That would mean a less than 10% increase in clock lead to a 10% increase in failure rates.That's the scale I'd like to see. I guess there's a curve of speed/temp/reliability and someone's picked...5% failure in 3 years, say, and 53 MHz would push that up to...5.5% (?)
The big component for device failure past manufacturing failure is some kind of degradation that has some relation to heat. Electromigration, mechanical failure, chemical degradation, degradation of SRAM, and the like all have a temperature component. The clock speed itself might have a minor contribution to some forms of transistor degradation, but most of the big ones are chemical or physical processes that accellerate with temps, current, and voltage bumps.
Since this is likely a bump that tries to stick within the same temp and a voltage bump would likely prevent that, we've clamped down both thermal and voltage factors that dominate.
Within a voltage step, the trend is linear. At steps this tiny, we can probably assume linear even without that relationship.What sort of power increase are we talking about here? It'll be more than a linear relationship to the clockspeed increase, so a 7% increase in clock will result in, say, 10% increase in heat. Is there enough AMD information out there to identify a realistic heat increase?