an E8500 on a Asus P5N-e SLI

[nefariouscaine]

glad to hear you got past your OC wall - try to test long than 3 minutes and let me know how you do and what temps you're getting with that AC 7 pro after a long run at full load

 

[Foxery]

Originally posted by: tacoturtle
5-5-5-31 2T Trc 23
OCCT: (10 minutes then i stopped it)

Using the a max value for part of your RAM timings doesn't make much sense. Make sure to set that back to normal before considering your build finalized.

Also, you have to run OCCT for a lot longer than 10 minutes

nefariouscaine: OCCT reads temps and voltages, and I think shows the same values as CoreTemp. (Subtract 10C if you prefer RealTemp.) Go to Settings -> Monitoring and add as many readouts as you like. I also set the Max Temperature on mine 15C higher so it would wait for an actual error to stop, not just temperatures.

 

[nefariouscaine]

Originally posted by: Foxery

Originally posted by: tacoturtle
5-5-5-31 2T Trc 23
OCCT: (10 minutes then i stopped it)

Using the a max value for part of your RAM timings doesn't make much sense. Make sure to set that back to normal before considering your build finalized.

Also, you have to run OCCT for a lot longer than 10 minutes

nefariouscaine: OCCT reads temps and voltages, and I think shows the same values as CoreTemp. (Subtract 10C if you prefer RealTemp.) Go to Settings -> Monitoring and add as many readouts as you like. I also set the Max Temperature on mine 15C higher so it would wait for an actual error to stop, not just temperatures.

Oh I already did all that - OCCT uses CPU-Z to read cpu voltages and FSB settings - just try to have them both open and you'll get an error. I never have any stress test setup to stop at a temp - thats one reason I tend to watch them at least for the first hour so I know if I need to cancel and turn back due to temps.

 

[tacoturtle]

The 2x2GB DDR2 1066 DOMINATORS came in today. Got them running stock 5-5-5-15-2T. The heatsinks get in the way of my air cooling so I had to mount the fans with brackets in funny spots. I'll post a pic later. Running occt for awhile yeilds cpu: 36 core 1: 51 core 2: 46 was the highest according to speed fan that was adjusted using real temp for over 15 minutes. I played cod4 for several hours and no problems also. I also got a Aero Cool Easy Watch. Pretty cool monitoring tool. Im gonna run some tests on temps without using the Aero Cool to contron strategic and loud fans and see if it does a better job than having them all run 100%.

Since the ram is running lower than spec'd, shouldnt I be able to lower voltage or timings some more?

 

[nefariouscaine]

update on my end - i haven't been able to get 4275mhz full stable yet - tried as much as 1.39v on one run of testing and got too high temps for anyones liking due to the lovely 90 degree+ los angeles weather. Started to think its my board holding me back and it sure as hell doesn't like any fsb over 450mhz - it booted after i cranked the VTT all the way up to 1.74v but errored out before windows even

currently running an orthos test to see if at 450mhz and a dropped CPU multi to x7 (3150mhz) I get any errors out of orthos and then OCCT

i'm thinking a new board is in my horizons its been good while it lasted but its not a shocker that 500mhz on the 650i is ever going to happen (wanted to test FSB wall on the cpu...)

 

[nefariouscaine]

well 5-6 hours of orthos was stable no prob at 450mhz FSB - temps didn't get above 40c on the nb heatsink (not sure about my placement on this temp sensor just yet, might need to move it)

possible my board just can push enough juice to my cpu at that high of clock - i know the 650i only has 3 phase power... -- currently looking for a cheap replacement (sub $200)

but it runs fine at 4275mhz - only issue i noticed was it won't pass more than 2-5 hours of orthos blend - usually won't get much passed an hour on large FFT's on orthos and firefox has an issue loading Gmail (i blame firefox on that one as Opera work like a charm)

back to running the memory at all auto settings and it defaults to 5-5-5-11 T1 (kinda interesting and it does do T2 on auto at higher speeds)

so i'm back to running at 1700 (425mhz) FSB and at 4037mhz cpu at 1.264v vCore

 

[nefariouscaine]

still kinda baffled at my 4ghz cap break - especially at the voltage my cpu will run 4ghz - got another 7.5 hour orthos large fft run

http://johndoe.hotlightbulb.co...%20change%20orthos.JPG - large image

i dunno lol

suggestions??? thinking about picking up a cheap MSI P35 Platinum board - really want to use this chip to the max

 

[tacoturtle]

Originally posted by: nefariouscaine
so i'm back to running at 1700 (425mhz) FSB and at 4037mhz cpu at 1.264v vCore

1700 @4037? u meant 1800 right?
Im running 1725@3881mhz
Im thinking about ditching the artic 7 for true or another cooler. I think thats whats holding me back from 4.0

 

[Tempered81]

hes got an extra 1/2 multiplier

 

[nefariouscaine]

Originally posted by: jaredpace
hes got an extra 1/2 multiplier

beat me too it - yeah my multi is 9.5

 

[VirtualLarry]

Originally posted by: nefariouscaine
it sure as hell doesn't like any fsb over 450mhz - it booted after i cranked the VTT all the way up to 1.74v but errored out before windows even

That much VTT is sure to kill a 45nm chip. Be careful!

 

[nefariouscaine]

Originally posted by: VirtualLarry

Originally posted by: nefariouscaine
it sure as hell doesn't like any fsb over 450mhz - it booted after i cranked the VTT all the way up to 1.74v but errored out before windows even

That much VTT is sure to kill a 45nm chip. Be careful!

Now this is going to sound n00bish but VTT is the NB chipset voltage right??? I would never set the vCore to something that high and now i'm back down again running 2.01v on vDIMM, 1.264v on vCore - and auto on NB voltage - running 4037mhz

looking at $150ish motherboards now - just sold some of my spare gear (my other 650i and the E6600 i had) so I got a little extra to work with right now on upgrading

 

[tacoturtle]

I went to a evga ftw 750i mobo and was able to go to 4.05ghz on the e8400 without changing anything except the fsb to 1800.

 

[BonzaiDuck]

Originally posted by: tacoturtle
I'll give that a try. Good thinking. I am using the same board you are and bios version 900. Im thinking its the ram. I got some ddr2-1066 on the way so I'll know more then. Your oc the e8500 to 4.0 so i would figure I could at least get to 3.7 on the e8400.

I'm making a fairly confident guess that what holds for a 680i/750i/780i chipset, in certain specifics, would also be true for the 650i. In fact, I think the imperative I want to raise here carries over into the Intel chipset line.

Myocardia mentioned raising the VTT (CPU-FSB or CPU_VTT) voltage. I think it's generally understood that this needs to be done when approaching or exceeding 400 Mhz.

And the caution to follow derives from Anandtech's experiment with a QX9650 (45nm) processor. You risk killing the processor in short order when you raise the VTT voltage above 1.4V. In the Anandtech experiment, it simply died when they set it to 1.5V.

Graysky's over-clocking guide notes that at 400 Mhz, about 1.30V is called for. I didn't make this adjustment with an E8400 on a 680i chipset until after several BSODs.

I only bring this up (with too much time on my busy-body hands) because all of tacoturtle's posts here don't mention the VTT voltage.

WHOOPS! Just realized I had skipped the second page of posts on this thread.

I just wish I could get my E8600 to run at its OC setting at these low VCORE voltages reported here. But I have to run it so that CPUZ reports about 1.33V idle and 1.29V under LinPack load. I didn't do a pencil-mod with my 780i board. I can push up to 4.25 Ghz with reported voltages I give here, but I don't want to volt it any higher than that.

 

[tacoturtle]

Im pretty satisfied with 4.05 for now. I haven't pushed much past that. Mostly due to inadiquete cooling. If I was on water I may try to push it some more. I think Im saving water cooling for the i7 line. I think Im turning into an overclocking junky/risk taker.