Transistor circuits take time and energy to switch.
The main reason is that real circuit nodes have capacitance.
Capacitance Stores Charge
A capacitor stores charge for a voltage difference.
The simplest useful relation is:
Q = CV
Q is charge, C is capacitance, and V is voltage.
A MOSFET gate behaves partly like a capacitor because an insulating oxide separates the gate from the semiconductor below it.
To change a gate or output voltage, the circuit must move charge.
Delay
Switching is not instant.
An output node must charge toward the supply or discharge toward ground. The available transistor path has resistance-like behavior, and the node has capacitance.
The useful first model is:
More resistance or more capacitance usually means slower voltage change.
This is why transistor size, wire length, fan-out, and layout affect circuit speed.
Dynamic Power
Changing voltage consumes energy.
For CMOS digital circuits, a simplified dynamic power relationship is often described as proportional to:
CV^2f
C is switched capacitance, V is supply voltage, and f is switching frequency.
The exact formula depends on activity and circuit details, but the lesson is durable: switching more capacitance, at higher voltage, more often, costs more power.
Boundary
CMOS is low static power in an ideal simplified model, but it is not free.
Real power includes:
- dynamic switching power
- leakage power
- short-circuit current during transitions
- clock and interconnect costs
The durable model is:
Digital logic has physical cost because changing bits means moving charge through real transistor circuits.