DRAM 產業十年興衰 – 探索日本企業管理困境與全球競合趨勢

譚仲民(Jonq-Ming Tang)

指導教授 ： 張重昭

國立臺灣大學/管理學院/高階公共管理組/碩士(2005年)

https://doi.org/10.6342/NTU.2005.01979

摘要

DRAM產業十年興衰 – 探索日本企業管理困境與全球競合趨勢
論文摘要

一九九五年前後，日本NEC、東芝、日立、富士通、三菱等機電集團，雄霸全球DRAM（動態隨機存取記憶體）市場，這些日系產業巨擘，擁有技術、品牌、資金、經驗、市占率和人才等經略市場的成功要件，也確實在一九九O年前半葉火紅的DRAM市場上呼風喚雨。那麼，為何在十年之間，整個日本DRAM產業土崩瓦解，僅存爾必達（Elpida）獨撐大局，全球市佔率更萎縮到10%以下？這是筆者投身半導體業之後經常自問的疑惑。

本論文是以一九九五年到二OO四年為時間主軸，除了簡單介紹記憶體產品、DRAM的種類和應用演變歷程之外，也透過產業數據資料的蒐集、整理，來說明這十年之間全球DRAM業界的興衰起伏實況，藉以表現這個資本密集高科技產業的變動不羈，來點出日本和其他地區從業廠商所面臨的挑戰。

至於探索日本DRAM產業由盛轉衰的原因，筆者是利用任職於力晶半導體的『職務之便』，輾轉向日本、台灣的半導體業先進和公司同事，拋出本文一開始就提出的疑問，通常受訪者都會先楞一下然後沈思幾許，然後說出他們的觀點；另外，筆者也因為工作關係，必須參與許多跨國會議和公司內部的討論，或是聽取睿智之士的想法，這些親身實地的經驗，也在消化之後融入本篇論文的內容，成為研究日本企業文化、管理乃至於日系DRAM公司潰敗的論據。

歷經過去十年的競合，今天的全球DRAM市場已經具備了寡佔的雛形，分佈在韓國、美國、歐洲、日本和大中華區的五個主要DRAM陣營，也各自採取不同的競爭策略。筆者根據目前的產業實況、技術前景，再加上生活用品數位化的趨勢（digital convergence）和快閃記憶體（flash memory）市場的快速成長，在本論文中嘗試歸納、預測未來全球DRAM、記憶體產業的主流趨勢，並針對幾大聯盟廠商的特質和可能的發展方向，提出剖析和個人的看法。

本論文雖然是以產業數據資料、訪談與實務經驗為主體，提出了對未來DRAM產業的預測，以及探討日本式企業管理的困境，但整個寫作的過程也讓筆者產生了『何以至此？良有以也！』的感慨。結論中筆者提到這種社會互助與企業獲利之間的權衡，但這恐怕是更龐雜、困難的議題，也非本文主旨了。

關鍵字

動態存取記憶體

並列摘要

The DRAM Industry’s 10 year rise and decline – an exploration of Japanese Business Management’s difficulties in coping with the trend of worldwide competition Thesis - Abstract
Around 1995, Japanese semiconductor firms including NEC, Toshiba, Hitachi, Fujitsu, and Mitsubishi were at the pinnacle of their success, as they made up the major players in the worldwide DRAM (Dynamic Random Access Memory) industry. These Japanese firms possessed the technology, brand name recognition, capital, experience, market share, and human resources (all necessary conditions) for success in this industry. The first half of the Nineties came to be known as a prosperous time for the Japanese as well as the worldwide DRAM industry. Which leads us to today, and the question why in the last ten years has the total Japanese DRAM industry almost entirely disappeared, with only Elpida surviving, struggling to maintain a meager 10% market share? The thesis that ensues will delve further into the raison d'être.
This thesis will focus on the period between 1995 through 2004. Following our introduction to the DRAM types, applications, development and evolution, and progress, we will then focus on the prospective of Japanese DRAM, as well competition between other DRAM manufactures-through the gathering and organization of statistical market data-in order to illustrate how the worldwide DRAM industry’s rise and fall in the last ten years in what is usually a stable highly capital intensive industry however in actuality it is very volatile industry.
As for the reasoning for the Japanese DRAM industries rise and fall, we will use Powerchip Semiconductor, as well as real world experiences and relationships in both Japan and Taiwan Semiconductor industries, in order to illustrate the central themes in this thesis. Furthermore, due to the nature of the Author’s occupation, it is necessary for participation in various multinational meetings as well as internal discussions or other input from various DRAM insiders. Thus we have organized and compiled these into this investigation of the Japanese enterprise culture and management, along with how this can explain the Japanese DRAM industry’s downfall.
Succeeding ten years of competition, the worldwide DRAM industry today is only a mere fraction of itself with only a few producers left—these being scattered around the world including Korea, United States of America, Europe, Japan, and the greater China region making up the five major DRAM camps, each with their own competitive stratagem. In using the compilation of current industry trends, technology prospective, digital convergence development, as well as our knowledge concerning the rapid growth of the flash memory market, we will try to forecast the future of the worldwide DRAM industry and distinguish probable development trends, while additionally make further suppositions based on our experience and opinions.
Even though this thesis tends to focus on not only statistical data and interviews, but also relative experience, as the focus of its research, the perception involving the future of the DRAM industry as well as the difficulties with Japanese management are identifiable but many of these questions may not be simple in their solution. Furthermore many of these difficulties arrive in the vary nature of corporate Japan’s culture, in its long standing devotion to social responsibility. Therefore these difficulties that occur in Japan’s very unique corporate culture have burdened many Japanese DRAM firms in their strife for profits. However this topic quite complex in nature is not the question that is explored in this thesis, but room for further exploration in further studies.

並列關鍵字

DRAM

參考文獻

4. Nicholas D. Kristof(紀思道)、Sheryl WuDunn（伍潔芳）、林添貴譯, 民92, Thunder from the East(東方驚雷),台北:時報文化

參考文獻

1. UBS Investment Research, 10 May 2004, DRAM Global Update, 10 reasons for caution on DRAM

2. Merril Lynch Global team, 16 August 2004, DRAM/NAND update: No replay of 2001 downturn; Supply constrains seen

3. JP Morgan Asia Pacific research, 1 March 2005, Global DRAM+NAND market, An interesting upturn: very strong 2004 followed by moderate but still healthy 2005E and 2006E