Determining safety stock for semiconductor manufacturing

Abstract

Semiconductor manufacturing requires many process steps performed by highly unreliable equipment. Such unreliability causes the means and variances of its cycle times to be generally larger than those of other industries. In addition, it makes the yields uncertain. While making the order quotation and calculating production plan, the safety stock must be used to guard against the uncertainty of the manufacturing processes to increase the on-time-delivery ratio and to provide better customer services. The randomness of cycle times and yields are the two major sources of the output uncertainty. Formerly, Hung [1991] proposed a method capable of obtaining the output lot distribution at a particular time on the planning horizon. The distribution could be obtained by using the production rate and the cycle time distribution which are readily available from the manufacturing database. However, the method does not adequately consider the yield uncertainty. While extending that method, this paper employs a simple variance calculation formula to incorporate the yield uncertainty when computing the variance of total good die output. After calculating this variance for a particular product at a particular time, the necessary safety stock for that die type at that time can be obtained.