Abstract
A cluster tool is the place where wafers are processed. To increase the productivity of wafers, semiconductor manufacturers would like to maximize the utilization of cluster tools. However, a cluster tool may not achieve its optimal utilization rate due to the schedule conflict under different combinations of process recipes. By exploiting the data in the log files of a cluster tool, we may find out the reasons of low utilization rate. In this paper, we propose an approach of cluster tool performance analysis to automatically analyze the root cause of low utilization rate in cluster tools. Experimental results show the effectiveness of the proposed approach.