What is Burning in?

[matrix]

hello guys..
im a real newbie at o/c ing
so..
whats this burning in everyones talking about?

if its a prog.. than can you give me a link to get it...

thank you for your tiime

[DocClock aka MadClocker]

Burning in is basicly letting your cpu get used to whatever you are trying...some people believe that if you overclock your cpu to it's highest stable speed, and then run it for a while,(burn in) to let everything settle down, and hopefully let you go further after a burnin period.
others will leave the cpu at the default speed, but turn up the voltage, and others are even turning off the cpu fan for twenty seconds or so.
People have different opinions as to how to burnin, but the idea is just letting the cpu get used to the high speed, or high voltage, depending on which method you are using..and the debate goes on...some people don't think burning in really works at all, but most here will say that it does work.
I hope this helps

[Phil]

I read an article a while back that explained how the transistors degrade over time and this degradation can improve the quality of the signal being transmitted to and from, it then explained that by increasing the voltage and running the chip at 100% load for a while can speed up this degradation, lowering the life of the cpu but improving how overclockable it is. When they say lowering the life they only mean to about 4-5 years becasue most chips are designed to last 10-20 years, but lest face most chips are completely outdated after 3 years so the extra speed you get in the mean time is a plus. They also said that this effect works better at lower temperatures and advised that burning-in should be done at default speed with a high voltage, I my self burned my celeron 600 in at 600mhz with 2v for a few weeks, this let me run it at 900mhz with at a lower voltage and allowed me to run it at 1008mhz stabally so I would vouche for it's effects.

[[OC]_SR20DE]

Burn-in also brings the CPU temperatures up stress the CPU to death. It's a good way to train your CPU just like people going to the army and get used to with its tough life. Exactly the same with treating your CPU. Burning-in does work and it worked for me..

You can go to download one of the well know softwares called "Seti". at: www.setihome.com

Dang.. Im having trouble using this html tag.. it aint working. Anyways, Seti program isn't the only good way to do it.
You can run 3D Mark 2000 and loop it all day. That's burn-in.
You can run Quake3A and loop it all day. That's another good way to burn-in.
I successfully ran Quake3A 8 hrs of Loop and totally stable. Had no crashes @ 1064mhz.

[Ben]

DocClock aka MadClocker (Feb 25, 2001 10:35 a.m.):
Burning in is basicly letting your cpu get used to whatever you are trying...some people believe that if you overclock your cpu to it's highest stable speed, and then run it for a while,(burn in) to let everything settle down, and hopefully let you go further after a burnin period.
others will leave the cpu at the default speed, but turn up the voltage, and others are even turning off the cpu fan for twenty seconds or so.
People have different opinions as to how to burnin, but the idea is just letting the cpu get used to the high speed, or high voltage, depending on which method you are using..and the debate goes on...some people don't think burning in really works at all, but most here will say that it does work.
I hope this helps


Burning in really seems to be to let the user get used to what they are doing. In other words I have spent XXXX.XX dollars and I am driving it past reported limits and now I am going to run it and get use to the risk associated with it. Its kinda like bending the spoon, all your really doing is bending yourself. hahaha. I know that when it comes to resistors when you run them at higher than specified wattage they will slowly fall out of tolerance as where after a few seconds of a 1/4 watt resister @ 5% will generally be at  of rated resistance. So think of this, you are maxing out resisters inside your cpu, you are driving transistors outside their limits you are truly causing a physical change in their environment, but it can never be anygood. I think the actual process under an VOM (Volt-ohm meter) analog looks like this. Apply 1/2 watt of power, and measure with VOM, the resister will short then open and you can watch it do so on your meter. Granted that is P out of tolerance where you guys are going perhaps %5 to  percent out in terms of voltage and associated current. The only thing saving your butt is the cooling system and that doesn't stop the physical change in the CPU environment. I know overclocking can work and some systems are stable for long periods of time, but has anyone ever put together a database on system lifetime and MTBF for components. Granted it may be longer then the 1-3 year lifecycle of systems which makes it in the end just fine. How many systems have run 10 years straight (like my 386sx-16Mhz). Has anyone ever done a case study on the %v_volts and %I_current that you can actually increase without causing a change in the cpu's environment? When do you start increasing the amount of carbon in the cpu's traces?? You have a trace that is .18 micron across that is 180 nano-meters how much current can that wire handle? You have PN junctions everywhere, how much overvoltage can they handle before the Junction capacitance changes for all time the resistance changes the current handling capacity changes the voltage handling capacity changes. And these result from physical changes and at the very bottom of all this the physical changes are changes in the makeup and actual chemistry of your CPU. How many more atoms of carbon are in the traces etc,etc,etc,