Tricky question: is heat the risk in increasing vCore?

[Cherub]

The subject says it all. I have an AK35GTR which supports vCore up to 2.3v. Even at 1.96, my AthlonXP runs very cool, so I'm wondering whether I should increase vCore to overclock a little more (I'm at 1725/150 with an 1800+).

Are there risks besides heat? Or stated, otherwise: if the temps are low, is adding vCore less risky?

I realize this is a tricky issue, but I think it would benefit MANY overclockers to hear a debate on the issue.

 

[gdawson6]

They say higher voltage means a shorter lifetime, but that could just be that a higher voltage means hotter temps which would mean less life.

Im not sure what I would do, I would maybe take it to 2 volts max, but thats just me.

 

[ST4RCUTTER]

This is from one of pm's posts. Pretty much says everything you want to know:


Increased voltage leads to three reliability concerns directly and one problem indirectly. Apologies in advance if I have resorted too much to engineering-ese in this post... but it is a complex topic.

The primary reliability concern for high voltages is gate dielectric breakdown. At a high enough voltage the dielectric of the gate of a MOSFET transistor will 'breakdown' essentially like air breaks down in the presence of a high voltage differential between the earth and a stormcloud. In the case of a MOSFET, however, the breakdown leads to a permanent highly resistive path through the FET and the transistor breaking. So, high enough voltages will lead to a dead chip once the breakdown voltage of the weakest FET on the chip is reached.

Two other related concerns are PMOS BTI and NMOS Hot-E. Both are fairly complex topics - in fact I'm still not absolutely certain that researchers entirely understand the causes of PMOS BTI - so I'll gloss over the details as to specifically what happens. From a macro point of view what happens is that higher voltages on the gate lead to increased threshold voltages on the transistor over time. This leads to reduced drive strength (AKA higher channel resistance during "on"), and eventually the reduced current drive will led to a critical path failure.

Lastly electromigration - the physical transport of atoms under constant current conditions - leads to movement of atoms in the wires which eventually thins out the wires so that they become too resistive to carry current (leading to IR voltage drop on the power rails typically, leading to eventual failure).

The indirect source of problems from voltage is, as many others pointed out, heat. Heat affects all of the above issues strongly. They would all have an increased effect if somehow the die temperature were a constant and the voltage were increased, but heat is an additional factor. So increased heat leads to these effects occuring more quickly. --pm