E8600 Prime Stable Overclocking Database

[JohnnyV]

I'm also playing with it atm but it's time to go to bed.
I also did 4100mhz at that voltage but it seemed like it hung when my screensaver kicked in. Could have been a vtt setting though. Now doing 4250mhz @1,232Vcore under load and it's been going for 1hr 20m now. First attempt failed after 17m so I upped the vtt 1 notch (to 1,2v) and it seems that was the right thing to do.
No difference between idle and load is very odd. Are all tools showing that? My temps at 1,232v are around 54c while priming which is much too high I think.
I looked up a screenshot of my Q9450 and at 3800mhz with 1,376v under load my temps were also around 53c/54c. A couple of reasons for these high temps remain possible:

- The E8600 ia a hot-head
- In my eagerness to install this baby I did not mount the block properly
- The IHS is crap
- ?

[Infa]

probably couse you got a very stuck sensor on your chip, got mine yesterday, and saw one of mine was that to, its stuck at 36, and wont go lower, but it scales upwards, seems getting "good" sensors also are a bit of a luck on these new 45nm

[JohnnyV]

When testing with Realtemp core 0 has a movement of 3 and core 1 a movement of 6. I guess that's just acceptable?

[Infa]

yeah, if its stuck its stuck, sure maybe u can RMA it, but what if u get one worse? or a bad clocker, and not all of the times will the retailer or Intel replace a chip with stuck sensor, it wont go to retail if it doesnt work completely, it will throttle at given temperature, and that what it was designed to do.

accept it, TRY to RMA it, or just buy a new one, thats the options , tested mine just, movement 8 and movement 5, not so bad though, when loaded they are equal.

regards.

[JohnnyV]

I selected this cpu on batchcode so no way I'm gonna return it :P
Just gonna re-seat the waterblock and see if it makes any changes.

[Infa]

heh ok, i thought it would work to "reseat" the cooler/waterblock, its just a mental block that says, IT CANT BE BROKEN

Been playing with my 8600 now, and ONLY GTL tweaking, actually a bit of fun, when adjusting those actually DO stuff, and not little , i began adjusting them both from 63x , and then one at a time 1-2 notches, then priming over and over, and see where it breaks, if it can to 8k pass and 10k pass, if do set it to 12k fom start, and reduce testing rounds to 5, to quicker see where it fails... interesting testing, and make me understand it a little more.

[JohnnyV]

Can you elaborate a bit more on that? I'm priming atm as well and noticed that from 4250 to 4500 it needed a lot more juice. Perhaps by tweaking the GTL's the Vcore can be lowered?

[Pisklink]

i can say that it looks good, and 4hours are good, and for ME thats enough! , but you may have to go mem test too, like OCCT RAM , or Blend in Prime, to make sure all is ok.

but why are you goin 5:8 divider? Use 1:2 , the highest one, that would be about 1850, and you can go 2.06-2.1 for that CL7, should work fine.. and put 1T on, or does the ram not like that?

Last post Offtopic here! head on in the P5e3 thread instead, people will get pissed here eventually :P

Your right we should continue in the P5E3 thread. Therefore I will copy your last reply into that thread. Ok? Would you please respond to my answer there? Thanks in advance.

@ JohnnyV

You should take a look at the GTL sticky thread, here at the Intel section. There is a lot of explanation going on about GTL settings. Maybe it can help you understand GTL settings if that's what you want to know.

[Dee]

Thwe GTL thread hasn't helped me at all. I still don't even know if I should be increasing the value or decreasing it.

[Infa]

me either, im just testing, and right now everything is out the window, in the beginning, adjusting GTL made diffrence in how it failed, but now, Core 1 is just failing all the time, nomather what setting, god "#¤""¤ , first when i popped the cpu in, it could prime atleast 5 mins at 1.37, now it fails 10s with 1.40 on core 1, stupid computer parts, i NEVER have luck even if i beleive so... but if it want 1.43 for stable 4,5 its not BAD its just over avarage what people have, my E8500 would barely keep it 4,2 at 1.45, 1-2 hours stable.

BUT ANYWAY, its really annoying though...

pisklink, ok

johhny , you have to start somewhere, like, putting both GTL at a x63 value, and then test it, do custom test with prime, shorten the test time to 5 mins instead of 15, and let it go on.

[Infa]

Ok, Im having trouble with prime and orthos, same shajjt diffrent names actually? but Doing OCCT dont fail, a program called CPU stability dont fail, and RUNNING Intel Burn test DONT FAIL and thats stressing lot more than the first named, so whats up? prime are foucked? using vista BTW, to late to install XP, but will do that tomorrow, this sounds weird.

[Infa]

Ok, now prime has gon for 2:30 mins, almost "validated" and i didnt change ANYTHING, just did a 5 pass of linpack, and then everything went flawlessly, strange thing this prime actually, can recall it being this strange before actually, it just freaks out after a while it seems, or its vista, Vista is lovely in every way, havent failed me yet, maybe just doesnt fit with all this "older" and benching programs? Lot of status reports of me, but where do i write it then, im just such a person, and havent got alot of friends that are computer freak like me? no one share to talk haha?

[Rattle]

4.4 17 hours, 500x9 is only stable 8 hours, but gaming flawlessly and even passes intel burn test, need more tweakage.

[mrcape]

Anyone know what batches are shipping from the egg recently?

[JohnnyV]

Batch: Q822A435 (yes another :P )
Vcore under load: 1,328

[flesheatinvirus]

Anyone know what batches are shipping from the egg recently?

I just got a 8600 from batch 820A324 from the Egg yesterday.

[thomas hobbes]

Q822A435
Vid = 1.1625v

4.3Ghz & 8 Hours Orthos Stable here. This is the sweetspot for temps and voltages for me, particularly as I'm running on air at the moment.

[dengyong]

I just got a 8600 from batch 820A324 from the Egg yesterday.

Same here a week ago from the Tennessee warehouse.

[jaredpace]

Q822A435
Vid = 1.1625v

4.3Ghz & 8 Hours Orthos Stable here. [/URL]

55C & 1.256v @ 4300mhz... Nice!

[mrcape]

I just got a 8600 from batch 820A324 from the Egg yesterday.

55C & 1.256v @ 4300mhz... Nice!

Thanks for the quick replies guys! No idea how that batch is, but I've had 4 820s so far - all great on air, but not so great under LN. By any chance have you guys pushed them under LN? I'm looking for 6.3ghz+ 3d bench stable, it may be a long and impossible hunt. Your info helps me already though.

What are you showing for a vid on that batch?

[jaredpace]

1.1625v ^^^

[uclajd]

So it looks like these are averaging like 4400 stable on stock volts with air?

How much better do these seem to OC than 8500's?

TIA. Please excuse my reluctance to read all 8 pages.

[mrcape]

Woah Jared that's ultra low vid. That may be the ultimate air/water cpu. Take care of it and keep it for your main rig. I got a nice Intel RMA 820a719 that is a great air chip.

It's funny how these are so different. Some of them that are horrible on air (max 4.8ghz or so) are getting 6.7+ under LN. My max under LN is 6.3ghz.

[DemonEyez]

I selected this cpu on batchcode so no way I'm gonna return it :P
Just gonna re-seat the waterblock and see if it makes any changes.

Just like Infa says, the tempsensors are stuck its a known problem on these 45nm chips, I got the exact same stepping as you and have the same problem. Movement of Core 1 is 0 and Core 2 is 3. My block is fitted right and makes even contact. I had these problems before with another E8400.
I mean even at 10x333 with 1.1v i get 41 degrees idle. And then when I Prime is stays at 41 degrees. So bad temp sensors, but great stepping.
Wouldn't wanna RMA it in a liftime :P

Can you elaborate a bit more on that? I'm priming atm as well and noticed that from 4250 to 4500 it needed a lot more juice. Perhaps by tweaking the GTL's the Vcore can be lowered?

I noticed that even when I set 1.25vcore and 1.35vTT i still couldn't do a SuperPi run 7x600. I tried lower and higher voltage, different mem dividers all kinds of GTL, nothing seemed to help. Then I changed FSB Ref which was set to 2A nd changed it to 29. Lo and behold, I can run SuperPi stable and also with lower voltages.
So you good give that a try.

[flesheatinvirus]

Thanks for the quick replies guys! No idea how that batch is, but I've had 4 820s so far - all great on air, but not so great under LN. By any chance have you guys pushed them under LN? I'm looking for 6.3ghz+ 3d bench stable, it may be a long and impossible hunt. Your info helps me already though.

What are you showing for a vid on that batch?

My vid is 1.26. This ones going on its own water loop first and if things look good perhaps some LN.