Sentential03's burn in's really do work for me.

I decited to try some Burning in using CPU Burn-In like he suggested. I lowerd my voltage @ 2500mhz to 1.50v which is like at the point of unstability where i can barely reach windows.

I ran that for a little over 8 hours last night. I was quiet surpised to wake up seeing my computer still working and burning in. I decited to stop the test and run prime95 Small FFT to test for cpu stability. I have been running for like 15 min now or more without one error. I just now passed the first 5 test's on the other set now.

This is using 1.50v which is something that @ 2500mhz i was barely able to get into windows with. Maybe my dream of getting 2600mhz will be a reality sooner then i think.

Thanx sen for suggesting this to me.

How exactly is this "burning in"? Perhaps there is another factor?

I dunno but thats what the prgram is called CPU Burn-in, downloaded it did what sen said to do and im still running prime95 @ 2500mhz 1.50v without any problems

Yeah, I've used that process before. It will increase your overclock usually because of the fact that it will allow you to use less voltage and therefor drop your temps. Hopefully it will increase yours, until I have enough money for the A64 I'm overclocking vicariously through Sen's Winchester Thread

Im glad it helped bro. As I said before, there IS a method to my madness How else do I manage to consistantally get the best CPU OCs on the forums, with the least amout of voltage.

You should see godm@n. He nearly **** a brick when my 2600+ AXP was stable in prime at 2750mhz @ 1.9v

where's the link plz?

Is there a link to how you do your burn-in, Sentential???

Will it work on a P4? Cause, my 2.6C could use higher clocks :/.

burn-in should work on any processor.

Know Nuttin said:

Is there a link to how you do your burn-in, Sentential???

Click to expand...

Basically you use minimal stable vcore. Lower the vcore to the breaking point. Where programs like Prime95 go for only 2-5 mins. Stable enough for windows to run, but not nearly enough for games.

From there you run a program like CPUburn for several hours. You should start to see results after the first burn. Unlike before where P95 would run for 2-5 mins, it should run well over 10 maybe even an hour.

From there you repeat the process, pushing the vcore lower and lower, until it is within the "acceptable / desired range".

From them you do the reverse. You hold that vcore and start pushing higher mhz in 1 -2 mhz (on the FSB) incriments. Repeating a burn in cycle each bump for atleast 4 hours. You keep doing this until you reach your original speed with alot lower vcore.

The final step is to "reset" your voltage settings to their original levels and push the mhz upward. You will see that the CPU will respond a hellova lot better. Generally speaking you will notice that you have both lowered the vcore and raised your top end anywhere from 150-350mhz.

I do this will all my CPUs and it works wonders. This work unlimited times and so far I havent seen a point where it hasnt helped. Granted it takes awhile, but for some its worth the effort

Thats what I do and exactally what I told Candy to do.He shaved off .05v from his OC from the first burn. Im sure he can shave off a full .1.

I personally have managed to shave off more than .150 volts off my 2600+ M. It went from 1.825v to 1.775, then later 1.75 at a whopping 2.61ghz. All settings were completely prime95 stable.

Right now Ive shaved off about .25v from my Winchester. It has been the most stubborn CPU Ive dealt with besides my 35W 2400+ M. Im gonna get it to cooperate, its just taking much longer than I thought it would.

I plan to get this thing stable @ 2.54 with 1.45v so help me god! I am currentally at 1.52v

thank you very much. I shall give it a go.

The prime stable results, are they running blend?

Cheers

Oli

Ol!ver said:

The prime stable results, are they running blend?

Cheers

Oli

Click to expand...

Either one, it really doesnt matter so long as you know your RAM is stable. However blend primararly tests RAM

i looked at dload page but it doesnt say that it is compatible with win xp?

Sentential said:

From them you do the reverse. You hold that vcore and start pushing higher mhz in 1 -2 mhz (on the FSB) incriments. Repeating a burn in cycle each bump for atleast 4 hours. You keep doing this until you reach your original speed with alot lower vcore.

Click to expand...

I dont get this part. For example,if i start with [email protected] and then do the first steps thus bringing my system stable at [email protected]. then i start increasing my FSB, won't I end up with a higher clock than my original? But in your description it says to keep doing this until I reach my original speed with lower vcore. did i lower my clock in one of the steps?

Also,how does the burn-in program work to stabilize your system? Is it similar to prime95 or completely different?

RIPSTER said:

i looked at dload page but it doesnt say that it is compatible with win xp?

Click to expand...

Odd.... Im running it just fine on XP

I'm curious about this too, are we supposed to decrease our oc and then turn down the vcore or do we just turn down the vcore only in the first step?

iNterBeiNg said:

I'm curious about this too, are we supposed to decrease our oc and then turn down the vcore or do we just turn down the vcore only in the first step?

Click to expand...

The CPU mhz remain the same. You drop your vcore first. Then prime at the same speed, and keep lowering it. Once you hit your desired vcore, then you start pushing up.

Not once do you back off the CPU mhz of your overclock. The only thing you should play with <downward> is the vcore. Mhz remain at stock OC and later are pushed up, never down

I should give this a try once I get my winchester, hoping to get 300x9 in 3dmark 2001 and break 21k with ti4200