IC Design House 之庫存管理改善 ~ 試行限制理論於 M公司之研究

Abstract

台灣IC 設計產業近年來成長相當快速，已成為世界上除美國矽谷之外的第二大無自有晶圓廠 (Fabless) IC設計公司之群聚地。幾家各領域之領導廠商：網路晶片 (如瑞昱)、光儲存晶片 (如聯發科) 和消費性晶片 (如凌陽) 等，在某些特定產品上甚至擁有全世界擁有超過五成的市場佔有率 (如聯發科的光儲存晶片、DVD播放機晶片等)。然而也因為無自有工廠，在所有製程皆委外的情況下，僅能依據市場預測提早2~3個月投單。景氣好時委外產能難求，景氣愈好時委外生產交期愈不可靠，委外廠也要求放大外包量做為保障產能之條件。一但景氣反轉，放大的安全庫存量及在製品量將產生數倍於平時的庫存等等特性。劇烈變動的消費性市場特性與愈來愈短的產品生命週期，不適當的庫存管理必須將造成跌價的損失，反而侵蝕獲利水準。本研究以限制理論為基礎，討論對IC設計公司生產規劃與庫存管理的方法的改變的可能 (包含以Demand-pull取代過去以 “Forecast” 生產；以 “動態庫存管理” 來管控庫存水準等)。透過理論的探討與實際資料的驗證，希望能找出一些有助於IC 設計公司因應市場景氣波動之對策，避免產生大量的庫存損失，進而提昇企業之彈性應變能力，提高獲利。

As a result of clustering effect and rapidly growing, next to the Silicon Valley in the States, Taiwan has become the 2nd large base for the fables IC design houses in the world. Some of the leading companies has even dominated more than 50% world-wide market share in some specific application area. Such as: Net work chipset (RealTek); Optical Storage chipset (MediaTek) and Consumer market chips (Sunplus), etc. However, as a “fabless” IC design house, most of the production jobs has to be outsourced to the subcontractors, and the production plan is done according to the forecast for the next 2~3 months. To be one of them, we can see the subcontractor’s capacity and stability present a reverse trend to the total market demand. And, the subcontractors might ask for more capacity reservation as a condition to secure the limited capacity during blooming market. So, once the market trend turn down, an IC design house has to face the fact of much higher inventory which can significantly affect to profit margin, especially in the drastically changed digital consumer market and short lift time products. Based on the Theory of Constraint (TOC), this paper is trying to study if there are other alternatives to get better result in an IC design house’s production plan and inventory management field. Which includes: the study of Demand-pull way vs. Forecast way and “real time inventory level monitoring method” in inventory management, etc. We are trying to evaluate those theories with a company’s history data. Then, we might able to say if these methods are able to reduce a company’s inventory loss, enhance its flexibility to market demand and increase its profit as well.