Semiconductor Yield Analysis & Process Defect Root Cause Identification
Correlate semiconductor defects to process parameters and drive yield improvements
What You Can Do
You can upload wafer-level defect maps and production logs to systematically identify which process parameters caused yield loss. The skill correlates defect patterns with tool settings (temperature, pressure, timing, material composition) and ranks root causes by likelihood and impact. You'll receive targeted recommendations to adjust specific process parameters and recover yield.
Features
Identify spatial and temporal patterns in wafer defect maps to distinguish random vs. systematic failures
Cross-reference defect locations and types with fab logs (temperature, pressure, dopant levels, etch timing, deposition rate)
Rank multiple hypotheses by likelihood using defect distribution, concentration, and historical process signatures
Categorize defects by production step (lithography, etch, deposition, implant, annealing) to focus investigation
Calculate how each root cause contributes to overall yield loss and prioritize by business impact
Generate specific, measurable parameter changes with expected yield recovery and implementation difficulty
Compare defect profiles across multiple lots to isolate systematic vs. random causes
Identify out-of-control process states using control limits on key parameters and defect trends
Example Output
Root Cause Summary — Lot ABC123 (Week 38 Yield Loss)
Primary Hypothesis: Furnace Temperature Excursion
- Likelihood: 78% (matches thermal failure signature)
- Impact: 8.2% yield loss
- Evidence: 94% of defects clustered in furnace zones 2–4; defect type = dopant diffusion edge effects
Secondary Hypothesis: Dopant Source Contamination
- Likelihood: 15%
- Impact: 1.5% yield loss
Recommended Actions (ranked by ROI):
- Reduce furnace setpoint from 850°C to 830°C; run qualification wafers (expect +7.5% yield, low implementation cost)
- Inspect dopant source vial for particulates; replace if cloudy (expect +0.8% yield, medium cost)
- Increase anneal ramp rate from 2°C/min to 3°C/min to reduce peak time at diffusion-critical zone (expect +0.5% yield, low cost)
Expected Total Yield Recovery: 8.8% → achievable within 1 production cycle
What's Included
- Defect Correlation Engine: Analyzes wafer defect maps and production parameters to identify process root causes
- Root Cause Hypothesis Framework: Systematic methodology for ranking competing hypotheses by likelihood, impact, and feasibility
- Process Signature Library: Reference patterns for common failure modes (thermal, dopant, contamination, etch, lithography)
- Yield Impact Calculator: Quantifies contribution of each identified root cause to overall yield loss
- Recommendation Generator: Produces specific, measurable process adjustments with expected yield recovery and feasibility assessment
- Executive Report Templates: Formats analysis results for fab management and engineering decision-making
Who It's For
- Process Engineers
- Fab Operations Managers
- Yield Engineers
- Quality Control Managers
- Manufacturing Supervisors
Best For
- Investigating unexpected yield loss
- Identifying systematic defect patterns
- Correlating defects with process parameters
- Prioritizing corrective actions by impact
- Supporting continuous process improvement initiatives







