A systematic study of the extended X-factor in relation to effective system capacity

Abstract

More and more productivity is demanded from semiconductor manufacturing lines, thus applying more pressure on the “throughput vs. cycle time” constraint. Cycle time and throughput are traditionally used as independent fine-performance measures. A new school of extended X-factor theory has recently emerged. Proponents of this X-factor contribution measure contend that such a metric, which explicitly takes into account utilization and raw processing time of each machine group, is effective for identifying system capacity constraints. This paper systematically studies the dynamics of a complex semiconductor line to reveal the relationship between the extended X-factor metric and the effective capacity of the line. This paper also studies the sensitivity of the extended X-factor metric to raw processing time and throughput rate to determine quantitatively its robustness and effectiveness when used to evaluate line performance. The analysis shows that the X-factor contribution measure correctly identifies capacity constraining machine groups and is more effective than utilization for identifying these machine groups when the difference in X-factor contribution is significant among the constraining machine groups. Additional capacity at the high X-factor contribution machine group lowered cycle time beyond adding capacity to the highly utilized machine group in a full-scale model. This study provides important insight on when the extended X-factor measure is more indicative of a system capacity constraint, as compared to a utilization measure.