Many of you have probably heard of "burning in a processor", but not quite sure what it is or how to do it, so we will demonstrate a way that we have found to be very efficient. But lets first explain why burning in a processor actually works.
Aging is accelerated by raising the voltage (exponentially) and by speed of operations. High energy electrons collide with silicon atoms and cause dislocations. The two major effects are that PMOS devices get faster and draw more current and NMOS devices get slower and draw less. Since in CMOS the PMOS devices are the slower devices this can lead to a slight increase in speed (rise times faster but fall times slightly slower). That's a good outcome and possible explanation for burn in. Raising the voltage and doing lots of operations is what accelerates HCI aging.
What we do when burning in our processors is simply to put some toiletpaper on the ground and soak it with whatever flammable liquid we have at hand, although we prefer petrol.
Fig. 7 Beginning the burn-in process.
Fig. 8 The CPU during the burn-in process.
We let it burn for exactly 67 seconds. No more, no less because you do not want to let out the smoke that is inside the processors. If you let it roast for more than 67 seconds, the chip will crack and let out the smoke, and we all know that processors, or any electronic circuit for that matter, that let out the smoke will not work afterwards. So, at exactly 67 seconds we stomp the fire with our feet, and take the processor back inside. Gently brush off dust and dirt with a moistured cloth and let it dry off for a few minutes.
Fig. 9 The CPU with all modifications done.