Integrating sequential process steps has proven helpful particularly in semiconductor manufacturing to increase process reliability. In addition, clustering of multiple process modules or chambers has made parallel processing possible. The results are increased manufacturing productivity through lower cycle times and/or greater throughput. However, due to cluster complexity it has proven to be difficult, even impossible to quantify the throughput of such integrated equipment using conventional methods. Simulation models of these tools provide a means to evaluate the various performance characteristics as a function of application specific process and equipment parameters.
This paper examines models for an integrated vacuum chamber tool and a chemical mechanical polishing (CMP) cluster. Parallel processing can take place within these tools at either the wafer or cassette level. Robot response can be insufficient to prevent product wait times at some locations within the integrated equipment or cluster, which causes the relationship of process time to cycle time to become nonlinear. Using simulation models to analyze the effect of changing chamber assignments or component availability can prove to be useful in avoiding bottlenecks in the processing sequence. In addition, because parallel processing equipment can accept more than one cassette of wafers at varying time intervals, the capacity of the tools is no longer inversely related to the cycle time. Conventional tool planning methods become inadequate as some of the queuing time for each cassette of wafers is spent inside the tool. Static throughput models cannot handle the dynamic behavior of the tool in combination with the dynamic behavior of product arrival in a typical manufacturing line.
Flexible simulation models are presented as a tool for evaluating throughput and cycle time for complex integrated equipment and clusters. The models described here are part of a group of standardized, PC based, and user-friendly capacity models intended for use by manufacturing and process engineers and technicians.