What is the difference between an NLDM lookup table and a CCS current-source model?
From PDVerse STA Mentor Guide · pdVerse Mentor Guide
Short Answer
NLDM, non-linear delay model, describes a cell's delay and output transition as a lookup table indexed by input transition time and output load capacitance, giving a single number for each combination. CCS, composite current source, instead describes the cell's actual output current over time, letting the tool calculate delay more accurately when the load is not a simple capacitor, such as when crosstalk or a resistive interconnect is involved. Both come from the same Liberty library file format, but CCS carries more detailed information at the cost of a larger library and more calculation.
Technical Explanation
The two models trade simplicity for accuracy on non-ideal loads.
- An NLDM table stores delay and output slew as two-dimensional tables, one axis for input transition and one for output load capacitance, and the tool interpolates between grid points.
- NLDM assumes the load looks like a simple capacitor, reasonable for a short, low-resistance net but less accurate for a long, resistive interconnect or when crosstalk is changing the effective load during switching.
- A CCS model instead describes the driving cell's output current as a function of time, letting the delay engine solve for the actual voltage waveform against whatever real load is connected.
- Because CCS models the driver-load interaction directly, it is what PrimeTime SI relies on to calculate crosstalk delay and noise accurately, which NLDM's simpler capacitive-load assumption cannot do as precisely.
- CCS library files are larger than NLDM files for the same cell, since they store current waveforms instead of two delay numbers per grid point.
Common Mistake
The Trap: assuming NLDM and CCS always produce nearly identical timing results, so it does not matter which one signoff uses.
- On short, low-resistance nets with simple capacitive loads the two models often agree closely, but on long interconnect or SI-sensitive nets the difference can change whether a path passes or fails.
- Mixing NLDM-based delay for some cells and CCS-based delay for others without a deliberate reason produces a report that looks consistent but is comparing apples to oranges.
Follow-up Question & Model Response
If CCS is more accurate, why would a flow ever choose NLDM instead?
Candidate Model Response: NLDM libraries are smaller and calculate faster, which matters for early synthesis and placement runs that go through many timing evaluations quickly and do not yet need SI-level accuracy. CCS becomes important once the design has real interconnect and crosstalk risk to model, typically closer to final signoff, so many flows use NLDM early for speed and switch to CCS-based analysis for the signoff runs where accuracy matters more than runtime. A team should confirm which model each stage of its flow is actually using.
Practical Example
On a net with 30ps of resistive wire delay and an active crosstalk aggressor, NLDM's simple-capacitance assumption reports 210ps of delay, while a CCS-based calculation on the same net reports 245ps, a 35ps difference that comes entirely from CCS correctly accounting for the resistive load shape and the coupling current the aggressor injects.
Complete STA Handbook
Master Signoff-Ready Static Timing Analysis
Get the complete 10-chapter STA handbook covering setup/hold margins, clock modeling, OCV/POCV, crosstalk noise, and PrimeTime closure.
Continue practising