Retail E8600 Overclock, Temps, Settings Thread

[Infa]

Heh thanks probably would get it higher though if i put more juice in it :P

[Ozzfest05]

can someone tell me if this batch number is good Q820aA405 SLB9L just got my E8600 today

[Jupiler]

can someone tell me if this batch number is good Q820aA405 SLB9L just got my E8600 today

Quick search here at the forums :

http://www.xtremesystems.org/forums/...2&postcount=43

[Counterpoise]

Pll 1.52
VTT 1.16
NB 1.26-1.28

mobo: p5Q-deluxe (ket 1406)
cooling: TRUE Copper lapped + slipstream fan
psu: pcpc quad 750w

[LagunaX]

Wow Counterpoise - what is that batch #?

[Infa]

woooot now that must be THE best one ever seen?

[OC4/3]

@Counterpoise,batch #??Can you hit 5.1ghz+ for suicide?

[carepolice]

@Counterpoise,batch #??Can you hit 5.1ghz+ for suicide?

I really belive so, much headroom left even for aircooled oc

[Counterpoise]

Hello guys, I believe 5.1Ghz would be extremely difficult to do without going over 1.45v which I'd rather not go for now as the chip is new. Chips don't really scale linear or as smooth as one would hope. At 4.7Ghz I'm seeing all kinds of weird things when priming, such as everytime I prime it either can fail w/in 5 min or can one time last as long as 2 hours, and it never does the 'insta-restart" of the PC anymore like it did at lower speed. There's really no pattern anymore.
I did boot up at 5.0ghz to do a superpi run though. The time is slow for the Ghz because my FSB is (relatively) low and just using 1 stick of ram w untweaked timings. And at that voltage windows was on the edge of instability and superpi run was barely stable.

[DJSUB]

@ Counterpoise

can you post your complete batch number?

thanks

[philbrown23]

yes batch # please?

[Ozzfest05]

Just did a quick test with E8600 and it looks decent havent messed around with volts may go lower like I said quick test on air. Oh Yeah Asus probe is reading CPU PLL wrong cpu z seems to mess up figures with it actual CPU PLL is set Standard 1.51006

[Ozzfest05]

Did some testing lower volts and seems fine so far no time left tonight for testing will hit the chip again tomorrow

[Counterpoise]

@ Counterpoise

can you post your complete batch number?

thanks

It's Q822 something, the rest I don't remember but I remember seeing it on this forum as a really good batch# to have. The rest I'm gonna have to get to you later since I left the box at my parents' house.

[Counterpoise]

Took forever to tweak the settings to get 4.7 stable.

[The Viper]

really awesome Counterpoise! Those week 22 chips are amazing, Im guessing yours probably is a Q822A435. What setting did you need to tweak to get stability at 4.7?

[LagunaX]

52c load @ 4.7ghz?!? Wow those TRUE coppers must be amazing!
I get 65c load @ 4.5ghz @ 1.3v with a regular TRUE/washer/lapped cpu!

[Counterpoise]

Yeah that batch sounds like it.

Well, at 4.6ghz all settings was easy and had a big margin. at 4.7Ghz, I had to have a specific PLL and VTT for the prime to not error out after 30-90 minutes. VTT I have at 1.22, PLL I have at 1.56. Change any of these 2 numbers to something else and it'll error out after a while. At first I didn't believe PLL and VTT did anything for dual cores under 500mhz FSB, but I guess it does now. NB is at 1.30 and that voltage isn't important for the Gromacs. LLC is enabled. all the REFS are left at auto though, because they don't seem to do anything, for dual cores at least, but I may be wrong again just like the PLL and VTT.

[Retro]

Took forever to tweak the settings to get 4.7 stable.

I had mine up to 4700 as well, needed about 1.475 volts on the v-core. Orthos still crashed in about two seconds

[BranNutz]

Yeah that batch sounds like it.

Well, at 4.6ghz all settings was easy and had a big margin. at 4.7Ghz, I had to have a specific PLL and VTT for the prime to not error out after 30-90 minutes. VTT I have at 1.22, PLL I have at 1.56. Change any of these 2 numbers to something else and it'll error out after a while. At first I didn't believe PLL and VTT did anything for dual cores under 500mhz FSB, but I guess it does now. NB is at 1.30 and that voltage isn't important for the Gromacs. LLC is enabled. all the REFS are left at auto though, because they don't seem to do anything, for dual cores at least, but I may be wrong again just like the PLL and VTT.

I thought I had mine stable too but it was crashing after 1 hour of P95, I tweaked the gtlref settings from auto to -60 and it wouldnt even boot, then to -50 and it ran P95 for 3 hours before it crashed. So I dropped it to -45 and now no crashes so far. So yeah they do make a difference for stability on dual cores. I run 4.5Ghz 24/7 on air with 1.368Vcore, and 1.35VTT, 2000FSB linked and synced with P1/P2 enabled.

I am willing to bet you can get that chip stable at lower Vcore if you set GTLref correctly.

[Retro]

I thought I had mine stable too but it was crashing after 1 hour of P95, I tweaked the gtlref settings from auto to -60 and it wouldnt even boot, then to -50 and it ran P95 for 3 hours before it crashed. So I dropped it to -45 and now no crashes so far. So yeah they do make a difference for stability on dual cores. I run 4.5Ghz 24/7 on air with 1.368Vcore, and 1.35VTT, 2000FSB linked and synced with P1/P2 enabled.

I am willing to bet you can get that chip stable at lower Vcore if you set GTLref correctly.

I've got good stability at 4400MHz with my current settings. I wondered about your suggested GTL Ref settings, my DFI board has the three settings for:
1.CPU GTL 0/2: currently set at 63
2.CPU GTL 1/3: also set at 63
3.NB GTL: set at 61
So, on your board, do you have all those values set at 45?

[carepolice]

I've got good stability at 4400MHz with my current settings. I wondered about your suggested GTL Ref settings, my DFI board has the three settings for:
1.CPU GTL 0/2: currently set at 63
2.CPU GTL 1/3: also set at 63
3.NB GTL: set at 61
So, on your board, do you have all those values set at 45?

i belive that his board got different GTL adjusting "table"

[Counterpoise]

I thought I had mine stable too but it was crashing after 1 hour of P95, I tweaked the gtlref settings from auto to -60 and it wouldnt even boot, then to -50 and it ran P95 for 3 hours before it crashed. So I dropped it to -45 and now no crashes so far. So yeah they do make a difference for stability on dual cores. I run 4.5Ghz 24/7 on air with 1.368Vcore, and 1.35VTT, 2000FSB linked and synced with P1/P2 enabled.

I am willing to bet you can get that chip stable at lower Vcore if you set GTLref correctly.

Did you move all your GTL REFs at the same time or one at a time?

[BranNutz]

Sorry, my board has 4 settings all for CPU only, GTLref lane 0-3
I set lane 0 and lane 1 to -45Mv my board doesn't use the multiplier like yours, it uses a voltage offset. I left lane 3 and 4 on auto as I've heard they are not used for dual cores. You need to calculate your gtlref differently than I do. You could use my number and work the equation backwards to find the number you need. GTLref settings will change when you change your VTT also.

VTT * 0.667 + Y = VTT * 0.635

my VTT is 1.35v
so for me Y= -43.1

my board only goes in 5mv increments +/- so I set it for -45 and it seems to be alot more stable so far.

[Counterpoise]

so you set both lanes to -45. on my board it suggests that the second lane be set to 0.670 and first lane 0.630. So I'm unsure as to set both of them to equal values, ie both 0.625, or decrease the same amount from each lane, ie 0.625 and 0.665