How does the static/leakage power formula translate into an actual placement-stage optimization technique?
From PDVerse PnR Interview Handbook · pdVerse Mentor Guide
Short Answer
Static (leakage) power: Ps = V * sum(Ij) -- supply voltage times the sum of per-component leakage currents, characterized per cell in the library. The practical optimization: replace low-Vt cells on non-critical paths with high-Vt cells, since high-Vt cells leak less. Excess static power is a limiting factor in high-performance deep-submicron CMOS, which is exactly why placement algorithms need to be leakage-aware, not just timing- and congestion-aware.
Technical Explanation
- Static (leakage) power: Ps = V * sum(Ij) -- supply voltage times the sum of per-component leakage currents, each characterized per cell in the library.
- The practical placement-stage optimization: replace low-Vt cells on NON-CRITICAL paths with high-Vt cells, since high-Vt cells leak less current.
- This only works on non-critical paths because high-Vt cells are also slower -- swapping a critical-path cell to high-Vt would trade leakage savings for a timing violation.
- Excess static power is a real limiting factor in high-performance deep-submicron CMOS -- exactly why placement algorithms need to be explicitly leakage-aware, not treat power as purely a post-placement cleanup concern.
set_multi_vth_constraint -low_vt_percentage <pct>is the real command implementing the low-Vt-to-high-Vt swap-on-non-critical-paths technique described here.
Common Mistake
The Trap: Swapping cells to high-Vt for leakage savings without checking whether they're on a critical path first -- the technique specifically depends on slack margin being available to absorb the speed penalty.
Follow-up Question & Model Response
"Why would this technique specifically require the placement tool to already know which paths are critical and which aren't?"
Candidate Model Response: Because swapping indiscriminately (without slack awareness) risks converting a currently-passing critical path into a violating one -- the technique's entire safety depends on only touching cells where the resulting slower delay still leaves acceptable margin, which requires real, up-to-date timing/slack information during the swap decision.
Practical Example
Debug Scenario: A design with excess static power shows several standard cells on well-margined, non-critical paths still using low-Vt variants. Swapping those specific cells to high-Vt equivalents (leaving critical-path cells untouched) reduces leakage power without introducing new timing violations.
Physical Design & Planning Handbook
Master ASIC Physical Design Planning & Floorplanning
Dive into 14 comprehensive chapters covering netlist sanity, FinFET grids, macro placement, power grids, CTS, and timing budgeting.
Continue practising