Compression architecture, EDT channels, pattern volume control, diagnosis loops, and yield learning workflows.
Start with the topic hub for architecture and ownership context.
Use mechanism and reports pages to map assumptions to metrics.
Use debug and worked examples for real closure patterns.
Finish with checklist and silicon impact before release decisions.
compression-architectures/ - Compression Architectures
pattern-volume-tradeoffs/ - Pattern Volume Tradeoffs
diagnosis-and-yield/ - Diagnosis & Yield Learning
compression-debug/ - Compression Debug
Compression saves tester time only when diagnosis observability remains credible.
COMPRESSION LOOP
EDT/decompressor -> compressed patterns -> compactor responses -> diagnosisPATTERN vs COVERAGE
coverage up -> pattern count up
compression up -> pattern count down (until aliasing risk)compression ratio dashboard
pattern count trend
X-source report
diagnosis bucket summary
Compactor overflow plus unknown inflation caused false diagnosis; masking policy and channel map corrected.
Separate X issues from silicon defects
Replay failing patterns uncompressed
Track tester memory budget
Ask: what evidence proves this DFT decision is safe for production?