Skip to main content...
S0 · CMOS Fundamentals + Digital Logic
30 min

Day 11: Dynamic vs static vs leakage power; the P = αCV²f equation

One equation governs a chip's power budget. Learn where each term comes from and you'll understand every low-power trick in Stage 5.

The power equation

A CMOS chip's power splits cleanly: P_total = P_dynamic + P_short-circuit + P_static. The headline term is dynamic (switching) power, P_dynamic = α · C · V² · f. Every symbol is something you can now reason about — and every one is a lever a real design team pulls.

The dynamic power equation, term by term
P_dynamic = alpha * C * V^2 * f

  alpha (activity factor) : fraction of nodes that switch per clock (0..1)
  C     (capacitance)     : total switched load capacitance
  V     (supply voltage)  : Vdd
  f     (frequency)       : clock rate

Energy per full 0->1->0 cycle of a node = C * V^2
(half is stored on the cap, half dissipated in the transistor each edge)

Where does C·V² come from? Charging a capacitor C to voltage V through the PMOS dissipates ½CV² in the transistor and stores ½CV² on the cap; discharging through the NMOS dissipates the stored half. So one full switching cycle costs CV², times how often it happens (f), times how many nodes actually toggle (α).

Dynamic power vs supply voltage (fixed α, C, f): the V² term means small voltage cuts pay off hugely.

Why voltage is the power lever

Power scales with but only *linearly* with f. Halving the voltage quarters dynamic power; halving the frequency only halves it. That asymmetry is the entire basis of DVFS (dynamic voltage & frequency scaling) in Stage 5 — and why lowering voltage, not just slowing the clock, is how chips save the most energy.

The other terms: short-circuit power is that brief both-on current per transition (kept small by fast edges). Static power is the Stage −1 leakage — negligible on old nodes, but a first-class problem on modern ones, which is why P_static earns its own place in the equation and its own Stage 5 toolbox (multi-Vt, power gating).

Key terms

Activity factor α
The average fraction of nodes toggling per clock cycle; clock nets have α≈1, random logic much less.
Dynamic (switching) power
α·C·V²·f — power spent charging/discharging load capacitance as nodes switch.
Short-circuit power
Power from the momentary both-transistors-on current during input transitions.
Static (leakage) power
Power burned by leakage while idle; dominated by subthreshold conduction (Stage −1).
DVFS
Dynamic voltage & frequency scaling — trading speed for large power savings via the V² term (Stage 5).

Before moving on, you should be able to

A design drops its supply voltage from 1.0 V to 0.7 V (holding α, C, f fixed). Roughly what happens to dynamic power?

We use cookies

We use cookies to enhance your browsing experience, serve personalized content, and analyze our traffic. By clicking "Accept All", you consent to our use of cookies. Learn more

    Day 11: Dynamic vs static vs leakage power; the P = αCV²f equation | RBTechIconX