This research regards ASE(Advanced Semiconductor Engineer) company as the subject of studying, combine with two subjects of resources construct and competitive advantage, to discuss that the ASE technology receives process and the situation of resource constructing. The main research purpose, is to probe into constructing of the resource and ability organized, how move and transfer knowledge that outside make to and organize the inside effectively, and via the study mechanism of, make useful knowledge organize the memory to store and spread after absorbing , melting and putting into inside, become the foundation of organization key ability. This research adopt ¡§case study method¡¨, choose ASE by the way of depth interview, to combine the relevant paper and develop the proposition to report the resource constructing process comprehensively deeply. The main conclusion has: The ASE, to accumulate and to study experience, and strong study willing, carry on with the technological provider FCT in the dynamic course of technological transfer, study to latent know-how and skill of management in a lot of different nature nearer and nearer in a proper order, and play an aggressive role in the course of technological transfer, learn tactic knowledge and cooperate with the characteristic of technology to launch the corresponding study activity during the process of transferring, and after accumulating certain knowledge resource foundation, have build up the competitive advantage by the knowledge learning. In addition, ASE set the cross-function mechanism that communicates and role crossing over department boundary person while organizing, in order to improve the horizontal communication. This suggestion on this industry development is: 1.Key ability is obtained, cultivation, subordinate to a ministry can not be reached shortly. So, short-term interests may not represent and succeed for a long time . 2.Key value of ability come from value of customer, because one urviv grasp customer need new value e in market early than competitor only.

As the the development stage of IC design industry changes, the IC design companies in Taiwan increasingly become more and more. Taiwan design industry ranks the second place in the world, following behind the American IC design industry. The technology becomes maturer and the company scale becomes larger. The demand of the integration is a very important trend for the IC design industry. The resources Taiwan companies have are not enough to sustain the competitive advantage in the global competition. From the viewpoint of markets and products, the systematic companies own the power to decide the product standards. The product development is inclined to the development of Digital Home Integretion. This study is based on the RBV theory to observe the way to compete during the different IC design companies, the industrial analysis by the Five Forces model, and the degree of commanding the demands of customers. Finally, we hope to find out some KSFs (Key Success Factors) of Taiwan IC design companies and IC design industry. This study finds that the scope of product line becomes more and more important. How to decide and develop the new product line efficiently and successfully is a KSF in the future of the IC industry. In the past, Taiwan IC design companies owned the competitive advatege by low-cost stratege, but it is not the KSF in the modern time. The company will focus on the development of integration, brand, and the demand of the product market. We interviewed the executives of the different IC companies in Taiwan, and regarded those companies as our target companies of case study to analyze their KSFs. We discussed their strategies and KSFs by the model of this study. We also find the KSFs may change as the time moves. Key Words¡GIC industry, resource-based theory, KSFs

Abstract Networks are a general feature of Taiwan¡¦s industry. Alongside rapid developments in technology, the island¡¦s industrial structure is being transformed; at the same time, new approaches to administration and management are developing. Hi tech industries are playing a vital role in Taiwan¡¦s economic growth, and this thesis aims to uncover the network characteristics of these industries. The thesis uses the extensive literature available today as the source material for analyzing and comparing the industrial and network characteristics of both traditional and hi tech industries. Three important underlying factors are presented in this thesis; together, these factors help to understand the differences between network characteristics in traditional and hi tech industries. Our research into the characteristics of networks in hi tech industries revealed some interesting results. Firstly, we found that these networks are more elastic, and better able to react to changes in the business environment than traditional networks. Secondly, we discovered that job status in hi tech industries is generally less obvious, more equal and more fluid. Thirdly, existing personal connections still play an important role in hi tech industries, but are less important than in traditional businesses. Networks in hi tech industries have inherited many characteristics of the traditional network; however, as the industrial environment has changed, so networks within hi tech industries have developed many interesting characteristics which differ from traditional industrial networks in important ways.

After decades of development, Taiwan¡¦s IC industry has become the world¡¦s third largest manufacturing center, and claimed the positions of the world¡¦s first largest IC foundry industry, second largest packaging industry, as well as second largest IC design industry. For an industry to be successful, in addition to the provision of various niches and resources from private businesses, it requires the coordination of industrial policy-making, peripheral conditions, and infrastructures. Therefore, to enhance current competitive advantages and facilitate industrial transformation for greater competitiveness, the government needs to make effective industrial policy to promote competitive environment. For this reason, this study intends to investigate the current competitive environment of IC industry, how the government responds in terms of stipulating rules and regulations, and what the primary concerns of policy adoption are. It is clear that aside from an effort by businesses, the assistance from government is pivotal to industrial development as well. This study is an attempt to first, collect literature related to industrial analysis and policy; secondly, to analyze the competitive environment of IC industry and thereby find out the relevant policies that influence the managerial effectiveness of Taiwan¡¦s IC industry; and lastly, to survey IC companies and experts on IC industry by utilizing a questionnaire designed based on Fuzzy Analytic Hierarchy Process. After prioritizing the concrete measures for a variety of dimensions of IC industrial policies based on the survey results, the researcher proposes top ten most emphasized including: the assistance in introduction of techniques, the training of technological talented, the aid in the development of SOC technique, the subsidy in R&D of innovative techniques, the accelerated depreciation on R&D equipments, the implementation of three-links, the tariff exemption for importing R&D equipments, the enhancement of National Defense Substitute Servicemen (NDSS), the training of marketing talented, and the enactment of patent rules. Accordingly, this study concluded by providing suggestions for government, IC industry, and future research.

Technological backwardness is the common problem for all developing nations. The acquisition and creation of advanced technologies are the common dreams of these catch-up economies. How to build up the adequate technological innovation system for high-tech industries is the major challenge to all catching-up economies. Because of the success of technological development, this research will investigate the technology innovation and diffusion system by taking the IC industry in Chinese Taipei as a case study. The purpose of this research is the provide a guide for the future direction of technological innovation and diffusion systems and of share experience on the development of technology in the IC industry of Chinese Taipei with other catch-up economies. This paper first presents hottest study topics: Knowledge Management and National Innovation System (NIS), next introduced two analytical frameworks --- the technological trajectory framework and the multiple cooperation framework --- which are used as tools to describe the evolution of technological development in catch up economies. It will then propose a technological innovation and diffusion system for catch up economies. The structure and components of the system will be discussed and a questionnaire survey designed for the IC industry in Chinese Taipei. The results of the survey will be discussed and future development of the technological innovation and diffusion system of the IC industry in Chinese Taipei will be proposed by this research.

處於競爭激烈、產品生命週期短暫和快速變動的時代裡，製程創新活動成為廠商競爭優勢的關鍵因素之一。Pisano（1995）研究指出廠商藉由製程創新活動來強化競爭優勢，包括：加速產品上市時間、加快爬坡速度、提高使用者接受性、強化專利產品地位等，其他學者也認為廠商應該重視製程創新活動，以因應未來的環境趨勢。國內製造業向來以製造能力見長於國際舞台，近年來半導體產業和資訊電子產業的表現更顯示廠商優異的製造實力。儘管如此，有關這方面的研究仍然非常少，因此本研究的目的是探討我國廠商製程創新活動之重要關鍵因素，期望能對這方面進行較為整體性的分析。 根據技術創新相關理論，本研究以技術網路、人員、組織和制度四構面探討半導體產業和主機板產業製程創新活動之重要因素和差異性，所得到的研究發現如下： 1. 製程創新活動可區分為四類型：獨立型、依賴型、主導型和協調型。獨立型廠商傾向於獨自開發新製程技術，依賴型廠商則傾向於透過技術網路提升製程技術水準；在對內和對外的製程創新活動上主導型廠商的整合能力較強，協調型廠商則較弱。 2. 不同製程創新類型有不同的製程知識來源。在來源方面，半導體產業（流程產業）顯示出較集中的現象，主機板產業（非流程產業）則顯得較分散。 3. 與設備供應商的互動關係為製程創新活動之重要影響因素，不同類型的廠商與設備供應商的互動關係不同。 4. 主機板廠商（非流程產業）藉由製程知識輸出來提高外包廠商的製程品質，製程創新程度越高的廠商與外包廠商的關係越密切。 5. 人員的技術和經驗為製程創新活動中重要的因素之一，半導體廠商（流程產業）特別強調研發人員的量產經驗。 6. 製程創新程度越高，廠商的製程技術單位之主導權越大。 7. 製程創新程度越高，廠商越傾向於設立製程技術協調單位。 目錄 / Abstract In highly competitive, short product life cycle and quickly changing environment, process innovation is one of a firm’s key competitive advantages. Pisano & Wheelwright (1995) consider that firms which do their best in process innovation can enhance their competitive advantages, including shortening time to market, increasing ramp-up speed, enhancing users’ acceptance, increasing product’s patent status and so on. Hayes & Wheelwright (1984) point out the concept of manufacturing strategy, consider that firms should escalate the role of the manufacturing division to decision-making on top management level, and claim that competitive advantages which are based on manufacturing capabilities will direct the decision-making process of marketing and engineering in future. To catch up adaptability for the trend of the future environment, other scholars also suggest that firms should actively make efforts to execute process innovation activities. Manufacturing firms in Taiwan display their manufacturing competence on international status, especially in IC industry and in information technology industry. Therefore, Taiwan is always named as “manufacture kingdom”. In the past, industrial firms’ process innovation is mainly thinking about employing cheaper labors and buying newer equipment. However, now they pay more attention to research and develop process technology, to enhance production process integration capabilities, to increase production flexibility and so on. For example, TSMC and UMC in Taiwan have more advanced process technology, which is about at the same technological level as IC industrial leading firms oversea. Their advanced process technology not only creates many benefits but also drives the development of other related industries. Their success shows us that process innovation will bring firms continuously high growing rate and take long-run competitive advantages. By forgoing observation, this paper wishes to discover what makes Taiwan the manufacturing kingdom and how do firms increase process advantages and execute process innovation activities. So the purpose of this paper is to find out what influences a firm’s performance of process innovation and how foregoing factors affect new process innovation and existing process innovation. Utterback (1994) finds assembled product and non-assembled product have different characteristics in innovation activities. Skinner (1992) explores how stakeholders are within their cognizance about a firm’s process innovation, and finds that there are different focuses between process industry and non-process industry. Therefore, this paper will go further to study how foregoing factors influence process performance in process industry and in non-process industry. According to some related technology innovation theories and innovation theories, this paper is conducted based on four key dimensions: technology network, people, structure and management. Then, this paper infers some important items of forgoing dimensions from past studies to build the research framework of the paper. These items in technology network dimension are equipment suppliers, material and component suppliers, satellite factories, group/strategic leaguer/consumer and research institutions. In people dimension, this paper mainly wants to explore how personnel’s skills and experience influence process performance. Structure dimension contains three items, which are organizational structure, task allocation and strategy. This paper lists important items in management dimension, including rewarding system, training & education, information system and coordination mechanism. Because related studies in the past explore only a part of process innovation, this paper employs case study to examine process innovation more systematically and completely. To explore a firm’s process innovation of high-tech industry in Taiwan, this paper chooses four IC manufacturing firms (process industry) and four motherboard firms (non- process industry) as research samples, which have better performance than others in their industry. Moreover, this paper classifies these samples into four groups according to the industrial characteristics and the process innovation level. In the field of IC industry (process industry), Fab-T and Fab-U are famous for advanced process technology and continuous research on advanced process technology. So this paper explores the development of the new process technology in Fab-T and Fab-U, and the improvement of existing process technology in Fab-M and Fab-K. In the field of motherboard industry, MB-A is the only motherboard firm in Taiwan, which has process patents in Taiwan now. And MB-G actively executes process innovation recently, including package and product vibrating tests. So this paper properly classifies MB-A and MB-G as a group with higher process innovation level, and explores their advanced process innovation activities. Then MB-T and MB-E are classified as lower process innovation on lower level group, and this paper explores how they improve existing process innovation. By a series of analysis about primary and secondary data of eight firms, the conclusions of this paper are: 1. There are four types in process innovation: Independent Firms, Dependent Firms, Directing Firms and Coordinating Firms. In IC industry, the Independent Firms almost develops new process innovation alone; the Dependent Firms tend to properly aided by members of technology network to enhance process innovation level. In motherboard industry, the Directing Firms have stronger relationship with menders of process innovation activities, which actively respond to the process technology unit, the Coordinating Firms have weaker relationship. 2. Different types in process innovation have different sources of process-related knowledge. The sources of IC Firms are more convergent, however the sources of motherboard firms are more divergent. 3. One of the key factors in process innovation is equipment suppliers, with which different types have different interaction. 4. Motherboard firms actively enhance the production quality of satellite factories through process knowledge output. Motherboard firms with higher process innovative level have more closely cooperating relation with their satellite factories. 5. The people is one of the key factors in process innovation. IC firms put more emphasis on the mass-production experience of engineers. 6. Whether it is in IC industry (process industry) or in motherboard industry (non-process industry), firms with higher process innovation level tend to give the process technology unit more power to integrate activities among different departments. 7. Whether it is in IC industry (process industry) or in motherboard industry (non-process industry), firms with higher process innovation level tend to set up the technology coordination unit. Finally, this paper has some advises from forgoing conclusions, and wishes to efficiently enhance a firm’s performance and production yields. These advises are: 1. The Depending Firms should do their best to improve relations with members of technology network. 2. To efficiently use resources and to increase performance in short terms, the first thing of the Coordinating Firms should do is to enhance process capabilities, instead of building good cooperative relation with members of technologic network. 3. Firms should set up a special technology-coordinating unit, which can efficiently increase interaction among different departments. 4. Firms should give more power to the process technology unit, so that it can cooperate with other departments more efficiently.

本研究主要是探討台灣半導體產業中晶圓代工與動態隨機記憶體製造廠商競爭優勢有哪些？並且由價值鍵模型、市場-技術生命週期論著、策略矩陣模型、策略性資源模型及鑽石模型來探討。前面是以一般產業模型探討產業具有競爭優勢。 　　從產業現象中推理，歸納哪些具有競爭優勢條件，再經過模型加以應用、推導、驗證，更能使理論與實務相結合。 　　本研究運用Porter鑽石模型探討，找出每一構面因素的相關競爭優勢內涵。第一因素是生產要素，第二因素是需求條件，第三因素是企業策略、企業結構和競爭程度，第四因素是相關及支援性產業，第五因素是機會，第六因素是政府角色。之後依序探討各因素實際上在產業中成功因素內涵加以分析。並配合個案研究與個案深入訪談專家，依序分析國內最具代表性廠商所具有競爭優勢形成的因果關係，使得探討產業競爭優勢更趨完整。 本研究結果使得下列相關命題更具實質的意義。 1. （價值鍵模型）中探討產業競爭優勢、產業分工與群聚的效果。 2. （策略矩陣模型中）分析得知競爭優勢之基礎與條件。 3. （策略性資源模型）得知廠商組織能力培養與產品創新及有形資產之土地、廠房、設備購置時間之重要性。 4. （後進、先進地區市場-技術生命週期論著）：探討產業發展策略中進入成熟期時台灣晶圓代工與DRAM廠商在進入21世紀深次微米技術時，採取何種方式及步驟，保持競爭優勢。 5. （生產因素）資本資源探討，Venture Capital 投入，不僅止於資金挹注且提供經營資訊，銀行融資及推介策略性合夥人。 6. （企業策略、企業結構和競爭程度）國內該產業廠商之策略、管理型態及組織結構如：未來經營策略有合併、策略聯盟與競爭又合作方向，良好事業策略制定與執行影響廠商生存利基與競爭條件。 7. （機會）產業之機會，如國際IDM大廠製造轉移至台灣晶圓Foundry與DRAM廠商。3C商品整合趨勢，提高DRAM需求與掌握國際市場開拓能力，增加市場通路機會。 第一章 緒論……………………………………………………………… 1 1-1 研究背景與問題………………………………………………… 1 1-2 研究的目的……………………………………………………… 1 1-3 研究對象與範圍………………………………………………… 2 第二章 相關文獻探討…………………………………………………… 4 2-1 競爭優勢觀念之探討…………………………………………… 4 2-2 經營策略觀念之探討…………………………………………… 9 2-3 價值鍵探討……………………………………………………… 15 2-4 產業分析觀念之探討…………………………………………… 16 2-5 其他研究架構因素探討………………………………………… 22 2-6 相關理論架構模式之比較……………………………………… 23 第三章 研究設計………………………………………………………… 26 3-1 研究架構………………………………………………………… 26 3-2 研究流程………………………………………………………… 28 3-3 資料蒐集與分析………………………………………………… 29 3-4 研究方法………………………………………………………… 29 第四章 台灣半導體產業之晶圓代工與動態隨機存取記憶體競爭優勢探討…… 31 4-1 前言……………………………………………………………… 31 4-2 全球半導體產業發展現況與展望……………………………… 32 4-3 影響晶圓代工及DRAM需求因 ………………………………… 48 4-4 台灣DRAM製程技術及研發探討 -市場∼技術生命週期論…………………… 49 4-5 具競爭優勢生產成本與價格-價值鍵模型……………………… 54 4-6 專業服務導向與多樣化產品之競爭優勢-策略矩陣模型………… 56 4-7 專業晶圓代工與動態隨機記憶體製造-策略性資源模型………… 60 4-8 彈性經營管理強度與國際競爭能力…………………………… 70 4-9 整體績效改善與達成…………………………………………… 71 4-10 生產力與獲利能力之優勢……………………………………… 72 第五章 晶圓代工與動態隨機記憶體產業競爭力分析………………… 75 5-1 鑽石理論模型構面因素………………………………………… 75 5-2 生產要素………………………………………………………… 77 5-3 需求條件………………………………………………………… 95 5-4 公司實力、策略與競爭…………………………………………100 5-5 相關與支援產業…………………………………………………104 5-6 機會角色…………………………………………………………116 5-7 政府角色…………………………………………………………119 第六章 公司個案討論……………………………………………………126 6-1 個案介紹-聯華電子股份有限公司 ……………………………126 6-2 聯華電子公司策略性資源模型（一）…………………………130 6-3 聯華電子公司策略性資源模型（二）…………………………133 6-4 聯華電子公司「晶圓專工」策略-價值鍵模型………………136 6-5 個案介紹-台灣茂矽電子股份有限公司………………………138 6-6 台灣茂矽電子公司總體策略-產業價值鍵模型………………141 6-7 台灣茂矽電子公司策略性資源模型……………………………142 第七章 台灣半導體產業競爭優勢-Foundry & DRAM研究探討……145 第八章 結論與建議………………………………………………………161 參考文獻……………………………………………………………………164 附錄一訪談問卷…………………………………………………………169

IC(積體電路)產業是台灣過去三十多年來發展最完整、且在世界上最具有競爭力的創新型產業。研發與技術專利成長快速，也代表這個產業的技術創新成果。台灣IC 產業技術創新和專利的研究累積了一些成果，但較缺乏使用研發創新發明人網絡所做的分析。本研究使用IC 產業專利發明人網絡資料，從結構洞(structural holes)和地位訊號(status signal)這兩個過去西方半導體和生物科技等創新產業研究中，最能解釋創新行動者與網絡系統和創新表現理論機制的學術研究累積成果出發，並反省台灣特有IC 產業分工技術利基位置、流動型勞力市場，發展出專利發明人網絡對專利發明成果影響機制之模型。研究發現，分析結果支持發明人地位訊號、結構洞網絡理論機制有助於專利創新發明的假設。而比較特殊的例外是，上游的專利發明網絡，結構洞低的類型反而有比較好的專利創新。這是由於上游的IC設計產業技術特性，需仰賴同一團隊的人重複合作方能完成設計，所以大多與重疊的人合作，結構限制較高，因此呈現與結構洞理論不同的情況。此外，研究發現，流動並不一定帶來較佳的技術創新，台灣IC產業專利人才流動對於技術發展與創新的正面影響，較多是在上游的IC設計產業。 因此創新發明鑲嵌在台灣特有半導體的不同技術利基環境中，還必須考慮到不同產業位置的技術特性，與產業技術發展所在的階段，才能正確解釋台灣IC產業的創新發明機制。 / The IC (integrated circuit) industry has been the most complete, competitive and innovative industry in Taiwan over the past three decades. The rapid growth of patents in this field demonstrates the marvelous outcomes of technological innovations. A huge amount of patent-based research has been accumulated in this industry, but only a little is related to the inventor network. Structural holes and status signal are two competing perspectives in explaining the dynamic mechanisms between innovative actors and network systems of the most innovative industries, semiconductor and biotechnology, in advanced countries. We use these two theories and network data of patent-based inventors to discuss the specific niche of vertical disintegration and fluid labor market in Taiwan IC industry, and develop models on mechanisms of how inventors’ networks affect the outcome of inventors’ patents. The findings reveal that the hypothesis on the effects of status signals and structural holes in inventors’ collaboration networks on the citations of inventors’ patents was supported. However, the patent inventions of the upstream IC design firms, the collaborative invention teams seem to be small or high structural constraints. Most of the patents of these newly growing firms in IC design industry are specialized in consumption products. The structural constraints are high, but the impact of is large. Besides, inventors’ fluidity would not necessarily create more influential patents. The fluidity of patent-based inventors was positive impact to the patent inventions, more was found in the IC designs company. The IC industry has been vertically differentiated into three sectors: the IC designs, the IC foundry, the package and test. In principle, these three sectors demand different levels of technology. The maturity of technological niches will affect the impacts of the patents.