Wafer Map Failure Pattern Recognition Using Artificial Intelligence
Wafer Map Failure Pattern Recognition Using Artificial Intelligence
Wafer Map Failure Pattern Recognition Using Artificial Intelligence
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
Phase 1 : Data preparation (2)<br />
• Some example of useful data<br />
–To detect system failure, trouble spots pattern<br />
Input Detection Verification Output<br />
1. X‐Y coordinates<br />
2. Soft Bins<br />
Soft Bins<br />
1. Start with each<br />
failure die as seed.<br />
2. Cluster failing dies to<br />
determine if there is<br />
grouping beyond set<br />
a threshold.<br />
–To detect “Known Recoverable Bins”<br />
1. Repeated signature<br />
in some wafers,<br />
adjacent wafers.<br />
2. Independent of<br />
tester or H/W used,<br />
cluster will still fail.<br />
Cluster failure True<br />
/ False?<br />
Quadrant location?<br />
Input Detection Verification Output<br />
Recoverable list of soft<br />
bins<br />
Recovery history on<br />
different wafers, first<br />
probe, online reprobe<br />
etc.<br />
Probability of<br />
subjecting to retest<br />
10
Change language