Synopsys’ DSSA Sentry™ is a next-generation solution for automatically detecting Defect Spatial Signature patterns on semiconductor wafers. Defect spatial signature patterns can be identified when systematic problems occur within processes or manufacturing equipment. Today, the large volumes of data within the fab and the subtlety of many of the defects make it nearly impossible to find and identify these signatures manually.

Signatures can be indicative of a yield problem that may get progressively worse or an overall systematic issue that has a cumulative affect on yield.

DSSA Sentry can easily find the faint or small signatures that can be missed by manual review of wafer maps.

Low Defect Count Signatures Captured with DSSA Sentry

DSSA Sentry also recognizes cluster details such as shape, length, position on the wafer, angle, etc. The signature data can be used with user-customizable rules to define the classifications. Rotation data may be used to set classifications based on specific angles, or rotation may be ignored to set single classifications for all possible rotation angles. Other parameters, such as layer, inspection recipe and inspection tool can also be used to set classifications.

This flexible, highly configurable system is able to support many classification groups, with very good accuracy and purity values for each group.

DSSA Sentry Runtime Environment

DSSA Sentry Benefits
DSSA Sentry is fab proven and extracts many variations of defect signature maps—far beyond the user's expectations.

DSSA Sentry shortens the time to detect a signature-related yield problem by as many as one to three weeks. The shorter time to detect could easily translate into a significant savings on a per-incident basis for a modest yield excursion. The system can easily pay for itself within six months of operation, frequently less.

Key Features

• Production Proven
• Running in a high-volume production environment since January 2004
• FAST – less than 5 seconds per wafer
• > 80% average accuracy across standard signature types
• Able to extract signatures from noisy wafer maps
• Support for multiple (compound) classifications per wafer
• Reporting and alarming (MES, e-mail, custom alarms)
• Direct connection to customer DDMS Systems (Klarity, Odyssey, etc.)
• Multiple classification algorithms – not dependent on only image-based wafer libraries
• Cluster Signature Analysis (CSA) identifies attributes about clusters detected on the wafer
• Repeaters – detects reticle repeaters
• Wafer Signature Analysis (WSA) compares wafer map to a library of known signatures
• Region Analysis matches defect distribution against 32 different wafer regions.
• Rules Engine uses Layer ID, CSA results, WSA results, repeaters, region analysis and other values to establish the final signature classification for a given wafer
• “Wizard” based editor for setting DSSA rules for each signature classification
• Rotation independent (users can assign rotation dependency to specific signatures, or not)
• Measure DSSA accuracy through confusion matrix

Setup Classifications in DSSA Sentry Manager

Success Stories
Read how a Texas semiconductor fab uses DSSA Sentry in their high-volume manufacturing environment in this Semiconductor International article: http://www.reed-electronics.com/semiconductor/article/CA604355.