半導體產業工程鏈服務整合者之探索性研究

Abstract

半導體產業製程節點微縮及晶片整合技術的演進使產業內原本垂直分工的各類型廠商必須更密切地合作解決諸多設計挑戰，即產業正面臨著一股許多業界人士所形容「整合（Consolidation）」力量的出現：產業內服務接受者IC 設計公司需要花更多資源，與其它廠商合作才能成功開發一顆晶片，但同時還得面對越來越強大Time-to-market 的競爭壓力；產業內服務提供者（包括EDA 工具商、矽智財廠商、設計服務廠商、晶圓專製廠及封測廠等）則莫不進行各類型產業內的整合活動（如，發起產業聯盟、與其它廠商合作開發等）幫助設計公司解決問題以取得市場領先地位，但也花費比過去更多額外的資源處理與其它廠商之間的合作關係。即業內廠商皆面臨需花費比過去大量之資源於與其它廠商協同合作之整合活動上。為解決這樣的問題並提升半導體產業鏈運作效率，本研究嘗試以交易成本理論及新近發展出比傳統供應鏈管理概念更適用於半導體產業之工程鏈管理概念，透過質性研究之個案研究策略歸納產業現象，推導並定義出「服務整合者」（Service Integrator）概念為半導體產業內廠商進行解套，使之可以達到縮短設計週期、快速量產等工程鏈之管理目標。 「服務整合者」為半導體產業內位於設計公司及服務提供廠商間一新廠商角色，它具備IC 設計支援能力、可取得製程及封測資訊之能力及建構完善IT 之能力，可以整合所有服務提供廠商之資訊（Data）或資源，提供給客戶即IC 設計公司，以協助其管理工程鏈。在此情境下，設計公司不用再自己管理整個半導體產業工程鏈，而是可以透過服務整合者去整合鏈上其它廠商，省下與其它廠商交涉之交易成本，得以更專注在市場產品的開發等利基事業；服務提供廠商則可以 透過整合者與其它廠商進行合作，而不用再像目前一樣進行各式多對多的整合活動，亦省下目前因為進行整合活動而花費之交易成本。因此，就交易成本的觀點來看，服務整合者的出現即是幫整個半導體產業廠商節省下諸多交易成本，而使工程鏈可以運作地更有效率。 此外，本研究分析現有五大類型服務提供廠商之能力後得出現有「設計服務供應商」、「晶圓廠」、或未來「晶圓廠內部設計服務團隊分割獨立之新公司」或「此獨立公司與已與晶圓廠為一對一合作關係之設計服務供應商合組新公司」為最可能扮演「服務整合者」角色之廠商類型；且發展初期會以特定晶圓廠為核心於產業內形成彼此競爭之工程鏈整合集團；待發展成熟並成長後，則有機會跨不同晶圓廠提供整合服務。本研究依此邏輯推導出共二十二項準命題。

Recent advances in process technology and design complexity have driven semiconductor industry structure from vertical disintegration to consolidation. Design houses are forced to spend more resources than ever before on cooperating with other firms to tape out new chips under the growing pressure of time-to-market. Meanwhile, EDA tool providers, SIP providers, design service provider, foundries, and assembly & testing houses require spending extra resources on more industry-wide consolidation activities to help design houses solve increasing design problems. To enhance the operation efficiency of the vertically disintegrated semiconductor industry, the study proposes a concept of “Service Integrator” based on the analysis of transaction costs and engineering chain management. The proposed Service Integrator is defined as one new business model, which coordinates data or resources from service providers to assist design house to design products and manage the engineering chain which involves the activities of design support, access to manufacturing and testing data, as well as constructing a reliable IT system. Emergence of Service Integrator decreases both design houses’ and service providers’transaction costs spent on consolidation activities and accomplish the goal of engineering chain management such as first tape-out success, design cycle time reduction, and rapid mass production etc. In addition, the study suggests that Service Integrator can be best run by 4 types of firms - design service provider, foundry, new type of firms evolving from design service teams inside foundry, and new type of firms merged together by design service teams inside foundry and design service provider. In the early phase of development of Service Integrator, each service integrator will lead one foundry alliance to compete with another foundry alliance. In this study, 22 propositions and several business implications are concluded.