[citation][nom]allenpan[/nom]i dont think the "power" cumsumption umber is wrong, assuming Power = Voltage * Current (P=VI or P=VIcos@) or Voltage ^2 * Resistance (P=V^2R) or Power = Current ^2 / Resistance (P=I^2/R)[/citation]I did Electronics in school, and know Ohm's Law....
...However, the power consumption in a CPU is more logarithmic and does not follow Ohm's Law. Due to various effects, even a 0.1 volt increase can, say, quadruple or septuple power consumption in certain types of digital circuits.
Think silicon breakdown voltage. Say, 0.3 volts for a certain type of diode. Just like when you try to push 0.2 VOLTS into the diode, you get 0 AMPS. But you try to push 0.5 VOLTS into the diode with no current-limiting resistors, you get a SHORT CIRCUIT.