Electromigration; I think that's when they use cattle prods to move herds?
Sorry My Bad, I'm in a really good mood today!
Seriously, electromigration is when the electrical pathways degrade enough to allow some electricity to short to another electrical pathway causing corruption and errors, and under some circumstances complete blowouts.
Electromigration is going to happen eventually to just about everything, it happens over time with age even to electrical motors, resulting in having to have them rewound, meaning the electrical wire inside the motor replaced, a process normally done to very expensive electrical motors, cheaper motors are just disgarded and completely replaced.
Overvolting anything will accelerate electromigration as it causes faster degradation of the electrical pathways.
There is no way to know where the breaking point is of any hardware component as some can take more voltage than others, just because of the quality of the product, and some can fail with just a small increase.
Temperature is an added factor when it comes to electromigration, lower temperatures do tend to allow more leeway than higher temperatures do, but there in again comes the quality of the component, and that's always the unknown factor.
Does over voltage cause the electrical pathways of a CPU or GPU to degrade faster, of course it does, but that's always been an acceptable loss to overclockers, willing to trade off hardware longevity for increased performance gains.
Even with that there is acceptable ground as some lean toward the conserving side, and some just flat go for it, those that completely throw caution to the wind and shoot for 24/7 max clocks, seriously increase the electromigration process, and shorten the life of the component, but for that type overclocker it was an acceptable loss.
Thanks to both of you guys, you really cleared my mind what is the effects of electromigration due to over voltage which eventually increase the temperature as a result of overclocking the CPU that will shorten the lifespan a PC components.
Thanks to both of you and I appreciate the detailed response to my post here in Tom's Hardware.
Because of Electromigration the life of a modern processor is only approximately 7 years (on 24/7). As the manufacturers go to smaller feature size this life span will reduce. When you overclock you increase the temperature and the current which increases electromigration even more which is why overclocked processors have a noticeably shorter life than non overclocked processors. If you want your processor to have a long life then keep the temperatures as low as possible and the CPU voltage as low as possible.
However most overclockers have the attitude that if the processor lasts a few years by the time that it fails it will be obsolete anyway and they don't normally have the computer on 24/7.
Now I know the consequenses of CPU overclocking. I hope every CPU overclockers out there knew this before they overclock their CPU's.
Good point and it may be something we need to directly address, long term overclockers have been aware of electromigration from way back and it has long been accepted and understood, however those new to the overclocking realm need to know how it will affect their overclocking intentions.
Electromigration is far from the multitude of dangers overclocking brings to the table, but is something that is a need to know consideration, because it's one factor directly responsible for pre-mature CPU failures.
The biggest danger in overclocking is doing it with only partial knowledge of how to do it in the first place, and the forums are flooded with too many wanting a quick fix, and be on their way, so considerations like electromigration may be something they are completely unaware of.
Good thread you've posted here, it should help others add electromigration to their overclocking plus and minus considerations.
Your post containing the link to overclock3d has been removed, if you want to input some solid factual information of tested CPUs to failure, solidly attributing it to electromigration that will be acceptable, however theoretical assumptions that even the writer is not sure about is not solid facts.
Additionally the thread perpetrator was adding his theoretical input on someone else work, which was clearly indicated on the thread header and his last conclusions were really not what we want to share with new overclockers here, we would just assume they ere on the side of caution, and not throw caution to the wind.
When you post links here at THGF read them thoroughly first and make sure they do not link to other websites, because you will be held accountable for what you post here.
The link you provided is very enlightening. Maybe my CPU would last about 2-3 years and maybe after 3 years the i5 2500K is already obsolete. Well actually if the Ivy Bridge CPU will be available, benchmarked and tested by Tom's Hardware and user's then I will upgrade my CPU(i5 2500K) to i7 3770K maybe a few months after it's release is the good time to buy it. What do you think?