| 摘要 |
PROBLEM TO BE SOLVED: To provide a failure analysis method capable of automatically sorting out the distribution of failures that occur in each wafer by the wafer. SOLUTION: The occurrence positions of failures occurring in a plurality of wafers are inputted. The wafer is partitioned into a plurality of regions. A computation is carried out so as to represent the amount of features of a failure distribution in each region of the wafer by, at least, one or more kinds of numerical values. As to the amount of features, the degree of similarity among the wafers is represented by a first numerical value. For each wafer, other wafers having the first numerical values above a predetermined first threshold value are extracted, and similar wafer groups composed of wafers which are similar to each other in a failure distribution are formed. COPYRIGHT: (C)2005,JPO&NCIPI
|
