MySTDF | STDF Conversion and Engineering Intelligence
Convert STDF at Scale. Analyze with Confidence.
MySTDF converts large STDF and archive files into structured, analysis-ready outputs, connects the converted data to wafer and yield analytics, and assists engineers with root-cause evidence, recommended actions, and automated reports.
Efficient local processing connects STDF conversion directly to wafer analysis, yield investigation, site comparison, AI-assisted root-cause analysis, and automated reporting.
- Site distribution shiftHigh
- Continuity tail behaviorMedium
- Edge-ring wafer patternMedium
STDF Conversion
Convert STDF Files into Analysis-Ready Outputs
MySTDF starts by converting raw STDF or archive files into structured outputs for downstream engineering analysis. The transfer engine is designed for large production files and uses efficient local processing without loading the complete dataset into memory.
- STDF and compressed archive input
- Chunked processing for large files
- Per-file SQLite database output
- Optional Excel output
- Optional automated PPT reporting
- Local processing without cloud upload
- Transfer progress and detailed status log
- Portable Windows deployment
- No Python installation required for the packaged release
Engineering-Ready Excel Output
STDF to Excel: Engineering-Ready Data, Not a Raw Dump
MySTDF converts STDF records into an engineering-oriented Excel workbook that combines test-item statistics, device-level metadata, and individual measurement values. Highlighted cells help engineers quickly identify failed or flagged results and trace them back to the corresponding device, site, bin, coordinate, and first-fail information.
Test-Level Statistics
Review limits, fail rate, Cpk, sigma, mean, percentiles, minimum, maximum, and device count for every exported test item.
Device-Level Traceability
Trace every value to its device number, site, hard bin, soft bin, wafer coordinate, first fail, last fail, and pass/fail status.
Fast Failure Localization
Highlighted results allow engineers to move quickly from an abnormal value to the affected test item and device without manually searching the complete workbook.
Conversion Pipeline
Convert Once. Reuse the Data Across Engineering Workflows.
The structured outputs become the reusable data foundation for TSR analysis, wafer maps, site comparison, bin analysis, correlation, project trends, and automated reporting.
Engineering Analytics
One Test. Multiple Engineering Views.
MySTDF helps engineers understand one test through an integrated workspace that combines fail ranking, wafer map behavior, spatial distribution, zone trends, by-site comparison, and correlation-based pattern review.
Fail Ranking Table
Automatically prioritize failing items using fail percentage, AI classification, test type, limit behavior, discrete behavior, outlier signatures, and heavy-tail detection.
Map & Distribution
Review wafer maps, parametric distributions, zone trends, limits, and by-site behavior for fast spatial and statistical verification.
Correlation & Pattern
Identify related tests, high-correlation candidates, upstream behavior, and repeated engineering patterns.
Detailed Evidence
Explore the Detailed Engineering Evidence
Open the detailed analysis view to inspect fail patterns, parametric wafer behavior, center-to-edge trends, cumulative distributions, specification limits, and site variation.
Project-Level Visibility
Project-Level Yield and Risk Visibility
Review lots, wafers, yield, hardware-bin loss, first-fail behavior, TSR results, and project trends before opening detailed per-file analysis.
AI Decision Summary
From Project Data to Decision-Ready Engineering Insight
MySTDF combines project yield, first-fail behavior, hardware-bin loss, ranked risk items, and engineering evidence into one decision-oriented summary. AI assists by organizing the evidence and prioritizing candidates while engineers retain control of the final conclusion.
MySTDF Decision Workspace
AI-Assisted Project Summary
Yield Trend and First-Fail Distribution
8 demo wafersYield trend and first-fail distribution help engineers identify unstable wafers and dominant loss contributors.
Top Risk Items
Ranked risk| Rank | First Fail | HB | Test | FF % |
|---|---|---|---|---|
| 1 | FF_001 | HB187 | DEMO_SUPPLY_CURRENT | High 3.58% |
| 2 | FF_002 | HB156 | DEMO_OUTPUT_LEVEL | Medium 3.16% |
| 3 | FF_003 | HB156 | DEMO_CONTINUITY_POS | Medium 2.58% |
| 4 | FF_004 | HB42 | DEMO_ADC_LINEARITY | Lower 0.67% |
| 5 | FF_005 | HB11 | DEMO_RF_GAIN | Lower 0.60% |
Key Insight
Illustrative AI-Assisted Summary- Yield recovered after an initial dip and is stable around 82-90%.
- Nine first-fail items were detected; the top three contribute most of the observed loss.
- Prioritize the highest-risk tests and verify their wafer, site, and hardware-bin evidence.
Wafer Performance Snapshot
Heatmap-style demo values| Wafer | Yield % | HB156 | HB187 | HB10 | HB42 | HB140 | Weighted Avg. |
|---|---|---|---|---|---|---|---|
| DEMO-W01 | 78.1% | High 3.9 | High 4.1 | Mod 2.7 | Low 0.6 | Mod 1.8 | 2.62 |
| DEMO-W02 | 79.4% | Mod 2.8 | High 3.4 | Mod 2.1 | Low 0.5 | Mod 1.5 | 2.06 |
| DEMO-W03 | 82.0% | Mod 2.2 | Mod 2.4 | Low 1.0 | Low 0.4 | Low 0.9 | 1.38 |
| DEMO-W04 | 83.1% | Low 1.4 | Mod 1.9 | Low 0.8 | Low 0.3 | Low 0.7 | 1.02 |
| DEMO-W05 | 84.0% | Low 1.2 | Mod 1.6 | Low 0.7 | Low 0.2 | Low 0.6 | 0.86 |
| DEMO-W06 | 84.5% | Low 1.1 | Low 1.3 | Low 0.6 | Low 0.2 | Low 0.5 | 0.74 |
First-Fail Detail
Short public demo table| First Fail | HB | Test | FF % |
|---|---|---|---|
| FF_001 | HB187 | DEMO_SUPPLY_CURRENT | 3.58% |
| FF_002 | HB156 | DEMO_OUTPUT_LEVEL | 3.16% |
| FF_003 | HB156 | DEMO_CONTINUITY_POS | 2.58% |
| FF_004 | HB42 | DEMO_ADC_LINEARITY | 0.67% |
| FF_005 | HB11 | DEMO_RF_GAIN | 0.60% |
Yield, first fail, and AI grouping
First-fail, hardware-bin, correlation, and pattern analysis
Wafer maps, site behavior, distribution, and HB heatmap
Focused verification and next engineering steps
Data-driven decision support for higher yield and faster time to resolution.
This public demo uses synthetic data to illustrate the MySTDF decision workflow.
AI Root-Cause Evidence
From Correlation to Root-Cause Evidence
MySTDF ranks upstream candidates using correlation, separation, lift, test role, and multi-view verification. Engineers can compare statistical relationships, distribution behavior, and supporting evidence before selecting the most likely root-cause candidate.
AI organizes and ranks the evidence. Engineers review the relationships, verify the supporting plots, and retain control of the final conclusion.
AI Root-Cause Summary
Summarize the target failure, statistical evidence, abnormal tail behavior, candidate type, investigation scope, and recommended next check.
- Target failure summary
- Outlier and tail evidence
- Candidate type and role
- Recommended next step
Upstream Evidence Ranking
Rank upstream tests using correlation, separation, lift, role, and test type so engineers can focus on the strongest root-cause candidates first.
- Correlation strength
- Distribution separation
- Lift score
- Device / program / monitor role
Multi-View Verification
Verify each candidate using correlation plots, normalized probability plots, distribution shape, specification behavior, and consistency across multiple perspectives.
- Relationship strength
- Tail and pattern behavior
- Specification-limit behavior
- Cross-view consistency
This public demo uses synthetic and anonymized data to illustrate the MySTDF evidence-ranking workflow.
Engineering Intelligence
From Engineering Evidence to Root-Cause Candidates
AI assists the investigation by organizing evidence, detecting patterns, and ranking root-cause candidates. Engineers retain visibility into the data, logic, and final decision.
Evidence
Correlation and Pattern Detection
- Wafer, site, bin, and parametric verification
- Upstream evidence ranking
- Repeated pattern and outlier review
Confidence
Root-Cause Candidate Confidence
- Candidate confidence scoring
- Supporting evidence for each candidate
- Traceable engineering assumptions
Decision Support
Recommended Next Checks
- Recommended next check
- Engineer-controlled final decisions
- Focused paths into automated reporting
Auto Report
Decision-Ready Output, Not Just Exported Charts
The report structure follows the way engineers communicate a problem: issue, root-cause candidate, supporting evidence, confidence, and recommended action.
- Issue summary
- Root-cause candidate and why
- Supporting evidence
- Confidence
- Recommended action
- Excel or presentation-ready output
Issue Summary
Yield loss is concentrated on one site with a repeated edge-ring wafer signature.
Root-Cause Candidate
Primary candidate: site-dependent contact or hardware path instability.
Supporting Evidence
Recommended Action
Verify the affected contact path, compare continuity distributions, and review impacted samples.
Adaptive Learning Engine
User-Driven Intelligence Without Constant Retraining
Engineering teams can keep classification and investigation logic current through durable configuration instead of rebuilding a large model for every mapping update.
Map continuity, supply, scan, analog, IDDQ, RF, and custom test families to engineering roles.
Case-insensitive matching can evolve as naming styles, test programs, and product families change.
Reusable rules remain available for future analyses and automatically refresh new results.
Capture site, bin, wafer, parametric, and correlation behavior in repeatable investigation logic.
Future analyses can reflect updated mappings without manual report rewrites.
Teams can update mappings and rules without rebuilding a model for each change.
Latest Release
Download the Latest MySTDF Release
Convert STDF files locally, generate structured analysis outputs, and explore the MySTDF engineering workflow without building custom scripts.
Portable Windows package | No Python installation required
Contact
Discuss a MySTDF use case.
Product inquiries, technical discussions, collaboration, and MySTDF demonstration requests.