How do you test floorplan sensitivity to cell-area growth?

Short Answer

" — this is about turning that worry into a number, not a guess. Take the current counted standard-cell area and apply the growth assumption (an RTL change, a late feature add, a resynthesis with a less-friendly library). Example: 600,000 um² growing 8% becomes 648,000 um².

Technical Explanation

• Example: 600,000 um² growing 8% becomes 648,000 um².
• Recompute the placeable area needed at your target utilization: area / U. 60, that 648,000 um² of cells needs 1,080,000 um² of placeable core area — not 1,000,000 um² as it would have been pre-growth.
• Rebuild (or resize) the candidate floorplan to that larger placeable-area target, keeping macro placement, aspect ratio, and PG planning consistent with the original intent.
• Rerun the same evaluation you'd run on any floorplan candidate: rows/legality, congestion (placement + trial route), timing (still meets the same constraints at the larger size), and PG capacity (does the grid still deliver current at the new demand).
• Compare like-for-like against the pre-growth baseline using identical methodology — this is what tells you whether the floorplan has margin or whether it needs to grow, add routing layers, or shrink logic before growth actually happens.

Formula Or Decision Rule

Anew = Aold x (1 + growth); required = Anew / U.

What To Check

• Warning sign: The plan has no continuous space for added buffers even though global whitespace exists.
• Inspect: choose one affected region, macro, row, pin, path, or net and trace the physical cause.
• Correct: Reshape or lower occupancy where growth must occur, then repeat the sensitivity run.

Healthy, Suspicious & Hard-stop Results

• Healthy: Recalculate required placeable area with the larger counted cell area, rebuild or resize the candidate, and compare rows, congestion, timing, and PG capacity using the same method.
• 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 the plan has no continuous space for added buffers even though global whitespace exists. ”