What does the si_enable_analysis variable actually turn on in PrimeTime?

Short Answer

Setting the si_enable_analysis variable to true tells PrimeTime to compute crosstalk delta delay for every net during timing analysis, using the coupling capacitance in the annotated parasitics, instead of timing each net as if it had no neighbors. It is a global switch: once set, it applies to the whole analysis, not to one net or one path at a time.

Technical Explanation

This variable is the entry point into the whole signal-integrity flow, so it is worth being precise about what it does and does not do on its own.

• It is set before linking or reading parasitics, with set_app_var si_enable_analysis true (PT), which is why the PrimeTime User Guide lists it as an early setup step, alongside the multivoltage and analysis-mode variables.
• It switches on delta delay computation, meaning the tool now looks at each net's coupling capacitance to nearby nets and computes an extra delay term, instead of using only the net's own driver and receiver capacitance.
• It depends on the parasitics actually carrying coupling data. The variable alone does nothing useful if parasitics were read in without preserving coupling — the data has to already include net-to-net capacitance, which the read-in step controls separately.
• It does not by itself turn on noise analysis. Crosstalk delay and static noise analysis are related but separately invoked; enabling this variable affects delay calculation, and noise analysis has its own setup steps and commands.
• It is a session-wide setting, not a per-net one. There is no such thing as "SI on for this net, off for that one" through this variable — narrowing scope, when needed, is done through separate include/exclude commands layered on top.
• Why it matters: forgetting to set it, or setting it after parasitics are already linked in a way that dropped coupling data, produces a timing run that silently skips crosstalk delay — no error, just a report that looks clean because it never checked.

Follow-up Question & Model Response

If si_enable_analysis is set to true but the design has no coupling capacitance data at all in its parasitics, what happens?

Candidate Model Response: The analysis runs, but effectively does nothing extra — with no coupling capacitance values present, there is no data for the delta delay calculation to act on, so every net's delay comes out the same as it would with SI turned off. This is a case where a run can look identical to a non-SI run for a reason that is easy to miss: it is not that SI found no violations, it is that SI had nothing to evaluate. I would confirm coupling data is present, for example by spot-checking a known tightly-coupled net pair with report_annotated_parasitics (PT), before trusting a clean SI-enabled report.