**Official e8400/e8500 Retail OC Thread

[Tancoll]

I'm going to try out mine when i get time.

[bonzo_catalin]

Don't you have 10 minutes ?

[CERO]

Wish i knew, i think they're about equal.

How's yours?

Hi Zeus , didnt test it yet, still debating with myself untill tomorrow if I should test it or get another 1.

[karateo]

Anyone with Q750A ?

Me!
Able to post and boot (hangs before win logo)
at 4500 with 1.38 (maybe can do as well with less)

[ghost_recon88]

Appears the X38 boards are handling these chips a lot better than the P35 boards are, at least the voltage necessary to hit certain clocks vs. the voltage necessary on the P35. Looks like I may need ton invest in a new board.

[CERO]

Well I dont think so... It is the Cpu, every cpu is unique, so it differs from voltage requirement, untill some1 test a curten cpu head on head on a P35 and X38 and see, but I doubt it.

[Viss]

Stil no E8500 here in NL so decided to try an E8400 first. Only been looking for max FSB, no spi tweaks at all. Q746A328 stock air:

Asus Commando
E8400 stock air
2x1gb Ballistix
8400GS
OCZ 520w

[ztop]

[QUOTE=Chri$ch;2722384]Why is this only a E8400/E8500 OC Thread?

here my first E8200 (Q745Axxx)

can you tell me the complete fpo\batch and pack date? and the cooling please??
thanks

[GLENBOY]

Stil no E8500 here in NL so decided to try an E8400 first. Only been looking for max FSB, no spi tweaks at all. Q746A328 stock air:

Asus Commando
E8400 stock air
2x1gb Ballistix
8400GS
OCZ 520w

hiya loco i have done the same i have an e8500 pre-ordered and got an e8400 which i should have tomorrow let us know the max cpu clock thats benchable on your chip please when you've time nice maxfsb , you'd like that with the 9 multiplier eh!

[Zeus]

Stil no E8500 here in NL so decided to try an E8400 first. Only been looking for max FSB, no spi tweaks at all. Q746A328 stock air:

Asus Commando
E8400 stock air
2x1gb Ballistix
8400GS
OCZ 520w

Nice FSB mate, did you max it out already?

[mrcape]

What is C-N's formula?

Hey, it's just a way to calculate your gtl vs vtt. In a nutshell it's keeping vtt voltage low (1.2v in my case) and setting gtl to around 55 & 58% of vtt. I shouldn't have called it "his formula", but it's something he pointed out in the DFI P35 thread and I kept going back to it when trying to get the right settings.

I think it's too early to count on this thread as evidence of good and bad chips. There are too many variables - mb & bios, stress method, ram, cooling, sensor accuracy, software accuracy, defined stability. I mean, I can boot into windows over 4ghz at crazy low volts but is it stable? Even if you can do SuperPi 32m I'm not so sure.

I think we have similar batch numbers so i'd like to see what you get for results. I also have an X38 P5E that I want to test withn the week to get performance on X38 vs P35 with the same exact chip. Noone seems to have that data that I can find.

[Plaicd]

Final 24/7 oc. P5k deluxe,Q746A381,Pack date-01/04/08

[Zeus]

Hey, it's just a way to calculate your gtl vs vtt. In a nutshell it's keeping vtt voltage low (1.2v in my case) and setting gtl to around 55 & 58% of vtt. I shouldn't have called it "his formula", but it's something he pointed out in the DFI P35 thread and I kept going back to it when trying to get the right settings.

I think it's too early to count on this thread as evidence of good and bad chips. There are too many variables - mb & bios, stress method, ram, cooling, sensor accuracy, software accuracy, defined stability. I mean, I can boot into windows over 4ghz at crazy low volts but is it stable? Even if you can do SuperPi 32m I'm not so sure.

I think we have similar batch numbers so i'd like to see what you get for results. I also have an X38 P5E that I want to test withn the week to get performance on X38 vs P35 with the same exact chip. Noone seems to have that data that I can find.

Okay but what is it good for? Stability? I' ve kept those at auto, does changing these values do anything for running lower Vcore?

[AriciU]

More here -> http://www.xtremesystems.org/forums/...=1#post2722819

[yenclas]

What is the best with this processor:

500*8 = 4000Mhz 1:1 mem=1000Mhz

or

450*9 = 4000Mhz 1:1 mem=900Mhz

?

[Devious]

Got a 8500 arriving tomorrow, it will have to be a good un' to replace my quad.

[Praz]

Hey, it's just a way to calculate your gtl vs vtt. In a nutshell it's keeping vtt voltage low (1.2v in my case) and setting gtl to around 55 & 58% of vtt.

On the DFI boards the GTL values in and of themselves have no numerical value in reference to the percentage of VTT that is being set. For instance, if a value of 60 is selected for GTL that does not imply that is the percentage of VTT being set nor does that represent a discrete voltage value. The voltage value that will correspond to 60 will change as VTT is changed. At the same time whatever percentage of VTT the value of 60 represents will also change as VTT is changed.

As no accurate measurements have been done when using these processors on DFI boards to assign any type of preferred percentage based on the value selected for GTL is a bit premature.

[Tancoll]

Don't you have 10 minutes ?

10 minutes?
Takes more to change my CPU

[mrcape]

On the DFI boards the GTL values in and of themselves have no numerical value in reference to the percentage of VTT that is being set. For instance, if a value of 60 is selected for GTL that does not imply that is the percentage of VTT being set nor does that represent a discrete voltage value. The voltage value that will correspond to 60 will change as VTT is changed. At the same time whatever percentage of VTT the value of 60 represents will also change as VTT is changed.

As no accurate measurements have been done when using these processors on DFI boards to assign any type of preferred percentage based on the value selected for GTL is a bit premature.

Interesting. Any insight on why setting VTT low and GTL manually at about 50 - 60% of the VTT set in bios seems to help? It works for me.

[mrcape]

Okay but what is it good for? Stability? I' ve kept those at auto, does changing these values do anything for running lower Vcore?

It seems to have done a lot for stability in my case on the DFI board. Leaving VTT high (like on 65nm) and the GTL at auto I would rarely be able to get in at certain speeds above 4ghz, regardless of vcore. With VTT low, and GTL set manually, I could get in and run stress tests. And as I notched the gtls down from there, the stability got better.

[Chri$ch]

can you tell me the complete fpo\batch and pack date? and the cooling please??
thanks

Batch: Q745A705 / Pack date: 01/11/08

cooling by water, is a hot CPU (my E8500 ES dont go over 35°C load @ 4Ghz)

[mrcape]

Batch: Q745A705 / Pack date: 01/11/08

cooling by water, is a hot CPU (my E8500 ES dont go over 35°C load @ 4Ghz)

Mine seems kinda hot as well, in core temp 96.1 but moreso at idle end at around 42c. In the bios it reads really cool. I can't imagine it's this hot at idle with low volts like 1.29vcore. Then at blend load it never gets over 59c even up in the 4-4.3ghz range.

[Praz]

Interesting. Any insight on why setting VTT low and GTL manually at about 50 - 60% of the VTT set in bios seems to help? It works for me.

I think you missed the point I was trying to make. Run GTL values at whatever works. But at this point in time nobidy knows that the values being set on the DFI boards equate to 50 - 60%. By themselves the numbers have no significant value.

[mrcape]

I think you missed the point I was trying to make. Run GTL values at whatever works. But at this point in time nobidy knows that the values being set on the DFI boards equate to 50 - 60%. By themselves the numbers have no significant value.

Got it, and thanks for the insight. Whatever the reason I'm glad it works! Did you get your chips yet?

[ztop]

Batch: Q745A705 / Pack date: 01/11/08

cooling by water, is a hot CPU (my E8500 ES dont go over 35°C load @ 4Ghz)

thanks... temp h2O?