The electrolyte does deteriorate due to water evaporation; it is no myth. That is the main reason most good-quality mobo manufacturers use solid polymer caps.

And I agree, the CPX-850 looks like a very interesting new twist. Let's hope CPX pans out well and doesn't go as largely ignored as BTX.

But that takes a very long time even in bad capacitors. Solid caps still have a lifespan, and they are typically about the same amount of time as electrolytics. They are just able to reduce ripple even more effectively, and are super low ESR.

There are often caps that look like solid caps, but are actually electrolytics, they usually have thinner metal so they can barely stamp a vent in it so you can't see it.

What causes low quality caps to fail prematurely is the electrolyte breaks down and produces hydrogen gas, that causes the electrolyte to dry up in the capacitor, increasing the ESR to a failure level. When too much of the electrolyte has broken down, it will cause a lot of hydrogen gas to build up and eventually burst the capacitor (a.k.a to vent, the reason why there is the vent stamped on the capacitor)

Every 10 degrees hotter a capacitor is, its life time is cut in half. This is true for polymer and electrolytic capacitors. Polymers tend to have a longer lifespan at a higher temperature.

Wow, a new statistic. I'll bet my broadcast engineers would love to know that..
I do not find it documented anywhere.

I know an article that complements that statistic wonderfully... let me try an find it.

The rule "every 10 degrees hotter a capacitor is, its life time is cut in half." Electrolytic caps folow that rule exactly, for polymer graphs the relationship is less linear. For both caps every ten degrees its lifetime is severly reduced.

Heres an article:
http://www.gigabyte.com.tw/FileList/..._all_solid.htm