Accounting for Frequency when Calculating Power Loss

[wwswimming]

the formula for power loss in a component is ...

P= (V^2)/R

in one of the other threads, somebody was asking about raising the CPU voltage .05 volts in conjunction with a CPU freq. increase from 2.4 to 2.5 GHz (on an AMD X2 3800)

i doubt that calculating the increase in power consumption is as simple as calculating the ratio (1.4^2)/(1.35^2), the ratio between the 2 voltage levels, squared. since the frequency is raised at the same time.

can anybody provide a fairly accurate formula for calculating the energy consumption of a CPU ? e.g. by curve-fitting data from TH or A'Tech ?

thanks !

 

[BrownTown]

there are two components of power consumption in a CPU, the static and dynamic power consumption. The static power consumption is due to leakage and is V^2/R. The second is the dynamic power which is V^2*F*C. Where F is the frequency and C is the capacitance. Basically both are proportional to voltage squared, so raising the voltage and keeping frequency the same should mroe or less increase power consumption based on the ration of the voltages squared. Increasing hte frequency will casue the dynamic pwoer to increase proportionally, but you firsth have to know what % of the power consumption is casued by dynamic power to figure out how much more power you are producing.