How do you create a repeatable early-congestion estimate?

Short Answer

The whole point of an early congestion check is to compare alternatives fairly — and that's only possible if every alternative is measured with the exact same yardstick. Keep the floorplan version fixed across all comparisons — if you tweak the floorplan between runs, you can no longer tell whether congestion changed because of your real variable or because of an incidental floorplan drift.

Technical Explanation

• Use the same coarse placement settings, the same active scenarios, the same layer availability, and the same PG/blockage assumptions for every alternative you're comparing — any one of these left uncontrolled invalidates the comparison.
• Run global routing in the same mode every time and report using the same settings — a congestion "map" generated with different global-route settings between runs is comparing apples to oranges even if it looks like the same kind of report.
• Know the limits of what this actually tells you: floorplan-stage global routing estimates demand vs. capacity, but it cannot prove your design will actually close detail-route DRC or hit final timing — it's a directional early-warning signal, not a signoff guarantee.
• Bottom line: repeatability comes from controlling every variable except the one you're actually testing, and honesty comes from remembering this is an estimate, not a proof.

Formula Or Decision Rule

Decision rule: archive inputs and settings with the map and report.

What To Check

• Warning sign: Two maps differ because one ignored blockages or used another floorplan checkpoint.
• Inspect: choose one affected region, macro, row, pin, path, or net and trace the physical cause.
• Correct: Reproduce both from the same baseline and rerun the identical route/report sequence.

Healthy, Suspicious & Hard-stop Results

• Healthy: Use the same floorplan version, coarse placement, scenarios, layer availability, PG/blockage assumptions, global-route mode, and report settings for every alternative.
• Hard stop: required legality or physical feasibility is missing or unexplained.

What The Interviewer Is Testing

” to measurable physical evidence and an owned correction.

Follow-up Question & Model Response

"* Candidate Model Response: Model response: “I would change it if the same controlled rerun shows that two maps differ because one ignored blockages or used another floorplan checkpoint. ”