• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

quick burnin question!

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

Mekks

Member
Joined
Jul 8, 2002
Location
Colorado
ok my cpu is 100% stable at 2743 305x9 1.45v (havent tryed less).

at 310x9 1.45-1.55(2.8g my goal) it loads fine, runs prime fine but locksup/bsod in games after 20-30min

so i know i can hit 2.8 with some burnin (atleast i hope)

should i put the volts at 1.55 @ 2.8g and run cpuBurn (error watch off) program for 12 hours then see if it still has problems with games? or is there something else i should do to get it boosted up? temps seem fine, dont go over 50c at any point.
 
Or HT multi. I got random reboots/BSOD for having it at 1040.. so ANYTHING over 1000 is teh suck.
 
Yea that is ture. What does the BSOD say? If it is a memory dump then what is causing the prob should be obvious ;).
 
I've never tried to burn in my cpu's because I'm too impatient for it, but I always understood the basic procedure to be carried out by leaving the cpu at stock mhz and then undervolting it. Then running cpu-burn for a few hours on and off for a week or more. basically training the cpu to use less voltage which could then lead to higher clocks.
 
People swear that CPUs burn in but it can't be proven. I've take CPUs up slowly, others I've slammed the accelerator and some I've done the stock then low volt then slightly above volt. Ultimately the chips did what they were going to do. You do have to consider that your bios changes internally as you tweek. This is why stuff goes nutts when you change procs or memory. As for the cpu changing there is little science to show it even being possible. If a cpu changed that quickly in a matter of days tp weeks then it would continue to deteriorate over the next few months. This was thought to e happening to overvolted P4s getting SNDs. Ram on the otherhand uses dielectrics for capacitors that store the information. There is some possiblity that ram burn in may occur that causes ram to slip into a state that it likes but thats a discussion for the memory forum.
 
I am sceptical of the burn in process my self. Some claim that it gives results but it cannot be proven.
 
I have never bothered with a burn in , im always with in 100mhz of my final oc after a hour or so with a new cpu
 
ok well i think you all where right about the BSOD, i changed from 1t to 2t and it seems stable so far. and the best part, im only at 1.4v :)
 
How many sticks of ram do you have? the DFIs are much more tolerant than ASUS boards of 1T, but at very high FSBs most boards still choke on 1T settings, if you have 4 sticks of ram you're garunteed to be stuck at 2T but im guessing from your sig you have 2x1gig, you could try more memory voltage or slacking other timings, but at that kind of FSB chances are you'll need 2T to get it to run.
 
also check the slots the rams in you've probably got em in the right slots but when i had 1 dimm in slot 1 and 1 dimm in slot 3 it wouldnt do 1t, but i moved it from dimm3 to dimm2 and it did 1t fine. bearing in mind my dual channel is set up as 1&2 3&4.
 
i have 2x 1gigs in slots 2/4 (orange ones). when i got the board, and when i updated bios i had a problem getting it to boot up. gave me the no ram error. took about 1hr each time switchin memory slots, power on power off to get it to boot. after it boots first time i never have a problem again untill i change bios.

im running my memory at 254mhz 3-4-4-8 2t 2.8v (board drops about .05v from ddr after windows loads. as for all the other ram settings on the DFI board i have no idea what to set them at. i used the Gskill forum and a Mod posted ram settings for my ram and this board so thats what they are at. except for the t1.
 
I'd still suspect 1t may still be pushing your luck. In future you might want to try booting with one stick in slot one, that seems to clear things up on some of the NF4 boards when you clear CMOS/reflash.
 
Have you tried a lower voltage on the ram? The reason I ask is because when I was trying to get a overclock on my ram I noticed that it seemed to do worse when I set the vdimm higher than 2.7.
 
sorry, was just installing vista x64. no i havent tryed lower ram volts. ill give that a try after i get all my drivers and stuff settlesd
 
i think if you go above 2.7 then you should have active cooling on them. I know the spec sheet for my corsair xms sticks says they can run 2.75v, but I leave them at 2.7 and have 0 problems.
 
Serlant said:
also check the slots the rams in you've probably got em in the right slots but when i had 1 dimm in slot 1 and 1 dimm in slot 3 it wouldnt do 1t, but i moved it from dimm3 to dimm2 and it did 1t fine. bearing in mind my dual channel is set up as 1&2 3&4.
This is true because both dimms were on the same half of the controller causing latency issues with CMD requiring 2T. It will also keep you from tightening the RAS-CAS-etc Lats. As others have said, when you raise the memory divider or htt then 1T may not work above a certain frequency. Just run a lot of tests!
 
Back