Burn-In your CPU??

Having your CPU running apps like Prime95 for hours can it be call a burn-in process???

nope I believe it is running your cpu at a higher voltage than it needs, so then it gets used to running that voltage, and you can lower it. Some people swear by it others, don't believe it works.

dicecca112 said:

nope I believe it is running your cpu at a higher voltage than it needs, so then it gets used to running that voltage, and you can lower it. Some people swear by it others, don't believe it works.

Click to expand...

Not only that, but underclocking the processor at high voltages, to keep temps as low as possible.

Kinda like taking a Barton 2500 and running it at 600mhz @ 1.9v.

Wait so if "burn in" works than what does it essentially achieve. Does it achieve higher oc at lower volts or just the ability for the cpu to withstand higher volts or what?

it varies cases from cases. some ppl about to achieve higher oc or lower volts or both. while others has no success at all. i burned in my p4 for about 12hrs, but couldn't get higher oc or lower vcore.

Leave your CPU at stock speed, but increase the Vcore as high as you can.

Personally, I never see any benefit in the burn in process. I have both 2600XP mobile and the 1800 JIUHB. I didn't see any gain in overclocking. The only benefit I see is "cooking" the thermal paste. Artic silver type thermal paste thats around 200+ hours of cycling before they are cook. When I say cycle, I mean leaving your computer on for couple of hours during the day, and shut it down during the night. Once the thermal paste is cooked, it transfers heat better, and you will see a drop in the CPU temperature.

The only true way to increase the overclockability of your CPU is good cooling, and L12 mod. 50% of the people here in the forum don't believe in the burn in process (I think).

The proper way is

The higher the voltage, the higher the effect, the higher the temperature, the lower the effect. Thus, you would run your CPU at minimum clockrate, maximum voltage and minimum temperature (remember, voltage and temperature are dependent of each other).

Quote taken from

http://people.freenet.de/s.urfer/conditioning.htm

I see the L12 mod mentioned often ..
I'm having a prob getting my 2400+ over 2.3 gigs. Its a tbred b, factory unlocked, and at 10X or above, I can't run anything over 185-190fsb... its really funky...
memtest runs perfect at 213fsb, and at 9.5X, I can run 210-213 fsb fine.
On same setup, my superlocked (other chip)2400+ was running 163X15..2445mhz, and I had my 1700+ running at 212X8...
What's up? Bad Cpu??

there is no reason for the burn in process to work.

the only thing increasing the voltage to high will accomplish is damaging the cpu.

you cant make transistors stronger

CPL.Luke said:

there is no reason for the burn in process to work.

the only thing increasing the voltage to high will accomplish is damaging the cpu.

you cant make transistors stronger

Click to expand...

And I'am guessing you have done testing yourself and this is your opinion correct. If this is all fact's I would like to know where you got it from.

jonspd said:

And I'am guessing you have done testing yourself and this is your opinion correct. If this is all fact's I would like to know where you got it from.

Click to expand...

No one can prove anything. Personally, I feel that the burn in process is still a myth to me. It may work for some, and may not work for others.

Paging Hitechjb1, Hitechjb1 there's a thread with your name on it... Any thoughts?

The proper way is

The higher the voltage, the higher the effect, the higher the temperature, the lower the effect. Thus, you would run your CPU at minimum clockrate, maximum voltage and minimum temperature (remember, voltage and temperature are dependent of each other).

Quote taken from

http://people.freenet.de/s.urfer/conditioning.htm

Click to expand...

Indeed. That's what I would have linked to.

CPL.Luke said:

you cant make transistors stronger

Click to expand...

I agree with that..

Stronger no but can you make them flow current faster?

No.

edit: not unless you crank the voltage waaay up to the point that the whole thing becomes one great big short circuit between vcore and gnd

jonspd said:

Stronger no but can you make them flow current faster?

Click to expand...

Do you mean after the burn in you can make the current flow faster thru the transistor??????? How is that possible?? Do you have any article to show this?? I never heard of that.

Can any semi-conductor engineer verifty this????

Kenshiro said:

Do you mean after the burn in you can make the current flow faster thru the transistor??????? How is that possible?? Do you have any article to show this?? I never heard of that.

Can any semi-conductor engineer verifty this????

Click to expand...

i dont think he has a article that shows that since he was asking about it not telling.

I don't think that burn in actually does anything, but I haven't had enough engineering to be all definititve about it. That said, I can see both sides of the issue.

I think it's more superstition than anything else, and that chips are fabricated with a certain voltage to frequency to temperature tolerance. That makes more sense to me.

There is no such thing as making current run "faster". I think people are a little confused about how a cpu and transistors work.

A transistor is essentially an electronically operated switch, which can be in several different states depending on what voltages are at what parts of the transistor. It doesn't matter what all these states are, what does matter is how fast a transistor can swith from one state to another. This switching is the process by which a CPU actually does calculations. To switch from one state to another, power is required. This power comes from the voltage applied. If more voltage is applied, the transistor will be able to switch faster. If too much voltage is applied, the transistor will fry.

In a CPU, there are lot's of transistors (over 100 million on a P4 I think). Not all of these transistors work, and not all of them will work at the same speed. This is why some chips will max out at 3ghz while other chips with the same core can do 4ghz.

As far as burning in, I do believe it can have some effect on how fast a cpu can run. Giving the chip more voltage than it needs will put more voltage into the transistors than they were designed to take, and the weak ones will fry. These weak transistors will no longer slow down the rest of the transistors, but they will also no longer do any work, so it is a trade off. It's kinda like killing on the weak animals in a herd. It will make the herd stronger at frist, but if you get too carried away you will weaken the herd.

As far as experience, I have never burned in a processor as described in the link above. I have noticed my chip becoming faster after being stressed. When my chip was new it would not do any more than 2.25ghz stable. After contunually raising the speed slowly and keeping my cpu at full load, after a few months I am now at 2.39 prime stable, and still pushing for more.