Abstract
A semiconductor supply network involves many expensive steps, which have to be executed to serve global markets. The complexity of global capacity planning combined with the large capital expenditures to increase factory capacity makes it important to incorporate optimization methodologies for cost reduction and long-term planning. The typical view of a semiconductor supply network consists of layers for wafer fab, sort, assembly, test and demand centers. We present a two-stage stochastic integer-programming formulation to model a semiconductor supply network. The model makes strategic capacity decisions, (i.e., build factories or outsource) while accounting for the uncertainties in demand for multiple products. We use the model not only to analyze how variability in demand affects the make/buy decisions but also to investigate how the correlation between demands of different products affects these strategic decisions. Finally, we demonstrate the value of incorporating demand uncertainty into a decision-making scheme.
Original language | English (US) |
---|---|
Pages (from-to) | 322-332 |
Number of pages | 11 |
Journal | International Journal of Production Economics |
Volume | 134 |
Issue number | 2 |
DOIs | |
State | Published - Dec 2011 |
Keywords
- Production planning
- Semiconductor manufacturing
- Stochastic programming
- Supply network capacity
ASJC Scopus subject areas
- Business, Management and Accounting(all)
- Economics and Econometrics
- Management Science and Operations Research
- Industrial and Manufacturing Engineering