Physical Design · All levels
PD Cross-Topic Interview Q&A
25 senior questions spanning floorplan through tapeout.
Cross-topic bank
Mix floorplan, CTS, timing, power, PV, and tapeout questions — as in staff panels.
Block fails setup at WC — first three reports?
[INT][PD][X-TOPIC]
Q: Block fails setup at WC — first three reports?
A:
WNS path report @ WC mode, clock latency/skew, net vs cell delay split.
FOLLOW-UP TRAP: Open optimizer without corner name.Hold fails after CTS only?
[INT][PD][X-TOPIC]
Q: Hold fails after CTS only?
A:
Hold corner report, short data paths, clock skew on capture — data delay or skew fix.
FOLLOW-UP TRAP: Blame placement before CTS review.Congestion + timing both red?
[INT][PD][X-TOPIC]
Q: Congestion + timing both red?
A:
Fix congestion first (util, macro channel) — timing opt on unroutable net useless.
FOLLOW-UP TRAP: Size cells in red zone.IR fail but timing pass?
[INT][PD][X-TOPIC]
Q: IR fail but timing pass?
A:
Strengthen grid locally — effective Vdd drop may soon fail timing — fix IR then re-STA.
FOLLOW-UP TRAP: Waive IR as unrelated.Macro channel floorplan debate?
[INT][PD][X-TOPIC]
Q: Macro channel floorplan debate?
A:
Trade util vs route — widen channel or rotate macro — narrate WNS/TNS impact estimate.
FOLLOW-UP TRAP: Shrink channel for util only.When shrink die?
[INT][PD][X-TOPIC]
Q: When shrink die?
A:
When util target met with clean congestion and signoff margin — not before route clean.
FOLLOW-UP TRAP: Die shrink to save cost mid-route crisis.LVS fail day before tapeout?
[INT][PD][X-TOPIC]
Q: LVS fail day before tapeout?
A:
Stop export — classify open/short — no waiver on connectivity — hot fix scope.
FOLLOW-UP TRAP: Tapeout with LVS fail.500 DRC remaining?
[INT][PD][X-TOPIC]
Q: 500 DRC remaining?
A:
Cluster by rule — fix top systematic — waiver only with approval.
FOLLOW-UP TRAP: Manual one-by-one without clustering.Post-mask ECO scope?
[INT][PD][X-TOPIC]
Q: Post-mask ECO scope?
A:
Metal-only per foundry — confirm cell rules — new tag — LVS/STA on ECO.
FOLLOW-UP TRAP: Add cells when forbidden.SDC change who owns?
[INT][PD][X-TOPIC]
Q: SDC change who owns?
A:
RTL + PD + STA lead — version tag — re-CTS if clocks change.
FOLLOW-UP TRAP: PD edits SDC silently.Useful skew interview answer?
[INT][PD][X-TOPIC]
Q: Useful skew interview answer?
A:
Setup help on critical path — cite hold risk on neighbors — structured budget.
FOLLOW-UP TRAP: Skew everywhere.Clock gating enable fail?
[INT][PD][X-TOPIC]
Q: Clock gating enable fail?
A:
Enable path timing — ICG placement — test mode override.
FOLLOW-UP TRAP: Only data path fix.Antenna after ECO?
[INT][PD][X-TOPIC]
Q: Antenna after ECO?
A:
Re-run antenna — metal change invalidates prior pass.
FOLLOW-UP TRAP: Skip re-check.Fill inserted — what regresses?
[INT][PD][X-TOPIC]
Q: Fill inserted — what regresses?
A:
Density DRC, extraction on nearby critical nets, optional SI.
FOLLOW-UP TRAP: Fill only PV team problem.Multi-mode signoff?
[INT][PD][X-TOPIC]
Q: Multi-mode signoff?
A:
Matrix corners × modes — worst cell per metric — document.
FOLLOW-UP TRAP: Func mode only.UPF isolation missing symptom?
[INT][PD][X-TOPIC]
Q: UPF isolation missing symptom?
A:
X on inputs when domain off — functional fail — boundary cell check.
FOLLOW-UP TRAP: Debug as timing only.Staff vs senior PD difference?
[INT][PD][X-TOPIC]
Q: Staff vs senior PD difference?
A:
Staff: cross-block tradeoffs, signoff judgment, risk docs, methodology.
FOLLOW-UP TRAP: More tool commands.Explain closure to VP in 60s?
[INT][PD][X-TOPIC]
Q: Explain closure to VP in 60s?
A:
Dashboard: timing/PV/power green or waived with approval — one top risk.
FOLLOW-UP TRAP: Technical deep dive without status.Biggest PD schedule killer?
[INT][PD][X-TOPIC]
Q: Biggest PD schedule killer?
A:
Late floorplan change, under-sized grid, ignoring hold post-CTS — pick one with story.
FOLLOW-UP TRAP: Blame tool only.RTL vs PD fix for setup?
[INT][PD][X-TOPIC]
Q: RTL vs PD fix for setup?
A:
Pipeline/RTL if arch allowed; else physical opt — justify with path report.
FOLLOW-UP TRAP: Always RTL or always buffer.Hierarchy reuse block?
[INT][PD][X-TOPIC]
Q: Hierarchy reuse block?
A:
Frozen pin abstract — soft macro handoff — parent integration checklist.
FOLLOW-UP TRAP: Change pins without parent notice.Scan mode timing?
[INT][PD][X-TOPIC]
Q: Scan mode timing?
A:
Separate SDC overlay — test clocks — signoff row in matrix.
FOLLOW-UP TRAP: Forget scan in signoff.Crosstalk on clock net?
[INT][PD][X-TOPIC]
Q: Crosstalk on clock net?
A:
Shield/NDR/spacing — glitch risk on clock — priority fix.
FOLLOW-UP TRAP: Ignore SI on clock.Tapeout tag mismatch?
[INT][PD][X-TOPIC]
Q: Tapeout tag mismatch?
A:
Hard stop — reconcile STA/GDS/SDC tags before export.
FOLLOW-UP TRAP: Ship wrong database.Why PD-aware RTL matters?
[INT][PD][X-TOPIC]
Q: Why PD-aware RTL matters?
A:
Long combos, bad hierarchy, no CG — limits closure — link rtl/senior/physical-awareness.
FOLLOW-UP TRAP: PD fixes all RTL mistakes.How to drill this Q&A bank
Use each question as a two-minute mock. The target answer is not an essay; it is a structured closure response with a metric, mechanism, risk, and follow-up check.
Answer template
MECHANISM: what physical effect or tool step is involved
WHEN: where it appears in the flow
PITFALL: one wrong junior answer
CHECK: report/map/checklist item that proves the answerScoring
5/5: metric, mechanism, experiment, regression, and release decision named without prompting.
4/5: technically correct and names regression, but misses ownership or escalation criteria.
3/5: correct concept but no decision framework.
1/5: tool command or buzzword with no mechanism.
Principal-level review bar
Q&A bank pages in this course should be read like real closure review material. For a 10+ year PD engineer, the bar is not remembering terminology; it is making a release-quality decision under ambiguity.
What excellent looks like
Names the failing metric, corner/mode, database tag, and analysis switches before proposing a fix.
Separates data, constraint, physical, tool, and methodology root causes instead of treating all failures as optimization problems.
Chooses experiments by information gain and reversibility, not by habit.
States regression blast radius across timing, route, power, PV, DFT, package, and tapeout manifest.
Turns recurring failures into methodology guardrails, dashboards, or checklist items.
Closure note template
STAFF / PRINCIPAL CLOSURE NOTE
Context:
stage: <pre-CTS | post-CTS | post-route | post-fill | signoff>
tag: <database / netlist / SDC / library stack>
failing metric: <exact report line>
affected scope: <block / hierarchy / path group / power domain / region>
Hypotheses:
H1: <most likely physical or constraint mechanism>
H2: <competing explanation>
H3: <methodology or input-data issue>
Decision:
next experiment: <cheap check that can falsify H1>
fix candidate: <minimal reversible change>
rollback trigger: <metric that says the fix is wrong>
regression set: <timing / route / power / PV / DFT / package>
escalation owner: <team or reviewer>Tradeoffs a senior engineer must discuss
Technical tradeoff
Interview performance is structured closure thinking under time pressure. Explain not only the preferred fix, but what margin or schedule you are spending to get it.
Cross-team tradeoff
What must RTL, synthesis, CAD, STA, DFT, package, IP, or foundry agree to before this decision is final?
Which artifact becomes the source of truth after the decision: report, waiver, manifest, ECO script, or methodology deck?
What is the cost of being wrong: one rerun, ECO churn, mask risk, performance loss, or silicon escape?
Leadership communication
"The current blocker is <metric> in <corner/mode/stage>. The leading cause is <mechanism>. I recommend <fix> because it is bounded and reversible. The regression surface is <domains>. If it fails, we escalate to <owner> with <evidence>."Key takeaways
Always connect the concept back to a measurable signoff artifact.
A fix is not complete until you can name the regression checks.
Common pitfalls
Optimizing by habit instead of reading the current report.
Forgetting that a local fix can regress timing, routing, power, or PV elsewhere.