Developing A Discrete-Event Simulation Model For Semiconductor Supply Chain

• Main Author: Lim, Si Ying
• Format: Monograph
• Language: English
• Published: Universiti Sains Malaysia 2022
• Subjects: T Technology; TJ Mechanical engineering and machinery
• Online Access: http://eprints.usm.my/55710/1/Developing%20A%20Discrete-Event%20Simulation%20Model%20For%20Semiconductor%20Supply%20Chain_Lim%20Si%20Ying.pdf; http://eprints.usm.my/55710/
• Institution: Universiti Sains Malaysia

Due to the volatility of demand of integrated circuits (ICs), it is vital to have master planning activity for the manufacturing supply chain to forecast the demand. Production decisions and production planning are based on the demand forecast. With accurate forecasting result, the benefits could be found in the reduction of inventory cost, improvement of order fulfilment, high level of customer satisfaction and many more. In contrast, the most significant impact causes by bad demand forecast are waste of money and time. Specifically for fabless semiconductor company, the shorter duration between placing of sales order by customer and the order requested date compared to the manufacturing cycle time needed for the processes are the most significant challenge faced by the industry. The typical sales order is booked at eight to twelve weeks ahead of the order requested date. Whereas manufacturing cycle time for an end-to-end semiconductor process takes anywhere from 20 to 30 week. Therefore, by developing a discrete-event simulation model, some of the crucial decision variables such as the production quantities of products at different stages, the release quantity of bare wafer to the wafer fab, the amount of inventory of product and bare wafer at the end of a period could be examined in term of planning and control. The simulation model that is constructed in Python will have input parameters such as customers’ demand for product and GDPW (good die per wafer). The model is programmed with stock alarm to alert the user when the quantity of products reaches certain critical level, thus this model could help the company to control the process. On the other hand, the company could utilise the model to simulate the manufacturing activities when comes to planning. Through running the simulation, the company could get to know the duration needed to fulfil the customer’s demand and the level of inventory at each step to avoid high inventory holding costs due to overstocking.