晶片組在電腦產業產品中是不可或缺的關鍵零組件，在應用上它提供用來作為中央處理器及記憶體等作週邊設備的協調頻率或工作的控制器。無鉛晶片組的單價高且隨時間與市場供需的狀況而變化，因此在少量多樣的產業特性下，買方因採購量過低且為賣方市場的限制下，在採購特性並無議價的能力與空間，促使衍生的存貨持有成本與跌價損失風險相對提高，因此工業電腦產業在存貨政策略上陷入高存貨持有成本或高缺貨風險兩難的困境。 無鉛晶片組為高單價的零組件材料，由於無鉛晶片組本身的價格高且隨時間呈隨機的波動，加上交易上無數量折扣與前置時間不定的特性，促使衍生的存貨持有成本相對高昂。存貨持有成本的壓力主要的是來自高單價零組件的庫存積壓，一般製造業者為避免缺料風險，均會準備安全庫存以降低缺貨風險，當市場供給量缺乏或需要量不明確時，無鉛晶片組等高單價零件的安全庫存量便成為資金積壓的主要來源。 本研究的研究方式是基於訂購點與訂購量(R，Q)模型構建，最佳再訂購點的求取是由再訂購點(Reorder Point：ROP)模型延伸而來，利用6週移動平均預估下期平均使用量與安全庫存計算ROP來控制庫存量。在訂購量上則是依照特定的購買量限制選擇大於ROP的最小允許購買量。為配合移動平均法的計算，在每次庫存量低於ROP進行補貨程序時，此（R，Q）模型需重新計算，以擷取最貼近趨勢的新資料。 本次研究之(R，Q)模型構建完成後，藉由庫存持有成本的績效指標評估現況、ROP、等模型計算結果並進行比較，以驗證本(R，Q)模型的有效性與其效益。 A公司缺乏對關係零件之供應商與顧客端的議價能力，本研究構建目的期望能在不降低服務水準的前提下能有效降低無鉛晶片組庫存，減少庫存持有成本以增加獲利能力。 Chipset is one of key components of product of computer industry. It offers cooperative frequency and operation channel between CPU、memory and I/O port. The lead-free chipset is pricing and it changes with the condition of supply and demand. There is no availability to negotiate the price base on the feature of chipset in procurement caused by low supply & demand quantity and seller market. It also raises the potential risk and loss of shortage seriously. Thus, how to solve the problem caused by chipset with efficient stock strategy becomes the key topic of industrial computer maker. Lead-free chipset is a pricing component. It changes by random and no discount for procurement quantity. Unstable lead-time for component complicates the problem. These features lead to high inventory hold cost for lead-free chipset mainly. For solving this problem, most companies will prepare safety stock for potential risk and loss caused by unpredictable demand and supply. Usually, safety stock will be the easiest way to solve the unpredictable demand from field. Lead-free chipset becomes the main source of inventory cost because of unpredictable demand and supply. Then, pricing lead-free chipset becomes the pressure of operation capital. This research sets up a model (R, Q) model to solve the problem of high inventory hold cost caused by chipset feature. The R model refines the parameter from typical ROP model (Reorder Point Model). 4 weeks based moving average method is applied to predict the average demand quantity and safety inventory to carry out the number of ROP for inventory control. The Q model decides to the minimum order quantity when the inventory quantity is lower than ROP base on the result of moving average method. The (R, Q) model has to be renewed when each refill up the inventory to get latest information. Performance of indicator is carried out and compared indicators performed with other models like current practice, ROP models model to verify the effectiveness and stability of model. Due to “Company A” is middle scale of manufacturer of industrial computing. Bargain power is on the vendor and customer’s hand. The purpose of this research is looking forward to declining the inventory host cost caused by lead-free chipset in stock and profit rate enhancement under unchangeable service level.
