| 摘要: | 商用伺服器產品向來以高穩定度、高效能以及易於管理為產品主要訴求，因此伺服器產品的開發必須投入大量的研發設計成本，其中產品功能測試與各項設計的驗證成本占了相當大的一部份。本論文個案中，代工廠的「系統整合測試」部門是整個伺服器產品研發各階段中最末端的專責測試單位之一，其主要的工作任務是執行伺服器各功能的整合測試與壓力測試，產品在完成各項研發測試之後才能正式導入產線進行量產。 然而，研發測試專責單位面臨測試設備與資源的成本問題，及研發測試需求失序與過量的問題，導致研發測試成本中心的執行績效難以被衡量與評估。因此，本研究參考過往組織內部移轉訂價交易之相關文獻，規劃研發測試專責單位導入「內部計價系統」做為相關問題的解決方案。其中將導入過程規劃成三個階段進行，先以虛擬計價的方式讓研發測試成本的觀念被建立與重視之後，再對研發測試服務的提供，執行真實計價交易。藉由導入「內部計價系統」，可以減少測試資源的濫用，同時更精準的預估研發階段中所需的測試人力與物力，使伺服器研發專案的測試成本能有效的控制與掌握。 當一個新的系統進行導入規劃時，會有許多不同面向的因素必須事先加以考量。因此，本研究亦從幾個常用的可行性分析面向進行探討，來檢視研發測試專責單位導入「內部計價系統」的規劃是否可行，以及系統導入過程中可能面臨的其他問題。 最後，因導入「內部計價系統」之計價交易所可能衍生的「交易成本」問題，可以透過相關資訊系統的建置輔助，降低因額外的作業流程所產生的交易成本，而計價資訊的透明與公開，亦有助於減低其他因「資訊不對稱」所產生的交易問題。Commercial server products are known for high stability, high performance, and easy management. These are the main features that attract customers. Therefore, during sever product development, considerable amount of budget are allocated to research and development (R&D). Much of these R&D budget are allocated to functional testing and design validation test for products. In this study, we focus on a server product ODM/JDP (Joint Development Program) site. The “System Integration Test” (SIT) Department is responsible for all testing task in the entire server product line, during the R&D stages. The main task for SIT is to perform system integration test and stress test for all functionality of sever, in addition to design validation testing. These functional validation test must be completed before a sever product formally enters the mass production stage. However, the SIT Department in Company A is faced with problems in high cost in testing resources, including manpower, equipment, etc. SIT is a “cost center” within R&D. This results in a phenomenon that testing requirements from related R&D teams are often unchecked and excessive, causing backlogs. As a result, the performance of SIT is difficult to measure and evaluate. Therefore, this study proposes a chargeback system, based on the transfer pricing literature. The proposed chargeback system is divided into three phases. In the first phase, invoice for testing costs are provided to related R&D teams, without actual chargeback. In the second phase, the invoices are charged against negotiated budgets. After the concept of R&D testing costs has been established and recognized, the final phase is to perform the real chargeback with the real budgets of each related R&D teams. The goal of the chargeback scheme is to reduce the abuse for testing resources, and more accurate estimates of the desired manpower and material resources in the R&D phases. We hope that by implementing the scheme, the server testing costs of R&D projects can be effectively controlled and tracked. A feasibility analysis was also conducted, and the proposed chargeback system is found feasible. Finally, “transaction cost” is another major issue in implementing the chargeback system. We further propose that an appropriate information system be developed to institutionalize the chargeback scheme. We also anticipate that transparent and open information of chargeback pricing will help to reduce the “information asymmetry” issues arising from implementing the chargeback system. |
