Spatial Modeling of Clustered Defects across a Wafer Map

Abstract

The need for accurate yield prediction is increasing for estimating productivity and production costs to secure high revenues in the semiconductor industry. To this end, we introduce new modeling approaches for spatially clustered defects on an integrated circuit (IC) wafer map. The spatial model shows more potential as an yield model by detecting spatial pattern of clustered defects which are commonly observed in an IC wafer map.