Post-CTS Hold Fixing Workflow — Reports & Metrics

On-call / interview prompt

Which report lines prove Post-CTS Hold Fixing Workflow is healthy vs failing?

CLOSURE CHAIN

1. METRIC     — name the failing report line (WNS, DRV, DRC count, IR %)
2. HYPOTHESIS — 2–3 likely causes ordered by probability
3. EXPERIMENT — one cheap check (corner, clock, report, map)
4. FIX        — minimal physical or constraint change
5. REGRESSION — what you re-run and what must not regress

Reports to inspect

• Hold violation histogram by magnitude and domain

• Common-path pessimism and clock path decomposition report

• Hold fix ECO summary (buffers/delay cells inserted)

• Setup regression report after hold repair

POST-CTS HOLD TRIAGE
total_hold_violations: 2847
violations_lt20ps: 1936
violations_gt80ps: 211
dominant_domain: CLK_APB
added_hold_buffers: 672
setup_regression_paths: 39
action: split fix pass by severity tiers; reserve skew tweak for >80ps bucket

Smoke check (5 minutes)

• Can you name the single worst line in the report?

• Can you tie that line to a physical region on the floorplan or layout?

How to read this like a signoff lead

The report is not a pass/fail artifact; it is a prioritization tool. Read Clock Tree Synthesis closure dashboard by severity, locality, trend, and fix cost before touching the design.

Report triage order

1. Confirm database tag, corner, mode, extraction state, and analysis switches.

2. Separate hard gates from tracked risks and cosmetic noise.

3. Cluster failures by physical region, hierarchy, clock domain, power domain, or rule class.

4. Compare against previous tag to identify new regressions, not just absolute failures.

5. Translate the worst line into an owner, experiment, and rollback plan.

SENIOR REPORT READOUT
  worst_line: <copy exact report line>
  cluster: <region/domain/rule/path-group>
  delta_from_previous: <new/worse/better/same>
  first_experiment: <cheap evidence-gathering action>
  decision: <fix now / assign owner / waive with approval / stop release>

Deep dive: how this shows up in real closure

CTS changes the timing problem by making clocks real.

Reports and artifacts to inspect

• clock tree summary: skew, latency, buffer count, sink count

• clock transition and capacitance violations

• post-CTS setup/hold WNS by clock domain

• ICG enable timing and test-mode clock exceptions

Mini case study

Pre-CTS setup is green, but post-CTS hold WNS is -60 ps on short register-to-register paths. This is expected: clock arrival spread is now real. Add data delay or adjust skew carefully, then check setup regression.

Debug branches

• If skew is high, inspect clock root, macro stops, NDR, and buffer placement channels.

• If hold explodes, classify short paths by domain and add targeted delay, not global padding.

• If ICG enable fails, debug enable path timing separately from clock skew.

Senior review question

Ask yourself: what single report line would prove this page's concept is either passing or failing?

What changes at 10+ years

• You are expected to predict what your fix can break before running it.

• You should recognize when the issue is methodology, not one block's implementation.

• You should communicate risk in tapeout language: owner, evidence, impact, mitigation, and decision date.

Principal-level review bar

Deep subpage 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

CTS changes the timing problem by making clocks real. 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.