E6600 Core Voltage

[DissonicEnt]

Hello everyone...I'm new sorta...
I've been poking around the forum for a while, though I have never posted.

Anyway...
I own an c2d e6600 on a asus P5B-DLX wifi
Now I've been researching OC results for this CPU/Mobo combo for some time now.
Several forums/tech sites have assembled databases of these results sorted by CPU mfg date & stepping, aswell as core voltage and mobo used.
I currently have the system @ 400x9 = 3600MHz @ 1.56 vCore.
1.65v NB, 1.4v NBt.
What I've notticed on other systems with identicle configs is that they are able to have similar OC results using MUCH less vCore. I CANNOT go lower than 1.55v without serious instability. Yet I see guy x running at 425x9 @ 1.35v core....I'm confused!
My concerns are the additional heat produced and ultimately the CPU's health and durability. With a Zal 9500 it won't break 55C loaded, however I'd like to run at a lower voltage if possible.
The only systems I've seen with a 1.5+ vCore have been uber liquid cooled and/or those netting 4GHz and up.
Can one of you folks please explain to me why my chip may need so much more voltage than others? And what is the MAX voltage that can be applied to the e6600 safely (provided that thermals were optimal). The P5B BIOS goes to 1.75v...though I'm sure that's for the 90nm Preslers/Prescotts and EEs.

Much Thanks in advance

 

[DissonicEnt]

BTW:

It's a B2-6 step CPU.

thanks

 

[LouPoir]

That is a nice overclock. Temps look good. I would also be concerned with the vCore that high even though the temps are OK.

IMHO

I have my E6600 running at 3.4 with 1.49 vCore running much cooler - 35c idle and 50c load on air - BigTyphoon Cooling.

Lou

 

[Noubourne]

Is CoreTemp reporting those temps or what?

56C seems low for the amount of volts you're pushing.

I'm not uber-familiar with C2Ds, but I would question the source of your temps. You should be able to get a temp for each core. If real core temps are going over 60C, I'd think about lowering my OC. Ideally close to 50C or below if possible.

3.6Ghz is nice though. If you had water you might be able to do that more comfortably temp-wise...

 

[DissonicEnt]

I've verified temps against readings from Intel TAT, Core Temp, PC Wiz, and PC Probe.
All within 3 deg of each other: 53 - 59C Loaded.
Idles around 42C.
System is very stable though.
Backing down to 1.50 requires a major step down to 360x9 = 3240 to run stable...but much cooler however.
I'll back off if you guys think 1.56 is too much...I don't want to damage the chip...or shorten its life.

thanks again

 

[Duvie]

Well they are obviously not Orthos dual prime temps...

At 3.4ghz (7x486) for me with only 1.4625v I will hit 59-60c...

My E6600 at 7x466 at 1.45v (1.44v actual) will hit 60c as well...

Both with Ninja Scythes and 120mm fans

I have riefly tested the E6600 at 3.4ghz with 1.525v set (1.5v actual) and I was hitting 67c

 

[bjp999]

With lapped HSF and CPU, I am running my E6600 at 400FSBx9 at 1.5625v with TAT temps maxing at 59C when running dual Prime95. Your temps seem pretty close to that, but without the extra work on cooling, I would be surprised that you're really that cool unless your ambient is dowright cold! I also have a Scythe Infinity with is reportedly a better cooler than the Zalman.

What are you using to apply load? If you are running dual Prime95, double check the affinity settings and run large TFTs. Or run Orthos which runs the dual prime95s for you. If you're running something else, switch. My guess is that once you really load up the cores, you'll see your temps head north of 65C (68C - 72C would be my guess). I hope I'm wrong for your sake, but that's what everyone is seeing at that FSB, vCore, and good-but-not-excellent air cooling.

The "B" stepping chips have a reputation for very good OCs. If you're really stable at 400x9 without lapping or hours of mucking with settings, there's a good chance that you could hit 3.7 or 3.8 stable with some effort.

To answer your questions about why others have faster chips, its just the luck of the draw. Also realize that many people post results on rigs that are just barely stable enough to run a benchmark, and not able to run dual Prime95 for more than a couple of seconds. People also exaggerate so that they can say "me too" with the fastest posters. If you worked in a computer store you might have an opportunity to personally sample dozens of chps and cherry pick the one that OCs the best. That would be an excellent way to get a great chip!

From all you've said, you have an excellent Conroe. Work with it and you'll get great results.

Good luck!

 

[DissonicEnt]

Thanks for the input and advice folks.

Using Prime95 to burn.
Now I understand the need to beat the hell out of it just to see how high it'll go....but I have yet to throw an actual application or task at it that it couldn't handle.
Runs many a game (oblivion, fear, q4) and loves to transcode DVR-MS files to MP2.

I have a Tuniq on order...should be here by Wednesday.
If that don't work I'll fab my old Swifty liquid cooler from my Gallatin system to do the job.

Thanks again.

BTW: I've backed the multi down to 8...while keeping @ 400 FSB = 3200
Allows a vCore drop to 1.45...much more comfortable here.

 

[DissonicEnt]

Also...
Still haven't unlocked the mystery of guys being able to run high (i.e. 425 x9)
with less than 1.4v on core...same chip same make same stepping same board.
they must have gotten a primo sample. They're running naturally aspirated too.
Infinity or Tuniq in most cases...but a few claim to be running 9500s...
Any thoughts?