/7 use, gaming included...which is why I'm not to worried about damaging anything, 1.5v is only used to push me to 4.4GHz and beyond for short periods of time to benchmark.

Typically with a decent h20 setup and (resulting) lower load temperatures it is possible to get by on slightly less voltage at almost any clock speed. It is because of this that I can't wait for the watercooling, I shouldn't need as much voltage to get 4.4 and 4.5GHz stable and I should be able to benchmark at closer to 4.6 or 4.7 if things keep going how they have been.

Cross your fingers for me guys, I might actually have a half decent chip to play with for once. :D
 

cornelious0_0

[H]F Junkie
Joined
Apr 6, 2003
Messages
12,783
Here's pics for you guys, I would say they're my max scores at the moment, but I haven't tried overclocking the GTX260's memory or pushing the E8400 to a clock between 4.4GHz and 4.5GHz.

3DMark2001:
44ghz72520003dmark2001gq6.jpg


3DMark03:
44ghz72520003dmark03yb0.jpg


3DMark05:
44ghz72520003dmark05qt7.jpg


3DMark06:
44ghz72520003dmark06eh0.jpg


Aquamark3:
44ghz7252000aquamark3mr4.jpg


SuperPi 1M:
44ghzsuperpi1mqv7.jpg
 

Matthew Kane

Supreme [H]ardness
Joined
Dec 1, 2007
Messages
4,233
Nice....the new c2d 8000 series are know for their light ease of reaching high overclocks...I read in my atomic mag that the E8600 is able to get 4.7GHZ on air no probs with a TRUE. And a month ago, they a did a test on the E8400 which was able to reach a 450MHZ FSB and it was even higher when they watercooled, to phase and single stage and then to DRICE and LN2. Another quick look at your current voltage...isn;t that a bit too high...I mean I think its 45nm c2d's that don't like 1.45v or higher or else they can easily die or is this only necessary for quads?:confused:

I like your XP theme, what theme is it..link?
 

cornelious0_0

[H]F Junkie
Joined
Apr 6, 2003
Messages
12,783
Another quick look at your current voltage...isn;t that a bit too high...I mean I think its 45nm c2d's that don't like 1.45v or higher or else they can easily die or is this only necessary for quads?

I like your XP theme, what theme is it..link?

The voltage is a little high, technically above 1.4v to 1.45v CAN be considered a little high for 45nm chips, but for 24/7 use I have it running at 4.2GHz /w 1.42v. I only bump the CPU up closer to 1.5v for benchmarking, maybe a half hour at a time, so I'm not worried about killing it, especially not when I see people running 24/7 @ 1.5v for months without a single issue.

If I do end up putting together a watercooling loop for the CPU I'm definitely gonna be aiming for 4.7GHz or so for benching and hopefully about 4.5GHz for 24/7 use with less then 1.5v. It should be exciting considering how well this E8400 is doing on air.

The XP theme I got from a friend years back, can't remember for the life of me where it came from. :p


Something else I've been thinking about doing is swapping out my Ultima-90 for my TRUE. The TRUE is the better cooler, but I would have to mount it horizontal as opposed to being vertically alligned with the rear case fan. There's still a good chance that mounted horizontally my temps would still come down, I just don't know if it's worth the time and trouble (especially if I'm gonna be doing water) considering the TRUE only beats out the Ultima-90 by a few degrees.
 

Matthew Kane

Supreme [H]ardness
Joined
Dec 1, 2007
Messages
4,233
Its "that" after a couple of months the chip dies away........

Have tried to lower the voltage a bit down to see if it still makes the system run stable at that speed?
 

cornelious0_0

[H]F Junkie
Joined
Apr 6, 2003
Messages
12,783
Its "that" after a couple of months the chip dies away........

Have tried to lower the voltage a bit down to see if it still makes the system run stable at that speed?

I have tried lowering it and this is as low as it'll go at this clock.

Again, I leave the system at 4.2GHz 1.42v 24/7 so there's really no issue of degradation, I can't get it to benchmark at 4.5GHz on air so I'm just going to leave it until I setup my watercooling. With the watercooling I should be able to lower my CPU voltage and get to the point where 4.4 or 4.5GHz is stable at close to 1.4v.

I like taking chances with my hardware and pushing it to it's limits, but I'm not dumb about it...don't worry guys. :D

I think the theme is Royal Noir - if not Royal Noir is very close to it. couple of versions out there.

http://www.google.com/search?hl=en&q=royal+noir+theme&aq=f&oq=

That's not it, but I actually prefer Royal Noir to what I've been using...thanks for the link! :)
 
Joined
Feb 8, 2005
Messages
3,343
I have tried lowering it and this is as low as it'll go at this clock.

Again, I leave the system at 4.2GHz 1.42v 24/7 so there's really no issue of degradation, I can't get it to benchmark at 4.5GHz on air so I'm just going to leave it until I setup my watercooling. With the watercooling I should be able to lower my CPU voltage and get to the point where 4.4 or 4.5GHz is stable at close to 1.4v.

I like taking chances with my hardware and pushing it to it's limits, but I'm not dumb about it...don't worry guys. :D



That's not it, but I actually prefer Royal Noir to what I've been using...thanks for the link! :)

How does better cooling allow you to lower the voltage? I would think that the same voltage is required to push 4.4GHz whether the chip is at 50C or 40C. Water will allow you to add more voltage while keeping your temps down, but it won't give you the same clocks at a lower voltage. Just at a lower temperature.
 

Sphere

2[H]4U
Joined
Apr 13, 2005
Messages
2,353
Going to water will not change your voltages.

Edit:-=Antimatter=- beat me to it.
 

cornelious0_0

[H]F Junkie
Joined
Apr 6, 2003
Messages
12,783
How does better cooling allow you to lower the voltage? I would think that the same voltage is required to push 4.4GHz whether the chip is at 50C or 40C. Water will allow you to add more voltage while keeping your temps down, but it won't give you the same clocks at a lower voltage. Just at a lower temperature.

I disagree...I have personally setup a couple watercooling loops and had it happen. The lower the temperature of a CPU it can typically run at slightly lower voltages.

I'll look into it by the slight chance that I'm completely out to lunch, but I'm positive that the last time I setup a watercooling system for an overclocked computer I was able to drop my CPU voltage as well as the load temperatures, I have always had the two go hand in hand.

I'm far from wanting to start any sort of argument, just saying that I could have swore I had it happen before.
 
Joined
Feb 8, 2005
Messages
3,343
I disagree...I have personally setup a couple watercooling loops and had it happen. The lower the temperature of a CPU it can typically run at slightly lower voltages.

I'll look into it by the slight chance that I'm completely out to lunch, but I'm positive that the last time I setup a watercooling system for an overclocked computer I was able to drop my CPU voltage as well as the load temperatures, I have always had the two go hand in hand.

I'm far from wanting to start any sort of argument, just saying that I could have swore I had it happen before.

How big was the drop in voltage, though? I doubt it was significantly less. The thermal condition will force you to drive voltage higher as temps climb, but frequency is the key here. Past a certain point (say 4.0GHz on a wolfdale) you won't be able to lower voltage much at all and keep the same clock, no matter what cooling you have.

Anything above 4.4GHz for me is just too uncomfortable with my cooler, and you really have to start pouring the voltage on above 4.2. You've got one heck of a chip there, and I look forward to your water results :)
 

cornelious0_0

[H]F Junkie
Joined
Apr 6, 2003
Messages
12,783
How big was the drop in voltage, though? I doubt it was significantly less. The thermal condition will force you to drive voltage higher as temps climb, but frequency is the key here. Past a certain point (say 4.0GHz on a wolfdale) you won't be able to lower voltage much at all and keep the same clock, no matter what cooling you have.

Anything above 4.4GHz for me is just too uncomfortable with my cooler, and you really have to start pouring the voltage on above 4.2. You've got one heck of a chip there, and I look forward to your water results :)

Maybe it's a difference in the way Wolfdale's respond to higher clocks/voltages and cooling beyond air, I'm not sure.

If beyond 4GHz (the point I'm already at on air) I'm not going to be able to drop my voltage much at all with watercooling then I might have to think twice about bothering with it in the first place.

It was always much understanding (as you actually just said) that you are typically forced to raise voltages as temps climb, I guess I was hoping that the lower load temps under water would make it possible for me to climb higher on 1.45v or so.

I guess it might have been the wrong way to say it when I said that watercooling would drop my voltages. I guess what I'm getting at is that because you do have to continue to add voltage as temps increase I was hoping to be able to get farther on my current voltage.

I have seen other E8400's hit higher clocks on lower voltage and even though YMMV still applies, who's to say that my required voltages on this specific CPU aren't partially due to temperature and that I might be able to climb a bit higher with a good h20 loop?

I think I'm still gonna stick with the plan to watercool this thing, if my lower temps don't loosen the voltage requirements for my OC then so be it, and if they do then that's great.

Thanks for the input everyone...nothing wrong with discussion in the down time.
 

cornelious0_0

[H]F Junkie
Joined
Apr 6, 2003
Messages
12,783
This is kinda what I was talking about..granted the review below is using a Vapochill phase change system on a s939 Opteron, but the same principle applies, cooler temperatures allow for higher overclocks and/or lower voltages required at the same OC when compared to air cooling.

http://icrontic.com/articles/asetek_vapochill_ls/10

I'm obviously not going to be seeing temperatures like the VapoLS, but a drop of 15-20c in load temperatures will mimick the results in the review, just to a slightly lesser extent.

I can't wait to get watercooling running on this thing, I'm still not entirely sure what order I'm going to setup the case mods/watercooling but when it happens it's gonna be sweet.
 

Matthew Kane

Supreme [H]ardness
Joined
Dec 1, 2007
Messages
4,233
The usual term used. Add more power and you will see a temperature rise. Get better cooling will lower the temps and gives you the change to lower power to gain better heat control. Not always the case. But has worked a couple times for me with the old days of Pentium overclocking. I might have to agree with both Cornelius and Antimatter for this one.

But I still want to know that XP Theme god damn it lols.:D
 

vsboxerboy

2[H]4U
Joined
Oct 17, 2005
Messages
3,661
Man, I'm sad if stuff like this doesn't exist in the Core i7 era.

Nice work, and at the prices of those chips now a days, it wouldn't be horribly disastrous if it did happen to fail before it's expected lifetime. Wouldn't be great either though!
 

cornelious0_0

[H]F Junkie
Joined
Apr 6, 2003
Messages
12,783
Regular Ultra-120 with a single yate-loon high-speed.

Sure does make me jealous...with I had an E0. :(

Awe well, as long as I'm above 4.5GHz on water I'm fine regardless of the revision, the lower voltages would sure be nice though.
 

CookiePuss

Limp Gawd
Joined
Dec 14, 2001
Messages
417
These e84-8600 chips are awesome.

Got this e8400 for 150 and its running 4.25 for me stable way under 1.4 volts.

I have booted into windows at 4.5ghz over 1.4vcore, but crashes quite quickly once there.

pic
 

ballistic

Gawd
Joined
Feb 8, 2002
Messages
743
Hey Cornelious,
I am glad to hear this build is running smooth.
I hated what happened to your Shuttle build, it was looking cool.
 

cornelious0_0

[H]F Junkie
Joined
Apr 6, 2003
Messages
12,783
Hey Cornelious,
I am glad to hear this build is running smooth.
I hated what happened to your Shuttle build, it was looking cool.

Don't you worry, once I'm done with the TJ08 it's gonna look like the Shuttle's bigger brother. I'm doing a few of the same mods to pay tribute to the system that never was along with a few improvements.

Gonna be one sexy beast once the case mods and watercooling is all done.
 

cornelious0_0

[H]F Junkie
Joined
Apr 6, 2003
Messages
12,783
wow I'm impressed!:eek:

As was I, damn E0 chips take so much less voltage...makes me jealous. :(

Funny to see how much extra voltage these chips start to need once you hit that 4.4GHz range, CookiePuss and myself need very similar volts to hit 4.5GHz but E0's do much better up to about 4.2GHz on air...which is where most people end up for day to day use.
 
Joined
Sep 18, 2008
Messages
755
Im quite impressed, i've always been to scared to stray to far from stock voltages for fear of damaging my chip ;)
 

GonzoP

2[H]4U
Joined
Jul 3, 2006
Messages
3,660
As was I, damn E0 chips take so much less voltage...makes me jealous. :(

Funny to see how much extra voltage these chips start to need once you hit that 4.4GHz range, CookiePuss and myself need very similar volts to hit 4.5GHz but E0's do much better up to about 4.2GHz on air...which is where most people end up for day to day use.

Damn I never noticed the first time yours isn't an E0. Even more impressive
 

amit_talkin

Limp Gawd
Joined
Jul 30, 2008
Messages
180
I have my E8400 E0 running at 4.05 GHz @ 1.25v, 24x7 stable.
Prime max temp. : 54C ( after 2 hours of test ).
IntelBurnTest max. tmep : 68C ( after 5 tests ).
091348cc0a5ce2b63.jpg


however it takes almost 1.43v to reach 4.5 GHz. but i dont wanna kill my CPU for just 450MHz extra. 4 GHz is enough for me :).
 

GonzoP

2[H]4U
Joined
Jul 3, 2006
Messages
3,660
How about 4.2Ghz on 1.3v ;)
That's the thing with the EO's. They seem to hit 4Ghz with way less voltage than previous revisions but once you get into the 4.3, 4.4Ghz neighborhood they really need the juice
 

amit_talkin

Limp Gawd
Joined
Jul 30, 2008
Messages
180
just tried 4.27 GHz on 1.31v. tested with orthos for 30 min. no errors.
427GHzstable.jpg


will post more results if I will go for 4.3 n more.:)
 

cornelious0_0

[H]F Junkie
Joined
Apr 6, 2003
Messages
12,783
Even if the E0's need a "healthy" amount of voltage around 4.4GHz+ itd still be nice to have a 24/7 OC of 4.2GHz or so at lower volts.

I found the fan clips for my Ultra 120 eXtreme and I'm gonna use it with my Scythe "F" in place of my Ultima-90 and see how the temps react.
 

cornelious0_0

[H]F Junkie
Joined
Apr 6, 2003
Messages
12,783
Technically, 30 minutes of Orthos isn't stable. You want to do it for 4 hours+

I agree, but I think all he was getting at was that it's a preliminary level of stability. I'd take pride in the fact that an OC is stable for 30 minutes of Orthos as it means I can use it to benchmark.

Damn I really wanna setup my watercooling now and see what this C0 E8400 can really do!!! I didn't realize at first that I was doing as good as I was, most results you see for E8400 OC's are of E0 chips...at least I'm not doing awful and there's still hope for my baby. :D
 

Mav451

Supreme [H]ardness
Joined
Jul 23, 2004
Messages
4,727
Maybe it's me getting older, but finding a setting that passes x hours of Stress Test doesn't seem to be enough anymore. I'm more comforted when I have a 10-15Mhz higher FSB that also still passes.

E.g. my 1.296 for 4Ghz turned out to be fool's gold (that or I had slight degradation). I tested up to 9 x 451 back in Feb 2008 and thought it was "good enough" after several hours of OCCT. Last week, after a few hours of L4D caused it to hard-lock. Mind you I've played hours of TF2 on this OC (Feb through August), but again, I haven't played much TF2 the last few months - so who knows.

I bumped my voltage 2 steps to 1.312 (Windows) and I find I can OCCT up to 458Mhz for several hours. But in my impatience I jumped to 464 for my next test, and that failed almost instantly. So I technically only have 13Mhz of "safety factor". Am I the only paranoid one here? Hrmm.
 

ATLPIMP

[H]ard|Gawd
Joined
Jun 27, 2002
Messages
1,124
I disagree...I have personally setup a couple watercooling loops and had it happen. The lower the temperature of a CPU it can typically run at slightly lower voltages.


I'm with you on this one... I'm thinking h2o allowed me a drop of about 0.05V, can't remember exactly. Also realizing h2o to allow a higher stable over clock with out voltage increase (same thing).
 
Top