• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

AMD® Phenom™ II Overclocking Thread

the CPU temp is the real temp.

it's impossible have the core temp lower than the CPU temp.... hence thats why the core temp is just software generated.... when i had a 720 the core temp was 18c idle which was impossible because my room was warmer than that.. the CPU temp was around 30c idle which was believeable and true..

yes i would say just another single bump to the vcore should do it
 
Last edited:
cheers for your reply Gareth,
I thought the temps were a bit strange, especially when I was getting 14c the other day :)

I'll bump the vcore up and see how I get on. If I can get it stable at 3700 I'll be chuffed!
my first ever overclock :)

will make up a bit for my chip not turning into a quad :(
 
Morning guys i really need some advice my 3.6ghz OC failed again :(. I set the vcore to 1.41 in the bios woke up this morning to find that my machine has restarted, i checked event viewer in windows and saw that it said blue screen at around 5.15 am so that was about 6hrs of prime started at 10.30pm last night. I was checking hw monitor for about 30 mins before i went to bed and i really believe my board suffers from vdroop quite bad, it was showing the lowest vcore voltage during the test was 1.31 and the highest it reached was 1.39 and like i said its set at 1.41 in the bios. I have checked the load line calabration in the bios and its set to auto but auto states it will run at 57% could this maybe be a cause to the lower voltages im getting?

The only thing i have changed is the vcore everything else is on auto, my temps were looking good cpu was showing 39c underload and the 4 cores were showing about
40-41c using hw monitor.

Hope someone can help

Many thanks
 
Liamcrane - As stated several times in the thread the best loadline calibration setting is 19% to eliminate vdroop. Try that. If it does not work then up the vcore and try try again. :D

Hi guys not posted on this thread for a while but I have finally got round to doing the next stage of fiddling with my chip.

I am retreating from my 3.79GHz and going for 3.74GHz (with a .04 drop in vcore) it takes me nicely away from 1.5 and seems to have dropped load temps by about 3 or maybe more.

3.7 passed 11 hours overnight at 1.45 vcore. I am pretty pleased with that so if this one fails I think I will stick at 3.7. I am so jealous of all those 4GHz clockers but I dont think my chip would come near! :D
 
Last edited:
could do with some advice please guys, I'm getting BSODs all the time now, and it's a Double Fault:

Double fault
A double fault occurs when an exception occurs while trying to call the handler for a prior exception. Normally, the two exceptions can be handled serially, however there are several exceptions that cannot be handled serially and in this situation the processor signals a double fault. The two primary causes for this are hardware and kernel stack overflows. Hardware problems are usually related to CPU, RAM, or bus. Kernel stack overflows are almost always caused by faulty kernel-mode drivers.
Click to expand...

I've run memtest on one stick of ram at a time, both were fine for 2 passes with 0 errors. I've no idea what a kernel-mode driver is or how to check for it, but I'm checking everything. I've reset the CMOS, done a full cleanout of the hard drive and registry and defragged, and I can't find anything wrong.

I'm also checking for CPU errors while I'm at it, and though you guys would be the best to ask? How do I check if my CPU is degrading?

Cheer for any advice!
 
gareth170:

Nope just 2 passes on each stick of ram. It won't stay on for 12 hours that's the problem. :P

Yep everything is stock now, reset the bios. It might be heat issues as the room is generally warmer now than it has been over the winter, but it's not the CPU that's warm, I ram P95 last night for 45 mins and it only got upto 42C.

According to G.Skill this ram is supposed to be run at 1.65v, so I've always run it at 1.64v, (closest I could get), now it's at auto though and comes out at 1.8v.

I have W7 v64 installed on an IDE hard drive, I'll stick that in tonight and leave it on running memtest or prime95, just to try and persuade it to crash. That wasy I can work out if it's a Windows/HDD/SATA controller issue.
 
Last edited:
According to OCUK they should be JEDEC standard 9-9-9-24 values at 1600MHz.

I've no idea how to check if they are though? I've never touched ram timings as I left my ram stock and overclocked on the multi not the fsb.

I forgot to mention, I ran two passes on the ram that took about 40 mins each to complete, but got 0 errors on both.
 
Just wondering if anyone has done any undervolting tests yet? . . . . either at stock clocks or with a mild overclock? . . . . interested to know if anyone can get an 8 hour blend test off using quite low vCore? . . . . I'm thinking approx 3.0GHz-3.4GHz should be possible with maybe 1.200vCore? maybe a touch more? :cool:
 
gareth170

it will happily do this overnight. i just saw the thread and thought id contribute. im too lazy and impatient...
although like i said, 4ghz seems to be the absolute limit, and i daren't go over 1.55v. even for dual - it may run fine just for bragging rights, but id prefer not to have a frazzled chip...

Big Wayne

:D

craftily photoshopped my name on ;)


just out of interest, would using the IGP hinder overclocking the reference speed (or whatever its called now...)? with my last board (same as sig, but i frazzled it :S), i could go upto 240mhz before artifacts came on screen and would crash at 245. the RMA replacement did the same, but if i tried to run a game or furmark, it would crash at anything above 230 (it would freeze and display a blue checkerboard pattern, then reboot after 10seconds ish. no artifacts though...).

i swear most boards ive seen could easily hit 260-270mhz... all i want is 250 so i can run my RAM properly...
 
Last edited:

Funny you should mention that...

When I first got my 945 I left the clock at the stock 3GHz and undervolted to 1.2v and it seemed to work ok, I even ran prime on it for 4 hours without a hitch..... ok so 4 hours isn't 8 hours but hey it was stable for that time :)

So yeah at 3ghz I would say it is definitely possible to run at 1.2v I'll try and run a full 8 hour prime blend on it in the weekend to properly confirm this
 
Last edited:
Hey thingemajib, there doesn't appear to be any disadvantage to overclocking using the IGP, at least I have not had a problem myself and have no read about a problem? . . . I'm aware that in the past that an IGP limited the clocks on some uATX boards but this situation seems to have improved greatly since then! :)

Btw: not sure if you have read the Welcome Post or how much of the thread you have read but did you spot the bit about me asking to avoid quoting if at all possible? . . .you seem to have ignored this and gone right ahead? . . . . sorry to nag you but can you delete the quotes from your previous post and maybe use the persons name? . . . thanks for your co-operation! :cool:

Fonz-Valo Please "No Quotes" . . . please!
 
Finally got my 3.6ghz stable after 10hrs+ of prime im very happy :). I set the vcore the 1.42 and i left the loadline calabration on auto which is 51%. Max temp was 41c underload and 27c idle, might try lowering the load line calabration to 19% and see if it will allow me to reduce the vcore a little. My voltages while under prime were at a low of 1.31 to a high of 1.45 with the vcore set to 1.42 thats quite an up & down ratio
 
Big.Wayne you may remember there was this undervolting thread a while ago :) Best I'd gotten to there still stands; "stock" voltage assumed as 1.35v (though my chip says 1.375), and taken from the various p-state settings for lower speeds.

3.5GHz 1.3375 = 1% under stock volts and 10% overclock
3.2GHz 1.2375v = 10% under stock volts
2.4GHz 1.0500v = ~16% "
1.6GHz 0.8250v = ~22% "
0.8GHz 0.7500v = 25% "

CPU NB is at 1.1v up to 2.4GHz, needed 1.175 for 3.2GHz and above.

Voltage requirements really tumble below 3.2, so yeah, I'd imagine you're right that you could do 3GHz at 1.2v :)
 
You must log in or register to reply here.
Back
Top Bottom