Read wafer maps to identify spatial defect signatures pointing to a specific process step.
Quality & Productivity
Semiconductor Fabrication Yield Improvement and Defect Density Analysis
Analyse wafer fabrication yield loss using defect density models, wafer maps and parametric data, and apply structured root cause and DOE methods to accelerate the yield ramp.
🗓️Duration5 training days
📚Content4 modules · 8 sessions
🎓On completionAccredited attendance certificate
About the programme
Course Overview
Semiconductor fabrication yield determines the economics of every wafer that leaves the fab, and small shifts in defect density or parametric performance can move a process from profitable to loss-making within a single lot. This course teaches participants how to read a wafer map for spatial defect signatures, apply yield models such as the Poisson and negative binomial models to separate random defect loss from systematic loss, and calculate defect density trends across process nodes and tool sets. Sessions cover distinguishing parametric yield loss, where devices are electrically weak but functional, from catastrophic and functional yield loss caused by particle or pattern defects, and tracing excursions back to a specific tool, chamber or process step using inline metrology and statistical process control charts. Later modules apply design of experiments to isolate the process parameters driving a yield gap, and structure a root cause investigation that moves from a wafer map signature to a confirmed mechanism and a corrective action on the tool. The course closes with planning a yield ramp curve for a new process or product, setting interim yield targets, and reporting yield trends to engineering and operations leadership. Participants leave with a defect density tracking template and a yield excursion investigation framework.
Expected Learning Outcomes
Apply Poisson and negative binomial yield models to separate random from systematic loss.
Distinguish parametric, catastrophic and functional yield loss using electrical test data.
Trace a yield excursion back to a specific tool or chamber using inline metrology and SPC.
Design experiments that isolate the process parameters driving a persistent yield gap.
Investigate root cause from a wafer map signature through to a confirmed failure mechanism.
Plan a yield ramp curve with interim targets for a new process or product introduction.
Who Should Attend
Process engineers responsible for yield performance on a specific fabrication process module.
Yield engineers analysing wafer map and parametric data across multiple fab tools.
Device engineers investigating parametric yield loss on new product qualification lots.
Equipment engineers tracing yield excursions back to a specific tool or chamber.
Quality and reliability engineers linking fab yield data to field failure investigations.
Operations managers planning yield ramp targets for new process node introductions.
Course Modules
Select any module to see its sessions and points.
01Reading yield data and defect density models
2 sessions · 8 points
Session 1Interpreting wafer maps and defect signatures
- Identify centre, edge, ring and cluster defect signatures on a wafer map.
- Distinguish random particle defects from systematic pattern or design-related defects.
- Overlay wafer maps from consecutive process steps to localise where a defect was introduced.
- Correlate wafer map signatures with tool, chamber and recipe identification data.
Session 2Applying yield models and defect density trends
- Calculate defect density from inspection and electrical test data across a process node.
- Apply the Poisson yield model to estimate random defect-limited yield for a given die size.
- Apply the negative binomial model when defect clustering makes the Poisson model unrealistic.
- Track defect density trends over time to detect gradual process or tool degradation.
02Distinguishing types of yield loss
2 sessions · 8 points
Session 1Parametric versus catastrophic yield loss
- Separate parametric yield loss, where devices are weak but functional, from catastrophic failure.
- Use electrical parametric test data to locate parameters drifting outside specification.
- Bin functional test failures by failure mode to prioritise the highest-volume loss category.
- Distinguish design-marginality yield loss from process-induced yield loss.
Session 2Detecting and confirming yield excursions
- Set statistical process control limits on defect density and parametric metrics per tool.
- Detect excursions using control chart rules before yield loss becomes widespread.
- Use inline metrology data to narrow an excursion to a specific process step or chamber.
- Confirm an excursion mechanism through targeted wafer sectioning or failure analysis.
03Root cause investigation and design of experiments
2 sessions · 8 points
Session 1Structuring a yield root cause investigation
- Build a fishbone or is/is-not analysis to scope a yield investigation from a wafer map clue.
- Prioritise investigation hypotheses by fab data availability and potential yield impact.
- Coordinate cross-functional investigation teams spanning process, equipment and metrology.
- Confirm root cause with a controlled experiment before implementing a permanent fix.
Session 2Applying design of experiments to yield gaps
- Select process parameters and ranges for a designed experiment targeting a yield gap.
- Run and analyse a fractional factorial or response surface experiment on fab equipment.
- Interpret interaction effects that a one-factor-at-a-time approach would have missed.
- Translate experimental results into a revised process window and control limits.
04Ramping and sustaining yield performance
2 sessions · 8 points
Session 1Planning the yield ramp curve
- Set interim yield targets across the ramp from engineering lots to volume production.
- Sequence yield-limiting issues by expected impact on the overall ramp timeline.
- Align yield ramp milestones with capacity planning and customer delivery commitments.
- Report yield ramp progress and risk to engineering and operations leadership.
Session 2Sustaining yield after ramp completion
- Maintain statistical process control on the parameters most critical to sustained yield.
- Review defect density and parametric trends periodically to catch slow degradation.
- Update the yield excursion investigation framework with lessons from resolved cases.
- Transfer yield learning to sister fabs or future process nodes through documented playbooks.
What the participant receives
4 course modules
A structured syllabus
8 training sessions
across 5 days
32 detailed points
Applied, detailed content
Accredited attendance certificate
On completing the programme
Complete your registration
We will contact you within one business day to confirm.
Ready to start?
Reserve your seat and start building the skill.