Graph Traversal & Slack: Expanded Case Study

Extended case study

Signoff review: WNS, TNS, path slack distribution, failing endpoint count fails after an apparently minor change affecting Graph Traversal & Slack.

Background

Team had closure last week on full MMMC matrix. New run shows regression clustered in one path group.

Symptoms observed

• WNS, TNS, path slack distribution, failing endpoint count regression

• Disagreement between synth and STA WNS

• Hold violations after setup ECO

Investigation timeline

1. Hour 0: freeze netlist, SDC, SPEF, lib tags

2. Hour 1: diff constraints and run manifest

3. Hour 2: isolate corner/mode that fails

4. Hour 3: read worst path — cell vs net, clocks

5. Hour 4: classify root cause

6. Hour 5: minimal ECO or SDC fix

7. Hour 6: full regression + signoff memo

Root cause

Root cause tied to Graph Traversal & Slack: STA builds a timing graph of cell and net arcs; setup slack = required time − arrival time along the worst path.

Fix and validation

• Bounded ECO or audited SDC change

• Re-run report_timing -path_type full, slack histogram, path schematic

• Full MMMC + SI regression

Lessons learned

• Tag every run

• Diff constraints before blaming PD

• Setup and hold together

CASE STUDY — Graph Traversal & Slack
baseline WNS / regressed WNS / post-fix WNS

Sequence under stress

TIMING PATH — Graph Traversal & Slack

launch FF --logic stages--> capture FF
     |                           |
  launch clock                capture clock
     |                           |
  required: arrival < capture_edge - setup
  metric: WNS, TNS, path slack distribution, failing endpoint count

STA deep dive

Slack is a path story — read the graph, not just the number.

Concept diagram

TIMING GRAPH

FF --cell-- logic --net-- logic --cell-- FF
launch                              capture

Metric graph

SLACK TAIL

paths |    ***
      |  *******
      |*********  <- fix WNS first

Reports and artifacts

• WNS/TNS

• path slack histogram

• cell vs net %

• path group summary

Mini case study

WNS -90ps but 80% net delay on one congested route — PD route fix beat upsizing three cells.

Debug branches

• Cell-heavy -> sizing/VT

• Net-heavy -> route/layer

• Check launch/capture clocks

Senior review question

Ask: what corner/mode/view proves this topic is closed or failing?

Key takeaways

• State corner, mode, view, and database tag with every slack claim.

• Run setup and hold plus MMMC regression after every ECO.

Common pitfalls

• Comparing STA runs with different SPEF or SDC tags.

• Broad false_path to green-wash violations.

• Setup-only ECO without hold check.

Principal STA review addendum

STA builds a timing graph of cell and net arcs; setup slack = required time − arrival time along the worst path.

Metric: WNS, TNS, path slack distribution, failing endpoint count