摘要
半導體產業在台灣晶圓代工業者的加入之後，快速的走向水平分工的產業運作模式。透過過去二十年的發展，各個產業鏈也趨向成熟，一方面使得產業的脈動加速，一方面也使得技術的發展遇到瓶頸。在傳統摩爾定律之外，近幾年半導體業者不斷的思考，在傳統的微縮尺吋之外，如何延伸、拓展成熟的製程技術。而有 More Than Moore(超摩爾定律)與Beyond CMOS(後CMOS摩爾定律)的More Moore的技術發展趨勢。超摩爾定律是以現有材料與機器設備(或極少改變) ，開發更多元件，應用在更多的領域。其中包括功率元件、影像感測元件…等。後CMOS摩爾定律是加入更多的製程與材料來延伸CMOS元件的極限。這使得半導體技術發展的構面，由摩爾定律分頭朝向後CMOS摩爾定律與超摩爾定律，來強化未來半導體產業的發展重點。
本研究將以產業雙螺旋理論的水平分工與垂直整合的各項驅動力檢視，對後CMOS摩爾定律與超摩爾定律的半導體產業，分別闡述其重要的驅動力與其未來發展趨勢。其中後CMOS摩爾定律下，半導體產業水平分工之後，在近幾年漸漸出現垂直整合的趨勢。其主要的驅動力是各項產業鏈上的主要供應商，已經取得相當大的競爭優勢、實力漸漸豐厚，為了更進一步追求技術與產業發展的需求，以技術與資本密集度最高的晶圓生產業者為中心，形成了虛擬垂直整合的經營模式。綜觀台灣的半導體發展史，從IDM角度切入的DRAM產業，因為是整合性的發展，在人力資金需求龐大的高難度與自主技術建立不易的情況之下，面臨崩解的窘境。反觀以半導體供應鏈中的晶圓製造為切入點的晶圓代工業，發揮了台灣製造業的既有優勢，已在這個半導體製造領域取得領先且領導半導體技術的發展，近年來朝向虛擬垂直整合，儼然已經是IDM的角色。這個發展的歷程驗證了產業雙螺旋理論，當其中一個產業區塊的競爭力不斷增強，而促使產業分解成水平分工。一旦競爭力與技術能力超越供應鏈的各個環節，即跨足產業供應鏈的各個環節 而走向垂直整合，在這方面台灣已經將半導體產業的雙螺旋走過了一輪。而目前以半導體晶圓代工為主，整合半導體產業的垂直整合商業模式，具有其獨特的整合優勢，不宜再向上整合產品行銷與設計。
而超摩爾定律下主要是類比產品，設計較為複雜，設計輔助工具、設計服務與製程元件的使用，比較偏向客製化，每家IDM都有自己的歷史與特性。然而，近幾年透過技術與生產釋放給代工廠，加上合作開發的逐漸增加，將使得這一個產業的水平分工逐漸加速，加上非常成功的數位產品設計公司的競相加入此一領域，都顯現出產業趨向水平分工的驅動力量正在加強。當類比半導體產業的雙螺旋循環循著摩爾定律的演變軌跡，走到水平分工的分水嶺，許多相對應的產業，例如：類比設計解決方案與封裝；正是蓬勃發展的絕佳時機。並且可以預期的是，其雙螺旋演變軌跡的速度是不斷的在增加當中。

Abstract
Semiconductor industry speed up its horizontal disintegrated structure，after Taiwan’s semiconductor foundry business model enter the supply chain. In the mean time , the whole semiconductor industry supply chain become mature after over twenty years evolution and development. It also accelerated the clock speed of whole semiconductor industry and forced the industry have to face the challenge of technology barrier. It is why more Moore and extended CMOS technology development strategy have been proposed in the semiconductor development roadmap during recent years. There are two main focuses : Beyond CMOS and More Than Moore. More than Moore is trying to implement existing materials and equipments with additional devices for more application area like power devices, sensors…. Beyond CMOS is to implement more materials or processes to enhance the device performance with existing process technology or speed up the process development to match Moore’s Law requirement.
In this research, we will demonstrate double helix theory of semiconductor industry’s vertical Integrate structure and horizontal disintegrated structure driving forces for Beyond CMOS and More Than Moore industry development focuses, to highlight the critical parameters we have found and try to predict the trend of the future. For beyond CMOS , the vertical integrated structure is becoming the major trend of the industry. Foundry company become stronger and dominate the capital investment and technology owner to integrate from EDA, SIP, technology, mask, manufacture and assembly/testing as virtual fab. for IDM or Fabless company. In the history of semiconductor industry in Taiwan, there is a failure starting point of DRAM semiconductor industry, it is started with a IDM model. It needs heavy human an capital resources investment and also the leading competitive technology, results in difficult business decision to continue the investment or not. On the other hand， semiconductor wafer foundry business model become a successful model. It is a typical double helix evolution of semiconductor industry, foundry company break into the supply chain base on manufacture strength of Taiwan industry and become a leader of semiconductor manufacturing and also the semiconductor technology field. Once the leading area becomes a leading edge of whole industry, it turn out to a driving force for this leader to broaden his area to vertical integration. It is a unique business model as virtual IDM that grapes business opportunities from different products and application to leverage the up and down turn of the market and among competitors.
For more than Moore trend, major products are analog process, design, EDA, SIP and process technology are application and user specific , customized in this field. But through IDM release and join development technology to/with foundry is accelerating the horizontal disintegrate of the structure. In the mean time leading fabless companies are also speed up their acquisitions with analog design house to join this battle field, it also indicate that more than Moore semiconductor industry field will have a quick horizontal disintegrate moving forward. For more than Moore’s semiconductor industry, it have 10 year’s technology lag to follow Moor’s Law evolution track, lots of EDA, IP and back end package, testing … technology are booming up to be worthy for focusing on the business development and for sure the clock speed is faster than Moore’s Law before.

參考文獻目錄
1.“終端設備解決方案與應用” , Texas Instrument , Taiwan, 2013.
2.司徒達賢, “資源基礎理論與企業競爭優勢關係之探討”, 行政院國科會專題研究計劃成果報告,1997.
3.陳玉娟” ARM軍團猛搶灘 行動晶片陷混戰 未來2年市場版圖恐重新大洗牌”Digit Times, Apr.,2,2013.
4.連于慧 “聯電與星科金朋合作 3D IC技術走開放生態模式” , Digit Times, Jan.30,2013
5.彭國柱, “IDM 委外代工新商機＿掌握T.S.M.C. 四個關鍵要點”, 工研院 IEK, Aug.,20, 2010
6.鄧榮惠” 設計工具就緒,晶片設計開始跨入3DIC領域”, EE Times Taiwan , May.17,2012
7.“Fab Lite/Asset-Lite Business Model Gaining Momentum” ,IC insights, Sep.29,2011.
8.“Global Wafer Capacity 2013”, IC Insight , Dec., 2012.
9.“IC Companies Close 49 Wafer Fabs from 2009-2011”, IC Insight, Feb.,15,2012
10.“Semiconductor Manufacturing”, Texas Instrument , Feb., 2013.
11.Bennett, Herbert, “International Technology Roadmap for Semiconductors, 2012 Update Overview”, ITRS, Jan.,2013
12.Clarke P., “tsmc plans 3-D IC assembly launch early 2013”, EE Times , Feb.,2,2012.
13.Constance, S.C, Gower, J. R. “A Value Chain Perspective on the Economic Drivers of Competition in the Wireless Telecommunications Industry” Master Thesis , Sloan School, MIT, 2001.
14.Ellis, J. “450mm fab for tsmc in New York”, 450mm , Nov.,25,2012
15.Feldhan, J., “ A March 2012 Snapshot Look at What the Gurus are Saying About the Analog and Analog ASIC Chip Market”, Signal Process Group Inc. Mar., 2012.
16.Fine, C.H, ”Clockspeed : Winning Industry Control in the Age of Temporary Advantage”, Sloan School, MIT, 1998.
17.Fitzpatrick, J., “An interview with Steve Furber on ARM microprocessor ”, Communications of ACM, P34, 54(5), 2011
18.Garrou, P., “Wafer Level Chip Scale Packaging (WL-CSP) an Overview”, IEEE Transactions on Adv. Packaging ,P198,23(2), 2000.
19.Hajdarbegovic, N., “3D IC market to see stable growth thru 2016 – Flash Memory and DRAM Lead the Way” , TechEye., Mar.,19, 2013.
20.Hergert, J.,“2012 analog IC market Update” , Electronics.ca Network, Feb.,14,2013.
21.Hill, C.W., Jones, G.R., “Strategy Management Theory”, 6th Ed, Boston, Houghton Mifflin, 2004.
22.Ho, Lora, “Tsmc Recognizes Outstanding Suppliers at Supply Chain Management Forum, tsmc, new release, Dec.,2,2011.
23.Holtberg, Paul, “International Energy Out Look 2011” ,Energy Information Association (EIA) , Sep.,2011
24.Hwang, S.H., “CMOS image sensor : Current Status and Future Perspectives”, Samsung White Paper, Jun.,18,2012.
25.Laker, J.F., “ Annual Report 2012”, Analog Device Inc., Feb., 2013.
26.LaPedus, M, “Analysts : Samsung Wins Apple’s Foundry Business” Semiconductor Manufacturing & Design Community, Oct.,4,2011
27.LaPedus, M.,”Fabless-Foundry Model Under Stress”, Semiconductor Manufacturing & Design Community, Jun.,26,2012.
28.LaPedus, M., ”Foundry ARMs Race Under Way”, Semiconductor Manufacturing & Design Community, Jun.,Feb.,21,2013.
29.LaPedus, M., “Maxim , Powerchip form 300mm foundry alliance “, EE Times Asia, Aug.,03,2012.
30.McCallum, John C.,”Memory prices 1957~2013”, JCMIT.com , Feb.,23,2013
31.McClean, Bill, “Pure-play Foundries and Fabless suppliers are Star Performances in Top 25 2012 Semiconductor Supplier Ranking”, IC insights , Mar.27,2013.
32.McKernan, Ed,” “How Apple Plans to Leverage Intel’s Foundry”, Semiwiki Dec.,9,2012.
33.McLellan, Paul, “A brief history of Semiconductors”, Semiwiki.com, Oct.,25,2012.
34.Mika, N., “ASML announced customer co-investment program aimed at accelerating innovation” ASML, Press Releases, Jul.,9,2012.
35.Moore, G.,”Crammping more component onto integrated circuits”, Electronics Magzine,P4, Nov.,11,2006.
36.Porter, M.E. “Competitive Strategy”, Free Press, N.Y. , 1980.
37.Randewich, N. “Intel takes big step in chip foundry business” , Reuters, Feb.25,2013
38.Rosso, Dan, “Global Semiconductor sales remain flat in August”, SIA sales report, Oct.,2,2012.
39.Shih, K.C., “ GUC’s Flexible ASIC Services” Globalunichip-Company Profile.
40.Thompson, S.E., “Uniaxial-Process-Induced Strained-Si : Extending the CMOS Roadmap”, IEEE Trans. On ED, P1010, 53(5), 2006.
41.Thurston, T.,”Intel NBI:Intel corporation’s new Business initiatives”, Harvard Business Review , Oct.,06,2008.
42.Templeton, R.K., “2012 Annual Report “, Texas Instrument, Feb., 2013.
43.Upadhyaya, P, “Here’s why Intel need CISCO’s foundry business”, Silicon Valley Business Journal, Jan.,17,2013
44.Wilsonville,Ore.” Mentor graphics announce comprehensive design enablement platform (Samsung’s 14nm IC Manufacturing Process)” Samsung.com, Press Center, Dec.21,2012
45.Wu, SL , Fine, Charles H. ,“Industry Dynamics within Semiconductor Value Chain IDM, Foundry and Fabless,” Master Degree Thesis, Sloan School, MIT , 2003.
46.Yogasingam, A., “Tear Down : Inside the Samsung Galaxy S3 Smart Phone , EE Times, Jun.,14,2012.