Japanese Semiconductor Industry Conference

April 12-13, 1990 Takanawa Prince Hotel Tokyo, Japan DataQuest mm a company of MMMB ThcDun&BradstreetCorporation 1290 Ridder Park Drive San Jose, California 95131-2398 (408) 437-8000 Telex: 171973 Eax: (408) 437-0292

Sales/Service Offices:

UNITED KINGDOM FRANCE EASTERN U.S. Dataquest Europe Limited Dataquest Europe SA Dataquest Boston Roussel House, Ibur Gallieni 2 1740 Massachusetts Ave. Broadwater Park 36, avenue Gallieni Boxborough, MA 01719-2209 Denham, Uxbridge, Middx UB9 5HP 93175 Bagnolet Cedex (508) 264-4373 England France Telex: 171973 0895-835050 (1)48 97 31 00 Fax: (508) 635-0183 Telex: 266195 Tfelex: 233 263 Fax: 0895 835260-1-2 Fax: (1)48 97 34 00

GERMANY JAPAN KOREA Dataquest Europe GmbH Dataquest Japan, Ltd. Dataquest Korea Rosenkavalierplatz 17 Shinkawa Sanko Building Daeheung Bldg. 505 D-8000 Munich 81 1-3-17 Shinkawa Chuo-ku 648-23 Yeoksam-dong West Germany Tokyo 104 Japan Kangnam-gu, Seoul 135 Korea (089)91 10 64 011-81-3-5566-0411 011-82-2-552-2332 Telex: 5218070 Fax: 011-81-3-5566-0425 Fax: 011-82-2-552-2661 Fax: (089)91 21 89

The content of this report represents our interpretation and analysis of information generally avail­ able to the public or released by responsible individuals in the subject companies, but is not guaranteed as to accuracy or completeness. It does not contain material provided to us in confidence by our cUents.

This information is not fiimished in connection with a sale or offer to sell securities, or in connec­ tion with the solicitation of an offer to buy securities. This firm and its parent and/or their officers, stockholders, or members of their families may, from time to time, have a long or short position in the securities mentioned and may sell or buy such securities.

Printed in the United States of America. AJl rights reserved. No part of this publication may be reproduced, stored in retrieval systems, or transmitted, in any form or by any means—mechanical, electronic, photocopying, duplicating, microfilming, videotape, or otherwise—without the prior written permission of the publisher.

© 1990 Dataquest Incorporated # Dataoyest nn a company of lilf Tlie Dun SiBiadstreet Corporation

1990 JAPANESE SEMICONDUCTOR INDUSTRY CONFERENCE ULSI Era: Challenges and Opportunities April 12-13, 1990 Takanawa Prince Hotel Tokyo, Japan

THURSDAY, April 12 8:00 p.m. Registration 9:00 p.m. Welcome Prince Room Manny Fernandez President Dataquest Incorporated 9:30 p.m. Worldwide Semiconductor Industry Trends ,,,;,;;. = . .Prince Room Hal Feeney Group Vice President and Director, Components Group Dataquest Incorporated 10:00 a.m. Japanese Semiconductor Industry Trends Prince Room Kazunori Hayashi Associate Director, Japan Components Group Dataquest Japan Limited 10:30 a.m. Coffee Break Crown Room 11:00 a.m. The Electronics Industry in the '90s and the Role of ULSIs Prince Room Dr. Tadahiro Sekimoto President NEC Corporation 11:30 a.m. Semiconductor Industry in the '90s and the Role of ULSI in Korea Prince Room Jin Ku Kang Chairman Samsung Electronics Co., Ltd. 12:00 Noon Lunch Crown Room and Zuisho-no-ma 1:15 p.m. Semiconductor Industry in the '90s and the Role of ULSI in Europe ..,:,..,;...,,, , .Prince Room Speaker to Be Announced JESSI 1:45 p.m. The U.S. Semiconductor Industry: A Perspective for the '90s . ; :,,...,;, ,->•-. .Prince Room Dr. Daniel L. Klesken Vice President, Senior Semiconductor Analyst Prudential-Bache Capital Funding 2:15 p.m. Prospects of the Semiconductor Business in the 1990s: A Spotlight on ASIC Business Prince Room Tsuyoshi Kawanishi Executive Vice President Toshiba Corporation 2:45 p.m. MOS Memory Trends in the '90s Prince Room Takeshi Sasaki Board Director and General Manager Semiconductor & Integrated Circuits Division - ' Hitachi. Ltd. 3:15 p.m. Coffee Break :. , • • • • • - -- • ...... ,— ,. Crown Room 3:45 p.m. The Trend of New Concept Devices . .:. i.-,, v,^ ..„>,;. .,.,-,,,. .;,. .;...... , -, Prince Room Dr. Hiroyoshi Komiya General Manager LSI Research and Development Laboratory Mitsubishi Electric Corporation 4:15 p.m. Ramping Up of the LSI Business .... ,„.,.. ...;.; ^ •.... i... - . - :..i.;..... ti;,.;,., .; .i;.; v. . Prince Room Nobutsune Hirai Managing Director Kawasaki Steel Corporation

(Over)

1290 Ridder Park Drive, San Jose, CA 95131-2398 (408) 437-8000 Telex 171973 Fax (408) 437-0292 4:45 p.m. Half-Micron Lithography Prince Room Shoichiro Yoshida Senior Managing Director Nikon Corporation 5:15 p.m. The Role of Thin-Film Technology in ULSI Prince Room Wiliam W.R. Elder President and Chief Executive Officer Genus, Inc. 5:45 p.m. Cocktails Crown Room FRIDAY, April 13 9:00 a.m. Artificial Intelligence and Hyper Media: The Semiconductor Impact on Hyper Media Prince Room Koji Yada President Hyper Media Corporation 9:30 a.m. Superconductor and Josephson Computer: The Roles and Impact of the 1990s Prince Room Dr. Hiroshi Kashiwagi Director-General, Electrotechnical Laboratory Ministry of International Trade and Industry 10:00 a.m. Neural Net Application in the ULSI Era Prince Room Masaka Ogi President Fujitsu Laboratories Ltd. ' 10:30 a.m. Coffee Break , Crown Room 11:00 a.m. HDTV—Its Present Status and Future Prospects Prince Room Kenji Aoki Managing Director, Planning &. Administration Headquarters for Affiliated Corporations Japan Broadcasting Corporation (NHK) 11:30 a.m. Evolution of Telecommunications Networks and LSI Technology Prince Room Dr. Iwao Toda Executive Vice President and Senior Executive Manager NTT Research and Development Headquarters Nippon Telegraph and Telephone Corporation 12:00 Noon Lunch .Crown Room & Zuisho-no-ma 1:15 p.m. The Evolving Role of RISC Technology in ULSI Prince Room Robert C. Miller Chairman, President, and Chief Executive Officer MIPS Computer Systems 1:45 p.m. Automotive Electronics—Today and Tomorrow Prince Room Shoji Jimbo Director, Member of the Board Toyota Motor Corporation 2:15 p.m. Multimedia in the 1990s Prince Room Richard A. Stauffer Marketing Manager Princeton Operation • - Intel Corporation 2:45 p.m. Closing Remarks . i^i ;>.;/.•;•;-w.:.. . vv.;. ....•^. sy,..-. ^; .^wiv^ t ,;:;.;:>.> i i^i ;>.•:>>;>.. .Prince Room Masahiro Miyagawa President Dataquest Japan Limited

3:00 p.m. Adjournment

Dataoyest a company of The Dun & Bradstreet Coq>oratK>n

Japanese Semiconductor Industry Conference

ULSI Era: CHALLENGE AND OPPORTUNITIES

Attendees' List

April 12. 13, 1990

Tokyo Takanawa Prince Hotel

Tokyo,Japan

-1990 Dataquest Japan Ltd. April 12 - Reproduction Prohibited = 104 mSiB+i^lESfJIII-S-l? |ffjl|=$b:;i. Mii (03)5566-0411 77''?v5U (03)5566-0425 ADVANTEST Corporation

Katsuaki Minani

Hajime Sasaki

AMD Corporation

Kenjiro Terada

AMD Japan Ltd.

Glen Balzer

Mitsuo Otsuka

AMP(Japan), Ltd.

Kazunari Hyakutake

ANELVA Corporation

Hideo Mi to

ASCI I Corporati on

Ryoichi Kurata

Kazuhiko Ni shi

Kenichi Takahashi

Advanced Micro Devices, Inc.

Robert G. McConne 11

Air Products and Chemicals, Inc.

Dean Duffy

Analog Devi ces K.K.

Kozo Imai

- 1 - Applied Materials Japan, Inc.

Jiro Kitaura

Yoshito Koriyaaa

Keiji Kawai

Yoichi Nakagawa

Seisaku Takata

Asahi Kasei Microsysteas Co., Ltd.

Seiji Azana

Kyoji Kurata

Masahiro Aoki

Bussan Electronic Technology Incorporated

Saburo Maruchi

C.Itoh Techno-Science Co., Ltd.

Hiroharu Isoaura

Norio Sato

CS First Boston (Japan) Ltd.

Vineet Nagrani

Steven Wheeler

Canon I nc.

Nobuyoshi Tanaka

Takashi Minagawa

Katsuni Monose

- 2 - Capital Research International

Masaaki Abe

Crestron ics Co. , Ltd.

Katsuhiko Ohara

Dai Nippon Printing Co., Ltd.

Kosuke Hirabayashi

Osamu Matsuoka

Daif uku Co. , Ltd.

Shinichi Nakatani

Dempa Publications, Inc.

Kinihiro Nishigaki

Disco Corporation

Hitoshi Mizorogi

Electronic Engineering Times

David Lammers

Electronic World News

John Boyd

Robert kuzbyt

Embassy of Canada (Tokyo)

Carl Kuhnke

Embassy of The United States of America

Sam Kidder

- 3 - Far Eastern Economic Review

Bob Johnstone

Fuji Electric Co., Ltd.

Toshio Hoshino

Fuji Electronic Coaponents, Ltd.

Peter Maly

Fuji Photo Fill Co. , Ltd.

Hirozo Ueda

Yuzo Mizobuchi

Fu j i Xerox Co. , Ltd.

Nobuaki Miyagawa

Fujitsu Device Corporation

Akira Honna

Fujitsu Laboratories Ltd.

Masaaki Ogi

Junichi Tanahashi

Fuj i tsu Lini ted

Masaichi Shinoda

Shinkichi Kuribayashi

Yoshisuke Kondo

Satoshi Maru

Kenichi Hori

- 4 - Genus, I nc.

Williaa W.R.Elder

HOYA Corporation

ShoiChi Harada

Tsuneyuki Yananaka

Harris K. K.

Tsuyoshi Kobayashi

Hitachi, Ltd.

Takeshi Sasaki

Masayuki Takegawa

Satoru Ito

Yoshio Toninaga

Hiroshi Nakagawa

J i nkichi Suto

Osaou Fujiwara

Hiromichi Koshikawa

Akio Hayashi

Takashi Ito

Osanu Kasagi

Hitachi Metals, Ltd.

Osamu Ohtani

- 5 - Hitachi Research Institute

Takayuki Masuda

Hiroto Matsumoto

Manoru Morita

Hyper Media Corporation

Koji Yada

IBM Japan, Ltd.

Hiroyuki Sato

Michael R. Crabtree

Takao Nojima

Kiyotaka Suetsugu

ICA Technologies Ltd.

Shigeaki Wada

Industrial Development Authority of Ireland

Dermot M. Tuohy

Innotech Corporation

Hiroyoshi Usuda

Intel Corporation

Richard A.Stauffer

Intel Japan K. K.

Nobuyuki Denda

Takashi Tomizawa

- 6 Intel Japan K. K.

Yutaka Karasawa

Tetsuro Fuj i i

International Semiconductor Cooperation Center

Masato Nebashi

Tatsuo Tanaka

Re ij i Suzuki

Ishikawajima-Harima Heavy Industries Co., Ltd.

Masaya Tanaka

Yukio Tamura

Makoto Nishimura

Japan Associated Finance Co., Ltd.

Katsuhiko Saito

Japan Broadcasting Corporation(NHK)

Kenji Aoki

Japan Computer Aid Ltd.

Hiroshi Yamamoto

Japan Macnics Corporation

Kiyoshi Nakashima

Japan Systhetic Rubber Co., Ltd.

Tatsuo Ichikawa

Nobuyuki Sonobe

- 7 - KOBOTA Inc.

Muneyuki Yaaaguchi

Kaneiatsu-Gosho Ltd.

Takahiro Shuda

Kawasaki Steel Corporation

Nobutsune Hirai

Tsuyoshi Fukutake

Takayasu Yaiada

Joe Fujiaoto

Masash'i Tonishina

Shinichi Hatano

Yukio Yanauchi

Masanori Nanba

Kidder, Peabody International Corporation

Peter G. Wolff

Kobe Steel, Ltd.

Toshiyuki Tanaka

Kodak Far East Purchasing Conpany Inc.

Hirotsugu Kodaka

Kokusai Electric Co., Ltd.

Masaki Hirata

- 8 - Koaatsu Electronic Metals Co., Ltd. Yoshikazo Hayashi

Kubota CoBpater Inc. Kenji Ikeda

Kyocera Corporation

Hartto Honda

Masaai Terasawa

Hidetoshi Aihara

Yukiaori Kaaeda

LSI Logic K. K. Masayuki Suzuki

Kunio Koaatsu

Lai Research Corporation

John Chang

MIPS Coaputer Systems Robert C. Miller

MIPS Coaputer Systeas Japan, K. K. Yasuhiro Nakagawa

Thoaas Laux

Richard Makino

Shinichiro Kuriaura

Takashi Ideta - 9 - MIPS Computer Systems Japan, K.K.

Masaki Matsumoto

Janes MacHale

MITI

Yoshitaka Mikani

Masahiro Hashimoto

Taizo Nakatomi

Harumitsu Suzuki

Mars Electronics International

Yoshiyasu Narahara

Matsushita Electric Industrial Co., Ltd.

Hideya Esaki

Koichiro Shoda

Ke ij i Tsuchi da

Hirofumi Goto

Hideya Oj ima

Kozo Ariga

Minora Shi mi zu

Matsushita Electronics Corporation

Toshinasa Asaka

Kazunari Aizu

- 10 - Merrill Lynch Securities Company Tokyo Branch

Matt T. Aizawa

Ministry of International Trade and Industry

Dr. Hiroshi Kashiwagi

Minolta Camera Co., Ltd.

Junichiro Takeuchi

Ichiro Yoshiyama

Mita Industrial Co.. Ltd.

Tsunehito Tachibana

Mitsubishi Bank Corporation

Terud Inani

Mitsubishi Corporation

Mitsuo Miyachi

Mitsubishi Electric Corporation

Dr. Hiroyoshi Komiya

Hideo Mori

Kinio Sato

Masahiro Yamane

Taizo Kadoma

Hiroyuki Saruhashi

Koichi Kanahara

Naoki Takayaaa

- 11 - Mitsubishi Electric Corporation Hideo Saeki

Atsuko Uchida

Mitsubishi Kasei Corporation

Fuaio Tokunitsu

Tateshi Yaaada

Mitsui High • Tec Inc. Atsushi Fukui

Mitsutoyo Corporation

Yasuyuki Yaaaguchi

KEC Corporation

Dr. Tadahiro Sekimoto

Tonihiro Matsumura

Hajiffle Sasaki

Tsuyoshi Maeda

Tetsuo Onikura

Terumasa laai

Yasuo Ii da

Kazuhiro Todokoro

Satoru Sato

Susumu Kitazawa

Kenj i Matsu i

- 12 - NKK Corporation

Yoshiyuki Kurita

Syoji Nagaoka

Susufflu Nakao

NMB Semiconductor Co., Ltd.

Shosuke Shinoda

Kazuo Yoneda

NTT Data Communications Systems Corporation

Takashi Katsukawa

Taichi Nakamura

Toru Yamashita

National Semiconductor

Eugene G. Taatjes

National Semiconductor Japan, Ltd.

Steven J. Heppner

Ronald E.Livi ngston

Takuya Ogawa

New Japan Radio Co., Ltd.

Shigeru Yamashita

Saburo Nagae

Masayoshi Kitamura

- 13 - Nihon Degital Equipment Corporation

Itsuo Suetsugu

Nihon Keizai Shimbun.Inc

Koichi Nishioka

Nihon Seniconductor, I nc.

Iwao Yamauchi

Nikkei Business Publications, Ltd.

Yoshio Nishinura

Hisashi Tabei

Nikon Corporation

Shoichiro Yoshida

Shigemasa Hisatsugu

Tsunehisa Yamashita

Nippon Motorola Ltd.

Naoko Kobayashi

Shozo Sugiguchi

Nippon Steel Corporation

Kazuo Takanashi

Toshio Wada

Hiroshi Hanafusa

Nippon Telegraph and Telephone Company

Dr. Iwao Toda

- 14 - Nippon Telegraph and Telephone Company

Takayoshi Nakashina

Kazuyoshi Matsuhiro

DR.Kiaiyoshi Yanasaki

Takahiro Makino

Mi npe i Fuj inani

Nippondenso Co., Ltd.

Masayuki Aoki

Yoshichi kawashioa

Masaii Yaiaoka

Nissan Motor Co.. Ltd.

Takeshi Ohguro

Saburo Tsutsumi

Nissei Sangyo Co., Ltd.

Hiroshi Goto

Tsuyoshi Matsukuma

Nissho Iwai Systec Corporation

Takashi Karube

Nissin Electric Co., Ltd.

Eigo Koguchi

Nitto Denko K.K.

Kazuo Iko

- 15 - Nomura Research Institute, Ltd.

Osanu Hayana

Manoru Kobayakawa

Novelius Systeis

Bob Chanberlain

Oki Electric Industry Co., Ltd.

Yoichi Tanaka

Yoshiyuki Honjo

Toshiki Yokogawa

Kazuhiko Shiaizu

Koichi Nakagawa

Kinihito Arai

Oaron Corporation

Akiyoshi Machida

Osaka Sanso Kogyo Ltd.

Michael Solomon

Osaka Titaniun Co., Ltd.

Shigeo Yamamoto

PFU Limited

Yutaka Miyakoshi

Philips K. K.

P. W. Bacon

Ken Funaki

- 16 - Prudential-Bache Capital Funding

Dr. Daniel L.Klesken

Bicoh Company, Ltd.

Haruo Hakayaaa

Hirosfai Nabetani

Makoto Hashiioto

ToiofuMi Nakatani

Yoshiro Hanawa

Ryosan Co. . Ltd. Tatstto Ui

Masakazu Oiezawa

Ryoyo Electro Corporation Morio Nishina

Yoshihisa Shinada

SGS-THOMSON Microelectronics Igor Dorochevsky

Yasushi Mochizuki

Saasung Electronics Co., Ltd. Jin Ku Kang

K.H. Cho

Y.B. Choi

F.Y. Chung

- 17 - Samsung Electronics Co., Ltd. G. T. Joo

K.H. Kia

Y. H. Lee

Saisung Japan Co. . Ltd.

Myung Bae Choi

J. H. Chung

Joo Ki-Taek

H. I. Kia

Jin-Hyuk Yun

Sanyo Electric Co*, Ltd. Masaaki Yanaauro

Katsuya Itoh

Seiko Epson Corporation

Yoshio Yamazaki

Kinio Takemori

Kenzo Nakanura

Seiko Instruments Inc. Yasunori Ebihara

Takao Yoshida

Semiconductor Equipment & Materials International

Shigeru Nakayama

- 18 - Sharp Corporation

Akihiko Kunikane

Sakan Yaaada

Hajiffle Nakajima

Shin-Etsu Handotai Co., Ltd.

Katsunori Kubo

Zenjiro Yanagisawa

Shinko Electric Industries Co., Ltd.

Motoyoshi Nakazawa

Siemens K.K.

Masatoshi Sato

Singapore Economic Development Board

Peng-wai Wong

Singapore Technologies

Peng-Koon Ang

Solomon Brothers Asia Limited

Koichiro Chiwata

Sony Corporation

Toshiyuki Yamada

Haruo Kozono

Haruyoshi Suzuki

Hatsumi Hamada

Yasushi Takino

- 19 - SuiitOBO 3M Linited

Aitio Harada

SuaitoBO Eaton Nova Corporation

Dean Wagner

SuaitOBo Metal Industries, Ltd.

Fuj io laai

SuaitoBO Metal Mining Co.,Ltd.

Akira Nakaaura

TDK Corporation

Masaaki Ikeda

Tachibana TectronCo., Ltd.

Seietsu Onodera

Taiwan SeBiconductor Manufacturing Co., Ltd.

Klaus C. Wieaer

Tencor Instruaents

Grahaa J. Si ddal1

Texs Instruaents Japan Ltd.

Hideo Yoshizaki

Hiroshi Tsuchiya

The Industrial Bank of Japan, Limited

Makoto Taaaki

Yasuhiro Nishi

The Mainichi News Papers

Yoshiyuki Itsumi

- 20 - The Nikkan Kogyo Shinbun

Shinichi Amano

Hiroya Otokosawa

The Sanwa Bank. Limited

Norimasa Kato

The Wako Research Institute of Econonic, I nc.

Takehiro Sagami

The Yoniurl Shinbun

Ikuya Shigeta

Tokin Corporation

Shizuo Asanabe

Tokyo Electric Power Corporation

SusuBU Yoda

Tokyo Electron Limited

Susumu Ichikawa

Ke i i chi ro Kur i yama

Katsuhiko Yamamoto

Hiroshi Odani

Hiroshi Yokoyama

Kiyotaka Kawamura

Joe Fukuchi

Tokyo Systems Laboratories, Inc.

Hi roshi Konno

- 21 - Tonen Electronics Corporation

Yasuyuki Fukuda

Toaoegawa Paper Co., Ltd.

Masaoni Nishimura

Toshiba Corporation

Tsuyoshi Kawanishi

Keizo Shibata

Yasuoni Uchida

Shoji Ariizuni

Hitoshi Hoshi

Kazuhiko Osada

Kazuei Semba

Minoru Tanaka

Takeo Tanaka

Hitoshi Hara

Makoto Uno

Yoshinori Fujii

Re i ichi Yanagisawa

Naohiro Kimura

Toyota Motor Corporation

Shoj i J imbo

22 - Toyota Motor Corporation Hiroshi Arai

Yoshinori Ohno

Hironobu Ono

U.S. Department of Commerce

Me 1 issa Ski nner

Unisys Japan, Ltd.

George T. Shima

Masanobu Utamaru

United Microelectronics Corp. John Hsuan

Venture Management Associates

Terry P. Hi Isberg

Veriflo Corporation

Paul Craig

Wacker-Chemicals East Asia Ltd.

Dr.Herbert H. Rauh

Xilinx K. K.

Ke izo Ichikawa

YAMAHA Corporation

Takatoshi Okumura

Hidehiko Ki ta

Takashi Murayama

- 23 - Yokogawa Electric Corporation

Mi chi 0 Yoshi oka

Yokogawa U-Systems Co. , Ltd.

Kazuaki Sakurai

- 24 - Dataquest Incorporated

Manny Fernandez

Hal Feeney

Toa Wang

Leonard Hills

Junko Matsubara

Dataquest (UK) Liaited.

Geoffrey M. Chaapion

Bipin Paraar

Daquest Korea

J. H. Son

Dataquest Japan Ltd.

Masahiro Miyagawa

Kazu Hayashi

SusuDU Kuraoa

Kunio Achiwa

Masanori Murata

Hideaki Nemoto

Suffliko Takeyasu

Satoko Kaji

Max Nanseki

- 25 - Dataquest Japan Ltd.

Tsuneo Saito

Takashi Kinura

Kuniki Abe

Satoko Hoshizaki

Eai Maki

Mia Morikawa

Keiko Sakimura

Junko Yairi

Masumi Sakoda

- 26 D B) "< DataQuest nn aa^ompanyo company off The Dun & Bradstreet Corporation

WORLDWIDE SEMICONDUCTOR TRENDS

Harold V. Feeney Group Vice President and Director Components Group Dataquest Incorporated

Hal Feeney is Group Vice Prsident and Director of Dataquest's Components Group. In this capacity, he has direct responsibility for all U.S. semiconductor and component research and coordinates European and Japan-based research. Previously, he was Vice President and Director of Dataquest's Technical Computer Systems Industry Service. Before joining Dataquest, Mr. Feeney was Manager of International Customer Marketing at Intel Corporation, where he was responsible for international marketing/sales support of microcomputer components and development systems. During his 14 years with Intel, Mr. Feeney held various positions in marketing management, product marketing management, and LSI design engineering. He designed the Intel 8008, the first 8-bit microprocessor. Prior to joining Intel, Mr. Feeney was a Component Design Engineer with General Instrument Corporation. Mr. Feeney received a B.S.E.E. degree from the University of Notre Dame and an M.S.E.E. degree from Stanford University. He completed additional graduate studies in Electrical Engineering at the University of Notre Dame.

Dataquest Incorpo):?'ted JAPANESE SEMICONDUCTOR INDUSTRY CONFERENCE April 12-13, 1990 Tokyo, Japan

1290 Bidder Park Drive, San Jose, CA 95131-2398 (408) 437-8000 Telex 171973 Fax (408) 437-0292 |B6«e7001.lMG 00/05AX>:JSt I ULSI Era: Challenges and Opportunities

Worldwide Semiconductor Industry Trends Hal Feeney Vice President and Director Components Group Dataquest Incorporated

1990 Dataquest Incorporated April 12 — Reproduction Prohibited tW42001.lMG 03tt1V00iFEE|

SUMMARY

Expect a slower industr^nn 1989 - Strong first half - Then three negative growth quarters Capital spending slowing

Souice: Ddaquest JSIS Conference April1989

|BW4g002.IMG 03«orao:FEE|

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |B904g003.Md Caa0/90:FEE|

Global industry trends

• Forecast for the 1990s ^ 7

1904201)4lUa oatawoKfel

GLOBAL TRENDS FOR THE 1990s

Regional manufacturing Electronic equipment standards become pervasive NICs become a greater force Wafer fabrication equipment - worldwide sourcing

1990 Dataquest Incorporated April 12 — Reproduction Prohibited 19042005JMQ O3my00-.FEE|

ESTIMATED WORLDWIDE SEMICONDUCTOR INDUSTRY REVENUE Billions of Dollars 100

1 I—I—I—I—I—I—I—r , T—I—I—I—r 1965 1970 1975 1980 1985 1990 Source: Dataquest

|904200e.lMG 03/gOraolFEEl

ESTIMATED WORLDWIDE SEMICONDUCTOR INDUSTRY REVENUE Percent Growth

60

A 40 k A y< J 20 A L- ( 0

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |Bi»42007.IMG flO«WOO:Ftf j

WORLDWIDE SEMICONDUCTOR BOOK-TO-BILL RATIO Three-Month Moving Average Book-to-BJII Ratio

h— 1985 —4— 1986 —4— 1987—4— 1988 —H 1989- Sourcetwsrs

IB904g023.IMO

U.S. SEMICONDUCTOR BOOK-TO-BILL RATIO Three-Month Moving Average Boolc-to-Bill Ratio 1.3

1985 -*\^ 1986 —4" 1987 4"—1988 4"— 1989-

Sou(Ce:WSTS

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |B90420a5.IMG 0QaoraO:FE^|

JAPANESE SEMICONDUCTOR BOOK-TO-BILL RATIO Three-Month Moving Average Book-to-Bill Ratio 1.3

_ _ _ T r iiTiiiiiMiiriMiiiii)iiii[iJ II I I I I I I I 1985 —4*—1986 A^ 1987 4*— 1988 —4*—1989 —«-| Source: wsrra

|B9

TOP 10 MERCHANT SEMICONDUCTOR COMPANIES' WORLDWIDE REVENUE IN 1989

NEC Toshiba Hitachi Motorola Fujitsu Texas Instruments Mitsubishi Intel Matsushita Philips 1 2 3 Biilions of Dollars Souice: Daiaquest

1990 Dataquest Incorporated April 12 — Reproduction Prohibited [BiWHgOOO.IMG 03yaOff»'FEE|

WORLDWIDE SEMICONDUCTOR PRODUCTION IN 1989

•I Top 10 Merchant Companies ^ Captive W All Other

Total = $59.1 Billion ($55.8 Billion without Captive) Source: Dataquest

Itaaamm tnUOkiai

ESTIMATED 1994 WORLDWIDE SEMICONDUCTOR INDUSTRY CONSUMPTION BY GEOGRAPHY

^Japan ^ North America (U Europe Bi Asia/Pacific 003 ROW M Captive

Total = $121.1 Billion ($113.3 Billion without Captive) Souice: Dataquest

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |b9042010,IMG O3/2(»0Q:FEE|

|B90420n.lMG QSISOfX.FEE]

Global industry trends

Markets and applications

• Forecast for the 1990s

1990 Dataquest Incorporated April 12 — Reproduction Prohibited locKzoigiMd oagaoofEFl

TRENDS IN SEMICONDUCTOR MARKETS

• Stronger order rates in all geographies • Firming prices - DRAMs • Production adjustments • Falling prices and consolidation in ASICs • Selective personnel adjustments • Concentration of vendor strength

|B9042014.tMG 03/20/90:FEE |

WORLDWIDE ELECTRONICS PRODUCTION Billions of Dollars 300

North Japan Europe Asia-Pacific/ America ROW

Souice: Dalaquest

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |P90A2015.IMG 03/aPflO:FEE|

WORLDWIDE ELECTRONICS MARKETS 1989 Shipments ^ Data Processing ^ Consumer H Industrial M Communications Em Transportation M Military

Total = $653 Billion Souice; Dataquesi

^g00^a016•I^^G 03/2tV90:FE:E|

WORLDWIDE SEMICONDUCTOR SHIPMENTS BY APPLICATION MARKET 1989 ^ Data Processing ^ Consumer M Communications M Industrial OD Transportation M Military

Total = $55.8 Billion Souice: Dataquest

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |B0042013.rMG 03/20^0:^EE|

ESTIMATED 1989 WORLDWIDE SEMICONDUCTOR INDUSTRY CONSUMPTION BY PRODUCT

MOS Memory MOS Logic Microprocessors Analog Discrete Bipolar Opto Captive

Total = $60.3 Billion ($55.8 Billion without Captive) Source: Dataquest

WiBaiioniMO mhuxfin

ESTIMATED 1994 WORLDWIDE SEMICONDUCTOR INDUSTRY CONSUMPTION BY PRODUCT

MOS Memory MOS Logic Microprocessors Analog Discrete Bipolar Opto Captive

Total = $121.1 Billion ($113.3 Billion without Captive) Source: Dalaquesi

1990 Dataquest Incorporated April 12 — Reproduction Prohibited l&WHgOITjMtd OateWMrffiEl

|B904gOie.lMG 03/20/90:FEE|

Global industry trends

Markets and applications

• Forecast for the 1990s

1990 Dataquest Incorporated April 12 — Reproduction Prohibited p042019.tMG 03Q

FORECAST TRENDS IN THE 1990s

• A slow start as inventories are balanced and DRAM prices stabilize • Promise of Europe 1992 becomes reality • Peak year of forecast growth in 1993 is paced by 4Mb DRAM shipments • Globalization continues as Japanese electronic equipment manufacturers move production to North America and Europe • Asia/Pacific provides the fastest growth opportunity

|B>g'JCiiaD WG OI/06flO:FE6 j

WORLDWIDE SEMICONDUCTOR INDUSTRY REVENUE FORECAST* Billions of Dollars

120 -

100 ::;=ii::i:::: 80 -iilli-liJ 60 -

•iLiiilll-li- •••':•':'::':•••• 40 iijiiijiyiii liiiiyiiii ::ii=:i:::;i: 20 - iiiiilliiijjl

1988 1989 1990 1991 1992 1993 1994 'Excludes captive SoiHce: potaquesi

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |0«g-X^2;iMG OI/U&fJ0FEE|

WORLDWIDE SEMICONDUCTOR INDUSTRY REVENUE GROWTH FORECAST

1988 1989 1990 1991 1992 1993 1994

Source: DataquesI

IBegofioaIMG oiwafiioFEfej

WORLD SEMICONDUCTOR INDUSTRY FORECAST Quarter-to-Quarter Percentage Revenue Growth Percent 15 Actual Forecast

frirmj 10

5 1

Ijil] j 0 iMMiiB—•••***i"C—A^^iin^ fTiiirr

Q1 Q2 Q3 Q4 Q1 Q2 Q3 CM Q1 Q2 Q3 CM ' 1 ' ' r —' ' 1 ' 1988 1989 1990

Source: Dalaqucsl

1990 Dataquest Incorporated April 12 — Reproduction Prohibited IBKoaigHMa inA»«oJgl

ESTIMATED SEMICONDUCTOR INDUSTRY PRODUCTION AND CAPITAL SPENDING Percent Change Year to Year 35 1989 1990 ^ 1991

Production Capital Value* Spending 'includes captive production Souice; Dalaquest

IBOixaaoiMG o3«(vao:gF|

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |904goei.tMG oewwooTggl

SUMMARY

Today, life is flat Capital requirements outpace industry growth Expect DRAM price stability and perhaps a mild shortage Production capacity is under better control than in 1984-1985 DRAM market still paces the industry, but the battle is no longer for market share Industry consolidation began in the 1980s; It will intensify in the 1990s

Igawaoama oa/ai»ao:ftfe|

1990 Dataquest Incorporated April 12 — Reproduction Prohibited Dataquest nn a company of MMMM TheDun&BradstreetCorporation

JAPANESE SEMICONDUCTOR INDUSTRY TRENDS

Kazunori Hayashi Associate Director Japanese Semiconductor Industry Service Japanese Components Group Dataquest Incorporated

Kazunori Hayashi is Associate Director of Dataquest's Japanese Semiconductor Industry Service (JSIS) and Japanese Components Group (JCG) and is based in Tokyo. Among his responsibilities are researching and compiling the Japanese fabrication plant data and the Japanese semiconductor company capital spending forecast and updating the Japanese equipment and materials data base. Prior to joining Dataquest, Mr. Hayashi was founder of Innov Japan and Techno Systems Research Corporation. He has had 12 years experience in the industry. During this time, he has also authored publications and performed research and consulting on the Japanese semiconductor equipment industry. Mr. Hayashi is a graduate of Meiji University with a degree in Commercial Sciences.

Dataquest Incorporated JAPANESE SEMICONDUCTOR INDUSTRY CONFERENCE April 12-13, 1990 Tokyo, Japan

1290 Ridder Park Drive, San Jose, CA 95131-2398 (408) 437-8000 Telex 171973 Fax (408) 437-0292 jBBeaTOQg.lMG o3W7rao!ja1 ULSI Era: Challenges and Opportunities

Japanese Semiconductor industry l^ends Kazunori Hayashi Associate Director Japan Components Group Dataquest Japan Limited

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |eg7iooi.ihid 'oyi3f6o.HAY|

AGENDA

Japanese semiconductor history Semiconductor consumption and forecast Semiconductor production and capital spending Future wafer fab plants Summary

|Ba71002.IMG tKViyjOHAVI

JAPANESE SEMICONDUCTOR CONSUMPTION - 1983-1989 Billions of Dollars 25

20 CAGR = ; :5% 15

10

1

1983 1984 1985 1986 1987 1988 1989 Source: Dalaquest

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |e37t003.IMG 03/iyjO:HAYl

JAPANESE SEMICONDUCTOR GROWTH RATE -- 1984-1989 Percent

1984 1985 1986 1987 1988 1989 Source: Dataquest

IB971004JMG 03ll3P!)0:HAY|

WHAT WE LEARN FROM HISTORY

Previous Worldwide Millions Annual Recessions of Dollars Growth/Decline

1967 1,926 (1.3%) 1971 2,487 (3.5%) 1975 4,496 (13.7%) 1981 14,828 5.0% 1982 15,261 2.9% 1985 24,823 (14.7%)

Souice: Dalaquest

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |Bfl071006.IMG 0af13A»:HAY|

TOP 10 MERCHANT SEMICONDUCTOR COMPANIES' WORLDWIDE REVENUE IN 1989

NEC Toshiba Hitachi Motorola Fujitsu Texas Instruments Mitsubishi Intel Matsushita Philips 2 3 Billions of Dollars

Souice: Dataquest

|Beg7100C.IMG 03/13/!]0:HAY|

TOP 10 JAPANESE SEMICONDUCTOR COMPANIES' WORLDWIDE REVENUE IN 1989

NEC Toshiba Hitachi Fujitsu Mitsubishi Matsushita Sharp Sanyo Oki Semiconductor Sony 2 3 Billions of Dollars

Sburt^e; Dataquest

1990 Dataquest Incorporated April 12 — Reproduction Prohibited jflfl07tOO7.IMG 03/l3«0:HAY|

TOP 10 MERCHANT SEMICONDUCTOR COMPANIES' JAPANESE REVENUE IN 1989

NEC B Toshiba Hitachi Fujitsu Mitsubishi Matsushita Sharp Sanyo Sony OIci Semiconductor 2 3 Billions of Dollars

Source; Dataquest

|Beo7iooB.lWG oa/ia/aoiHAVi

SEMICONDUCTOR SHIPMENTS INTO THE JAPANESE MARKET IN 1989

HI Japanese Companies ^^ North American Companies ijjjjji European Companies Asian Companies

Total = $22.2 Billion Source: Dataquest

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |Be07l6cO.IMG 0a/13TO0.HAYl

JAPANESE SEMICONDUCTOR INDUSTRY REVENUE FORECAST* Billions of Dollars 50 Actual Forecast 40

30 : Hi 20 i : 10 :^ii lij iiiliiiiii _1 1988 1989 1990 1991 1992 1993 1994 *Excludes captive Source: Dataquest

|Bfla710tQ,IMG ttKTO/'JO.HAYI

JAPANESE SEMICONDUCTOR INDUSTRY REVENUE GROWTH FORECAST Percent 150

100

1989 1990 1991 1992 1993 1994

Source: Dataqu^

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |Ba07t01g-IMG Oa/19/gO:HAY|

WORLDWIDE ELECTRONICS PRODUCTION Billions of Dollars 300

North Japan Europe Asia-Pacific/ America ROW

Source: Dataquest

18371011 IMG 0a/03>Q0:HAY I

ESTIMATED JAPANESE SEMICONDUCTOR INDUSTRY CONSUMPTION BY PRODUCT

(Millions of Dollars) Actual Forecast 1989 1990 1991 1992 1993 1994 CAGR

Bipolar 1,529 1,451 1,488 1,554 1,644 1,558 0% MOS Memory 6,233 6,125 6,773 8,270 11,817 12,834 16% Microprocessors 2,588 2,681 3,193 3,927 5,263 5,789 17% MOS Logic 3,677 3,785 4,550 5,688 7,309 7,638 16% Analog 3,626 3,632 4,251 5,096 6,190 7,118 14% Discrete 3,192 3,047 3,404 3,853 4,462 4.962 9% Opto 1,237 1,277 1,455 1,673 1,957 2,212 12%

Total 22,082 21,998 25,114 30,061 38,642 42,111 14%

Souice: Dataquest

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |8971013.IMG 03f14/'JOHAY|

JAPANESE ELECTRONrC EQUIPMENT FORECAST

(Billions of Yen) Actual Forecast 1989 1990 1991 1992 1993 1994 CAGR

Data Processing 6,608 7,080 7,588 8,299 9,069 9,455 7% Communications 2,752 2,934 3,116 3,321 3,570 3,832 70/0 Industrial 2,998 3,205 3,373 3,783 4,018 4,111 7% Consumer 6,742 6,627 6,733 7,157 7,211 7,269 20/0 Transportation 2,123 2,178 2,323 2,578 2,727 2,851 6%

Total 21,223 22,024 23,133 25,138 26,595 27,518 5%

Source:; Dataquest

|83710M.1MG OaW3raO:HAY]

ESTIMATED SEMICONDUCTOR INDUSTRY PRODUCTION AND CAPITAL SPENDING

(Millions of Dollars) Actual Forecast 1989 1990 1991 1992 1993 1994 CAGR

Production 30,074 29,556 32,294 37,596 48,706 52,154 12%

Capital Spending 5,368 5,089 6,635 8,802 11,090 11,234 16%

Source: Dalaquest

1990 Dataquest Incorporated April 12 — Reproduction Prohibited pig71015IMG 0»14/OOHAY|

ESTIMATED SEMICONDUCTOR CAPITAL SPENDING BY REGION

(Millions of Dollars) Actual Forecast 1989 1990 1991 1992 1993 1994 CAGR

North America 3,822 3,759 4,465 5,835 7,409 7,805 15% Japan 5,368 5,089 6,635 8,802 11,090 11.234 16% Europe 1,201 1,273 1,686 2,330 2,943 3,262 22% ROW 1,854 1,877 2,333 2,985 3,433 3,983 17%

Total 12,245 11,998 15,119 19,952 24,875 26,284 17%

Source: Dataquest

|fta971019.IMG (ia/19flW:HAV| JAPANESE WAFER FAB PLANTS BY REGION IN 1989

'Includes company purchases and joint ventures * Source: Dataquest

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |Be07t016.IMG 03^14y00:HAY|

NEW FAB LINES WORLDWIDE R&D and GaAs Not Included Number of Fabs

nil ^^^ ^^^ ^^ "^ ^^ ^^ ^^^ "^^^^ ^^^ "^^^ ^^^ "^^^^ ^^^ "^^^ ^^ ^ ^^^F ^^^^^ ^^^ ^

RH uU 1 1 An 'HJ iiii::;:: — ! :;:::.;

30 -iiii! rr;:;;:;: :::::

20 IV\\: — — : iii 10 Iiii ~ — - — — i;li- n 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991

Source: Dataquest

jeg?1017IMG 03/H/00:HAY|

ESTIMATED JAPANESE SEMICONDUCTOR EQUIPMENT REVENUE

(Millions of Dollars)

1985 1986 1987 1988 1989 CAGR

Lithography 379.1 240.3 325.7 562.1 665.0 15% Automatic Photoresist Processing Equipment 54.3 51.5 61.8 119.7 140.0 27% Etch and Clean 184.3 172.0 211.3 391.0 450.0 25% Deposition 259.8 228.1 245.7 407.0 475.0 16% Diffusion and RTP 95.7 78.7 91.4 127.4 215.0 22% Implantation 140.6 40.3 81.8 212.5 225.0 12%

Total Fab Equipment 1,113.8 810.9 1,017.7 1,819.7 2,170.0 13%

Source: Dataquest

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |e07TOie.lMG 0yt4W0!HAYl

SUMMARY

ULSI era - Advanced products - Progress in high technology using 8-inch wafers - Various applications Competition and cooperation - Alliances - Investments

1990 Dataquest Incorporated April 12 — Reproduction Prohibited DataQuest nn acompanjToa company off The Dun & Bradstieet Corporation

THE ELECTRONICS INDUSTRY IN THE '90s AND THE ROLE OF ULSIs

Dr. Tadahiro Sekimoto President NEC Corporation

Dr. Tadahiro Sekimoto is President of NEC Corporation. Dr. Sekimoto has been associated with NEC since 1948 except for a two-year period that he was on loan to COMSAT. Among his various managerial positions with NEC, Dr. Sekimoto has served as Executive Vice President; Senior Vice President; Board Director; General Manager, Transmission Division; Manager, Communication Research Laboratory, Central Research Laboratories; and Chief, Basic Research Department, Communication Research Laboratory. Dr. Sekimoto graduated from the Physics Department, Faculty of Science, Tokyo University. He also received his Doctor of Engineering from Tokyo University.

Dataquest Incorporated JAPANESE SEMICONDUCTOR INDUSTRY CONFERENCE April 12-13, 1990 Tokyo, Japan

1290 Ridder Park Drive, San Jose, CA 95131-2398 (408) 437-8000 Telex 171973 Fax (408) 437-0292 |BBa87003.1MG Oa«F/QO-JSl| ULSI Era: Challenges and Opportunities

The Electronics Industry in the '90s and the Role of ULSIs Dr. Tadahiro Sekimoto President NEC Corporation

1990 Dataquest Incorporated April 12 — Reproduction Prohibited AT OUR GUEST SPEAKERS REQUEST,

PRESENTATION MATERIAL

WAS NOT PROVIIDED Dataquest nn a company of Ifli TheDun&BradstreetCorporaitio n

SEMICONDUCTOR INDUSTRY IN THE '90s AND THE ROLE OF ULSI IN KOREA

Jin Ku Kang Chief Executive Officer and Chairman Samsung Electronics Company, Ltd.

Jin Ku Kang is Chief Executive Officer and Chairman of both Samsung Electronics Company, Ltd. and Samsung Electro-mechanics Company, Ltd. He has direct responsibility for consumer electronics, semiconductor, information systems, and computer business within SEC. Mr. Kang joined Samsung in 1965. He became President of Samsung Electronics Company, Ltd. in 1973. Mr. Kang has served as Chairman of the Korean Electronics Industry Association. Since 1979, he has held a variety of executive-level positions in Korea and at present serves as Vice Chairman the of Korea Chamber of Commerce and Industry, having once been its Chief Director. He has also been a Committee Member of Policy Advisors to the Ministry of Science and Technology, and Chairman of the Korea-Belgium Economy Cooperation Committee. Mr. Kang received a B.S. degree in Electrical Engineering from Seoul National University.

Dataquest Incorporated JAPANESE SEMICONDUCTOR INDUSTRY CONFERENCE April 12-13, 1990 Tokyo, Japan

1290 Ridder Park Drive, San Jose, CA 95131-2398 (408) 437-8000 Telex 171973 Fax (408) 437-0292 IBBBB7W4 IMG OVOSraO JSI \ ULSI Era: Challenges and Opportunities

Semiconductor Industry In the '90s and the Role of ULSI In Korea Jin Ku Kang Chairman Samsung Electronics Co., Ltd.

1990 Dataquest Incorporated April 12 — Reproduction Prohibited AT OUR GUEST Sr>EAKERS REQUEST

I>RESE1SITATI01ST MATERIAL

WAS NOT PROVIDED DataQuest

BS TheDunSiBradstreetCorporation

SEMICONDUCTOR INDUSTRY IN THE '90S IN EUROPE- EXAMPLE OF THE IMPACT OF ULSI ON THE NONVOLATILE MEMORIES EVOLUTION

Igor Dorochevsky Corporate Vice President President, Japan SGS-THOMSON MicroelGctronlcs K.K. Mr. Igor Dorochevsky Is a Corporate Vice President and President of SGS-THOMSON Microelectronics Japan. His main task is to develop the presence of SGS-THOMSON in Japan. Mr. Dorochevsky has worked for SGS-THOMSON for more than 28 years. Before coming to Japan, he was Vice President Marketing in Asia Pacific based in Hong Kong, Mr. Dorochevsky has spent most of his career In West Germany. Before being assigned to Hong Kong he was Managing Director of the components operation of the THOMSON Group in Central Europe, including West Germany, Austria and Switzerland. Mr. Dorochevsky has graduated as an Electronic Engineer at the French Ecole Superieure d'Electriclte in Paris.

Dataquest Incorporated JAPANESE SEMICONDUCTOR AND ELECTRONICS TECHNOLOGY CONFERENCE April 12-13, 1990 Tokyo, Japan

1290 Bidder Park Drive, San Jose, CA 95131-2398 (408) 437-8000 Telex 171973 Fax (408) 437-0292 |toi 84001 .tMG oaao/qo!!Bl ULSI Era: Challenges and Opportunities

Semiconductor Industry in the '90s in Europe Example off the Impact off ULSI on the Nonvolatile Memories Evolution Igor Dorochevsky Corporate Vice President President, Japan SGS-Thomson iVIicroelectronics K.K.

1990 Dataquest Incorporated April 12 — Reproduction Prohibited Semiconductor Industry in the '90s in Eu Example of the impact of ULSI on the NV Memories Evolution

ID490 - Dataquest conference 4/90 SC Industry evolution in Europe

* Europe has still a large proportion of older tech Example: Capacity by Line Geometry EUROPE USA above 3.0 um 54% 20% 1.0 to 3.0 um 46% 80% 100% 100% * However, the future looks much brighter...

ID490 - Source : Dataquest Dec.89 Wafer start capacity by line geome

EUROPE USA JAPAN over 1 um 81% 91% 78% below 1 um 19% 9% 22%

^H • *^ ^" ^" ^^ ^^ ^tm ^^ •« HHl MHi HH 100% 100% 100%

ID490 - Source: Dataquest Dec.89 Evolution of ULSI in Europe in the '9

* Europe is state of the art in submicron technology * Around 20 advanced tabs will be built in Europe in the next 3 years * The environment is very favorable - Cooperation programs (JESSI) - Government policies (EUREKA) - Europe 92 - largest consumer market * European companies show an aggressive process technology roadmap

ID490 PROCESS - PRODUCT STRATEGY OF NEW MICRON-SUBMICRON PROCESSES

--•

Min.features pm Processes Produc

1.0 0.8 0.5 N-V- Memories EPROMs 1Mb- FLASH EE —

1.5 1.2 O.S 0.5 Logic Multiapplication MCU-CC-PL

I

256K - 1Mb 1-2 0.7 0.5 Static RAMs High Complex - i - . Transputer - SO - 1.2/1.0 0-8 0-5 Logic High Performance Cell Library - S

1.2 0-8 0.5 Logic BICMOS ASICs - Teleco

1,

1 Sour ce : SGS-THOMSON -ID490 EPROM Technology in Europe

EPROM are one of the key technology drivers in E

ID490 EPROM TECHNOLOGY TREND

1MB 4MB 16MB

CM0S-E4 CM0S-E6 CMOS-E6 G mini.feature (um) 1.0 0.8 0.5 act. area w/s(um) 1.0/2.0 0.8/1.6 0.6/1.5 gate pitch w/s(um) 1.2/1.6 0.8/1.4 0.5/0.8 mtl. pitch w/s(um) 2.8/1.6 2.0/1.0 1.2/0.8 contact size (um) 1.4x1.6 1.0x1.0 0.6x0.6 gate oxide (A) 280 200 , 160 cell area , {um2) 19 9 4 die size (mm2) 46 90 130 access time Cnsec) 100 100 <100 masks n. 15 16 18 waf.diameter (inch) 6 8 6-8

Source : SGS-THOMSON -ID490 NV IVlemories in advanced system architectures

The market is going towards more customisation

ID490 ULSI - Trend to customisation

CA % of logic market Total logic 100 • ppppppmwwwnwwnw^ DEDICATED LOGIC

80 -

STANDARD LOGIC

1982

ID490- Source: BCG 1989 NV Memories in advanced system architectures

NV MEMORY •*• RAM + LOQIG in a multiapplication process open th# way t powerfull syat0fTi architecture^

Example..

ID490 Example : Multiapplication proces

* 0.8 um CMOS double metal - 13 to 22 ma * Allows combination of logics with NV Memo features implanted capacitors for analog appl * Tunel oxide for EEPROM/FLASH EEPROM * Memory size : up to EPROM 1 MB ROM 1 MB EEPROM 64 KB FLASH 1 MB SRAM 64 to 128 KB

ID490 - Source : SGS-THOMSON Example : Multiapplication process

* Combination possibilities: LOGIC LOGIC RAM RAM ROM FLASH EEPROM EPROM ANALOG EEPROM ANALOG * Key products^ - Sea of gates and standard cells - Microcontrollers and smart card chips - Programmable logic

ID490 - Source : SGS-THOMSON Example: Multiapplication proces Microcontroller with modular architecture

MEMORY BUS

LU H o 2LJ^A& o o Q < u. u en lU cr X cc o > cc O O m tr w o o I- a: z c O o Q. < w« CL o O < cc c o S3 CJ (L a. IT •D o 6 a. 0. Q. CD ui OC z U cc til o o CM 0. ! I " 1

REGISTER BUS/INTERRUPT-DM1 nA BU ES 1 u

ID490 - Source : SGS-THOMSON SC Industry in Europe in the '90s

* A battle is never def nitively won... * A battle is never def nitively lost.... * ULSI will be the chance of Europe to win again battle.

•••

ID490 Dataquest nn a company of lis The Dun &Bradstreet Corporatiol n

THE U.S. SEMICONDUCTOR INDUSTRY: A PERSPECTIVE ON THE NINETIES

Dr. Daniel L. Klesken Vice President Senior Semiconductor Analyst Prudential-Bache Capital Funding

Dr. Daniel Klesken is a Vice President and Senior Semiconductor Analyst for Prudential-Bache Capital Funding, the investment banking, institutional, and capital markets arm of Prudential-Bache Securities Inc. Prior to joining Prudential-Bache, Dr. Klesken was a General Partner and Senior Semiconductor Analyst at Montgomery Securities. Previously, he was a Vice President and General Manager of Dataquest's Semiconductor Group. He was with Dataquest for seven years and was one of three individuals largely responsible for launching Dataquest's Semiconductor Industry Service. Earlier, Dr. Klesken was associated with Texas Instruments for nine years where he was a Member of Technical Staff working on NASA and DOD-related programs. He also supported their corporate staff with market research and long range planning. Dr. Klesken also worked at IBM and Bell Telephone Laboratories. Dr. Klesken received both his B.S. and M.S. degrees in Electrical Engineering from Lehigh University, and his Ph.D. in Electrical Engineering from Carnegie-Mellon University.

Dataquest Incorporated JAPANESE SEMICONDUCTOR INDUSTRY CONFERENCE April 12-13: 1990 Tokyo, Japan

1290 Bidder Park Drive, San Jose, CA 95131-2398 (408) 437-8000 Telex 171973 Fax (408) 437-0292 |flB68700G.IMG OaW7/OQ\gr| ULSI Era: Challenges and Opportunities

Tlie U.S. Semiconductor Industry: A Perspective for the '90s Dr. Daniel L. Klesken Vice President Senior Semiconductor Analyst Prudential-Bache Capital Funding

1990 Dataquest Incorporated April 12 — Reproduction Prohibited (eo7g66i IMG oaoaoolaT)

OVERVIEW

• Review of the eighties • Current status • Issues for the nineties • Perspective

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |Ba9720Qg.lMG Oe«i3«0:KLE)

THE EIGHTIES

1990 Dataquest Incorporated April 12 — Reproduction Prohibited te972CW3-IMG OZ/tOKiOKLE]

WORLDWIDE SEMICONDUCTOR MARKET Billions of Dollars

1980 1985 1990

Source; WSTTS

1990 Dataquest Incorporated April 12 — Reproduction Prohibited IBBOTgOWIMG Og/00/Ql):KLE|

WORLDWIDE SEMICONDUCTOR MARKET Billions of Dollars bO

40

30

20

in 1980 1981 1982 1983 1984 1986 1986 1987 1988 1989 1990

Source: WSTS

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |8J7aOOSlMG 03;i5/OO.KLE|

REVENUES OF U.S.-BASED COMPANIES Billions of Dollars

1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990

Source: Dataquest WSTS

1990 Dataquest Incorporated April 12 — Reproduction Prohibited jBeoraOOCMG 0eJO3/UO:KLE|

REGIONAL SHARES OF SEMICONDUCTOR MARKET (Percentage of Worldwide Market)

1980 1985 1989 United States Ell Europe Japan •• Asia-Pacific

Source; Dataquest

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |a97aoo7iMG oaoafjoKiEl

1985 INDUSTRY RECESSION

• Industry's worst nightmare • Participants lost $6 billion • Five U.S. DRAM producers drop out • Japanese market share jumps

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |80720oe.lMG 02«)3fiW.KLE|

KEY EVENTS IN THE EIGHTIES

• Microprocessor revolution • Growth of ASICs • Insatiable demand for memory • Extraordinary competition • Increasing customer requirements • Shifting consumption patterns

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |a97gOQ9-iMG""02/63/90KLE|

CURRENT STATUS

• 1990: a year of no growth • User inventories low • Realistic capital spending • Greatly improved manufacturing capability • Strong balance sheets

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |Ba9720IO.IMG 0em3raO:Kl,E|

THE NINETIES

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |Ba97g011.>MG 0g/0a;0OKLE|

WORLDWIDE SEMICONDUCTOR MARKET Billions of Dollars 225

1990E 1995E 2000E

Source: Prudential-Bache

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |6072012IMG O3/1SO0:KtE|

EXPECTATIONS FOR THE NINETIES

• DRAM memory -1990: $50 per megabyte - 2000: $3 per megabyte • Microprocessors -1990: 25-50 mips - 2000: 2,000 mips

1990 Dataquest Incorporated April 12 — Reproduction Prohibited (8072013'.IMG i}2t03/d6KL£\

EXPECTATIONS FOR THE NINETIES

LInewidths -1990: 1 micron - 2000: 0.3 micron Chip density -1990: 4 million transistors - 2000: 100 million transistors

1990 Dataquest Incorporated April 12 — Reproduction Prohibited I8Q72014.IMG 03/1

ISSUES FOR THE NINETIES

• Cost of capital • l\4arket shares • Lack of critical mass • Lack of consistent profits • Shifting geographic markets

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |e97201S.IMG OaoafJOKLEj

COST OF CAPITAL

Tlie number one issue Americans pay significantly more Low cost capital a must for DRAMs and SRAMs

1990 Dataquest Incorporated April 12 — Reproduction Prohibited leoTgoiGJMG oacwoolaEl

WAYS TO CLOSE THE GAP

• Advance payments • Long-term contracts • Partnerships • Use of foundries • U.S. government policy • Shifting capital markets

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |BQ7g017IMG 0a03rJ0:KLE|

DECLINING U.S. MARKET SHARES

• The number two problem • Greatest changes occurred in eighties • Market share key to economies of scale

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |ea72oie.iMG oaisflwlaE]

MARKET SHARE SOLUTIONS

• Leading-edge products & technology • World-class manufacturing • World-class customer service • Global scope • Government trade policy

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |a072oi9.iMG oaoaraoKLEl

LACK OF CRITICAL MASS

• Top 10 Japanese suppliers = 90% of Japanese production • 20-25 Japanese suppliers • Top 10 American suppliers = 72% of American production • >100 American suppliers

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |e972oao.iMG oaiaoolaEl

CONSOLIDATION

• '70s and '80s: era of startups • '90s: era of consolidation • Marriage of convenience

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |Be072021.11^ OgA)4;uO:KLE|

LACK OF CONSISTENT PROFITS (Net Profit Margin) Percent

1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989

1990 Dataquest Incorporated April 12 — Reproduction Prohibited |aQ72og2.ifcte feroawdJKLE]

ROAD TO MORE CONSISTENT PROFITS

• Proprietary products • Intellectual property • High R&D investment • Enforce anti-dumping laws

1990 Dataquest Incorporated April 12 — Reproduction Prohibited [BeQ7g023.IMG Oe/04/aQ:KLE|

SHIFTING GEOGRAPHICAL MARKETS Percent of Sales • '*

1980 1985 1990

^H United States Japan t::;:::| EurOpe Asia-Pacific

Source: WSTS

1990 Dataquest Incorporated April 12 — Reproduction Prohibited ia972flg4.tMG 0»1gOO:KLE|

SERVrNG ASIA-PACIFIC REGION

• Provide local sales & FAE support • Establish local manufacturing • Involve top management in Asia-Pacific • Make worldwide customer support a priority

1990 Dataquest Incorporated April 12 — Reproduction Prohibited lB«)720e5LlMG OeftB;Qa:M.E|

U.S. COMPANIES IN 2000

1990 Dataquest Incorp>orated April 12 — Reproduction Prohibited |BQ7gog6.iiL*G oaoarjo.KLEt

PERSPECTrVE

• Cost of capital narrows • Market share declines - Americans - Japanese • Significant consolidation in U.S. • More consistent profits • Market share gains in Asia

© 1990 Dataquest Incorporated April 12 — Reproduction Prohibited Dataquest nn a company of MMMM The Dun &Bradstreet Corporation

PROSPECTS OF THE SEMICONDUCTOR BUSINESS IN THE 1990s: A SPOTLIGHT ON ASIC BUSINESS

Tsuyoshi Kawanishi Executive Vice President Toshiba Corporation

Tsuyoshi Kawanishi is Executive Vice President for Electronic Components and Materials at Toshiba Corporation. He joined Toshiba in 1952 and has held various manufacturing and management positions including General Manager of their MOS LSI plant, Vice President and Group Executive of the Semiconductor Group, and Senior Vice President. Mr. Kawanishi graduated from the Department of Electrical Engineering at Tokyo Institute of Technology.

Dataquest Incorporated JAPANESE SEMICONDUCTOR INDUSTRY CONFERENCE April 12-13, 1990 Tokyo, Japan

1290 Ridder Park Drive, San Jose, CA 95131-2398 (408)437-8000 Telex 171973 Fax (408) 437-0292 |BB6e70O7.>MG mmsio^^ ULSI Era: Challenges and Opportunities

Prospects of the Semiconductor Business in the 1990s: A Spotlight on ASIC Business Tsuyoshi Kawanishi Executive Vice President Toshiba Corporation

1990 Dataquest Incorporated April 12 — Reproduction Prohibited SPEECH CONTENT

1. SEMICONDUCTOR BUSINESS IN THE 1990s. (D A New Era Has Come — Mega-bit, ASIC, Interna ® What is tlie Mega-bit Age ? ® What is the ASIC Age ? ® What is the International Age ? ® 21 century — peering into a crystal ball — 2. ASIC BUSINESS IN THE 1990s. ® Why ASICs ? ® Trends in the ASIC Market ® Trends in ASIC Technology ® Key factors of ASIC Business 3. TOSHIBA'S STRATEGY FOR THE 1990s. ® Strategy for the 1990s ® Keys to the Mega-bit Age ® Keys to the ASIC Age ® Keys to the International Age

— SPEECH CONTENT

1. SEMICONDUCTOR BUSINESS IN THE 1990s. ® A New Era Has Come — Mega-bit, ASIC, Internat (D What is the Mega-bit Age ? (D What is the ASIC Age ? (D What is the International Age ? (D 21 century — peering into a crystal ball — 2. ASIC BUSINESS IN THE 1990s. (D Why ASICs ? ® Trends in the ASIC Market (D Trends in ASIC Technology ® Key factors of ASIC Business 3. TOSHIBA'S STRATEGY FOR THE 1990s. ® Strategy for the 1990s (D Keys to the Mega-bit Age (D Keys to the ASIC Age ® Keys to the International Age r A NEW ERA HAS COME r TRANSITION OF AGE KILO-BIT, MEGA-BIT, AND GIGA-BIT

Age KILO-BIT MEGA-BIT

Year 1980 — 1990 —

DRAM 64K, 256K, 1M 4M, 16M, 64M

Total bits 1.5X10^^ byte 6.0X10^5 byte 1 Total data Per capita 1.2 M byte 50 M byte (Japan) Equivalent units 150 Mp 48,000 Mp 1 of 1M DRAMs +AM + INFORMATION IN DATA RECORDING RECORDING Application of RE Data of Computer, Character and still memory Movi PC, WS, WP, etc. picture of PC, WS, musi WP, TV, etc. Vide KILO-BIT AGE VS MEGA-BIT AGE

Kilo-bit Age Mega-bit

Business Style Share oriented Profit cau

Key Strategy Capacity Expansion Advanced Product

Easy to oversupply In short s Demand & Supply (Requirement Era) (Adjustme

Possible and Faster Difficu New Entry (Top ranker isn't far ahead) (Technical bar

Days International Trade Days of Friction Complementarity /: 3 ATTRACTIVE MARKETS FOR SEMICONDU IN THE MEGA-BIT AGE

1. Portable OA equipment

2. High performance AV equipment

3. Replacement of other recording med COMPUTER EVOLUTION AND REVOLUT Mainframe to Personal

Speed : 10 MIPS Memory Capacity: 128MB^

^mm •^ '\ \ Mainframe Personal WS 1980s 1992-1995 r CUSTOMERS' NEED

1. to aim at originality

2. to respond to changing requirem (short life cycle)

3. to pursue cost reduction c ASIC AGE means Diversification and Charact

ASIC is Tailor-made merchandise supported by advanced wafer fabrication facilities and exce

Customer

Customer

Customer V ASIC FAMILY

CUSTOM

SEMI-CUSTOM FULL-C

GATE ARRAY STANDARD CELL GLOBALIZATION OF SEMICONDUCTOR

1. Technology without borders • frequent exchange, within established rules. 2. Customers without borders • economic principles work. 3. Logistics without borders • in consideration both of customer requests and political issues. 4. Prices without borders • merchandize and related Information circulate wo in real-time. 5. A "crude oil industry" for all countries' economies • how to survive tough world class competition. 6. No single company can control and cover all • Multinational alliances and collaboration required. V. RIGHT KEYS IN THE 1990s - GLOBAL INTEGRATION OF THE VERY BEST OF TH • Teamwork will be superior to One (Super) man performance.

5^11^^

V. 21 CENTURY

Peering into a crystal ball

Component Giga-bit memory Biological m

Multi media Equipment Humanoid Palm-top computer

Transportation Linear motor car Flying carp

System Perfect weather forecasting Environmen r SPEECH CONTENT

1. SEMICONDUCTOR BUSINESS IN THE 1990s. ® A New Era Has Come — Mega-bit, ASIC, Interna (D What is the Mega-bit Age ? (3) What is the ASIC Age ? ® What is the International Age ? (D 21 century — peering into a crystal ball — 2. ASIC BUSINESS IN THE 1990s. ® Why ASICs ? (2) Trends in the ASIC Market (D Trends in ASIC Technology ® Key factors of ASIC Business 3, TOSHIBA'S STRATEGY FOR THE 1990s. ® Strategy for the 1990s (D Keys to the Mega-bit Age ® Keys to the ASIC Age ® Keys to the International Age

— WHY ASICs ?

ASICs

1. fit the era of differentiation and individuality.

2. expedite system integration.

3. enjoy the fruits of micro-lithography.

4. are key components of all kind of electronic

5. are a super hero of the information era. Auto Vending Machine

Facsim Auto Teller Machine Copy Machine Controller o H Q.

C/) z o o QC o o z> z o o HI 0) Q HI

(/)us ffl 1 1 r 6 6 6 6 o o o;i o o o o o CM O CO CO •<* CM CM O CO (D ^ V CM CM T- T- T- LOGIC DEVICES MARKET BREAKDOW {%) 100

80- OTHER DEVICES

60-

40-

20- ASIC LOGIC DEVICES

V 1988 1990 1992 1994 1996 199 r TRENDS IN PROCESS TECHNOLOGY 5 0^(Design Rule: n)

2.0- ASIC (Gate Array)

1.0- MEMORY (DRAM) 0.5-

0.2-

0.1 1980 1985 1990 1995 TRENDS IN INTEGRATION (Gates) IOM1

1M- GATE ARRAY (Usable Gates) 100K-

10K MEMORY (DRAM)

IK-

V. 1980 1985 1990 1995 r (Delay Time/Gate: Sec) 10m

in-

lOOp-

1980 1985 1990 1995 r TRENDS IN ASICIZATION (COMWJXITY)

SYSTEM

CPU, STANDARD LSI FUNCTION

MEMORY FUNCTION r ASICs FOR

A •••Advanced Process S^^- Short Turn-Around Time I •••Interface between Customer & Vendor C^^^CAD

A*-' Alteration S"- Speciality I •••Individuality C •••Change & Chance

A^*^ Active & Able S"^ Service & Sensitivity I •••Innovate & Improve C •••Characteristic & Conspicuous SPEECH CONTENT

1. SEMICONDUCTOR BUSINESS IN THE 1990s- d) A New Era Has Come — Mega-bit, ASIC, Internat (D What is the Mega-bit Age ? (3) What is the ASIC Age ? ® What is the International Age ? (D 21 century -• peering into a crystal ball — 2, ASIC BUSINESS IN THE 1990s. ® Why ASICs ? (D Trends in the ASIC Market (3) Trends in ASIC Technology (D Key factors of ASIC Business 3. TOSHIBA'S STRATEGY FOR THE 1990s, ® Strategy for the 1990s (2) Keys to the Mega-bit Age (3) Keys to the ASIC Age ® Keys to the International Age V TOSHIBA'S STRATEGY FOR THE 1990

1. 3 BASIC CONCEPTS OF SEMICONDUCTOR BU (D Foresight (D Consistency (D Balance

2. 3 BASIC STRATEGY OF SEMICONDUCTOR BUS 0 Simultaneous 3-generation R&D (D International shareing based on CC & C po ® Management fitted to ASIC Age

— • KEYS TO THE MEGA-BIT AGE

1 • Key processes for the Mega-bit Age ® Micro-lithography • Stepper ( 4MDRAM:gLine .16MDRAM: i Line (2) Three dimensional structure • Trench/Stacked capacitor • Multilayer metalization (3) Mixed device technology • Bi-CMOS • Memory-Logic (ASM) • Discrete-Logic (Intelligent Power De V KEYS TO THE MEGA-BIT AGE

2, Key Materials for the Mega-bit Age (D Low Cost - High Grade Epitaxial Wafer • Price < 1.5 X Mirror Polished Wafer (2) Super-Flat 8" Mirror Polished Wafer • LTV<0,5/^m (D Low Cost Multilead Package (4) Low-Stress Molding Compounds (D High-Resolution/Sensitivity Photo Resist (D High-Reliability Metalization Materials 3. Key Equipment for the Mega-bit Age ® Super High-Resolution Photo Stepper (2) High Throughput Reactive Ion Etching Machine for Tren (D Low-Damage CVD Machine for Stacked Capacitor (D High Throughput Epitaxial Reactor Vertical Diffusion/CVD Furnace KEYS TO THE ASIC AGE

Implementaion of

1. the most advanced process technology

2. highly advanced CAD

3. a powerful central engineering center

4. the best in customer service in convenient des

5. ASICs oriented plant

6. alliances with partners worldwide r KEYS TO THE INTERNATIONAL AGE

1- G-H.S- Philosophy G Global View H Excellent Hardware S Software & Service 2- Philosophy of Globalization Positioning Internationally

Sharing Internationally • Sharing of products & technologies • Restructuring based upon each country's stre • Many kinds of alliances DataQuest nn a company of MMMM The Dun &BradstFeet Corporaitiol n

MOS MEMORY TRENDS IN THE '90s

Takeshi Sasaki Board Director and General Manager Semiconductor and Integrated Circuits Division Hitachi, Ltd.

Takeshi Sasaki is Board Director and General Manager of the Semiconductor and Integrated Circuits Division at Hitachi, Ltd. He has been with Hitachi for over thirty years and has held several managerial positions including. General Manager, Musashi Works; and General Manager and Deputy General Manager, Takasakl Works. Mr. Sasaki graduated from Hiroshima University.

Dataquest Incorporated JAPANESE SEMICONDUCTOR INDUSTRY CONFERENCE April 12-13, 1990 Tokyo, Japan

1290 Bidder Park Drive, San Jose. CA 95131-2398 (408) 437-8000 Telex 171973 Fax (408) 437-0292 |Beag7ooe.»iiG m/mtoo-JSii ULSI Era: Challenges and Opportunities

MOS Memory Trends in the '90s Takeshi Sasaki Board Director and General Manager Semiconductor & Integrated Circuits Division Hitachi, Ltd.

1990 Dataquest Incorporated April 12 — Reproduction Prohibited f MOS MEMORY TRENDS IN THE 90'S

APRIL 12, 1990

TAKESHI SASAKI HITACHI LTD.

V 0 HIT r THE AGE OF ULSI

w- W MARKET Patt B$ .. 16K \^M. IIHHL f-. 1M 300 - ^^u :;:•:.:4M •'4^^^; :16M MOS MEMORY IN THE 90'S CONTINUES -|:U:64i TO BE • MAJOR BUSINESS OPPORTUNITY SUB- MICRON L ULSI 200 & • PROCESS DRIVER T

100

0 1970 1980 1990 2

SOURCE : DATAQUEST, HITACHI 0 HIT /" BIT DEMAND AND LIFE CYCLE OF DRAM

Million PCS

BIT DEMAND

X1.6/Year 1,000 UNIT SHIPMENT

500 X1.8/Year

0 1980 1990

SOURCE : HITACHI V^ ^^ HITA MOS MEMORY WILL LEAD ULSI AGE

1987 : 1Mbit DRAM 1.2p.m

1990: 4Mbit DRAM O-Sjim

1993 : 16Mbit DRAM O.Spim

1996: 64Mbit DRAM O.Biinn

1999 : 256Mbit DRAM 0.2nm

1990 IS THE YEAR OF SUB-MICRON MEMORY

ANOTHER 1/10 AREA SCALE DOWN BY THE END OF

CENTURY

ULTRA HIGH-DENSITY, HIGH PERFORMANCE

MICRO/ASICS WILL FOLLOW.

0 HIT r ISSUES

(TECHNOLOGY)

FINE PATTERN & LARGE » R&D INVESTMENT SCALE INTEGRATION \ MANUFACTURING INVEST. HIGH RELIABILITY ' % HIGH YIELD HIGH PERFORMANCE CLEAN TECH. PACKAGING FAULT TOLERANT TEC % HIGH THROUGHPUT PROCESS/EQUIPMENT • DIE SIZE/WAFER SIZE

MARKEZ T WORLDWIDE OPERATION OPERXA ( TRADE ISSUES K SILICON CYCLE

^ ^HIT MEMORY AS A TECHNOLOGY DRIVER

• EVOLUTION OF DRAM PROCESS TECHNOLOGY

• ULTRA-CLEAN TECHNOLOGY

• REDUNDANCY

• INCREASING WAFER SIZE

HITA EVOLUTION OF DRAM PROCESS TECHNOLO

PROCESS STEPS: 16M/1M DEPOSITION 1.95 OXID./IMPLANT 1.06 KEY TECHNOLOGIES IN THE ETCH 1.11 DEPO. ® WIDER VARIETY OF M PHOTO 1.11 © HIGH THROUGHPUT/

TOTAL 1.24 0.3pm WAFER PROCESSING ^ " O.Siim -MULTI-CHAMBER [ ? ] O.Siim [STACK] 0 LOW TEMP. LARGE W (7^2^ [STACK] OXID. (D RAPID THERMAL ANN [PLANER] IMPL. © DEFECT CONTROL

to ETCH Q. ® LOW TEMPERATURE LU I- 1/1 MAGNETRON (p - WA LO (/> © LOW DAMAGE/DAM LU U O HIGH THROUGHPUT/ O a: Q. WAFER PROCESSING

PHOTO ® HIGH NA, i-LINE/EX 0 WIDE FIELD STEPPER ® PHASE SHIFT LITHOG 1 1M 4M 16M 64M 0 HITA CLEAN TECHNOLOGY

i/> 50% YIELD oo UJ U sV O Q: a. ~1mm Partic E a1 whole baseb Ln 80%YIELD X o ^^ V. 1 A ULTRA CL >• N^ to N» 2: V. LU Q h- 0.1 U LU 256K 1M 4M 16M 64M Q 1 . 1 I.I A 10 10 10 10 10 ELEMENTS/DIE 1 L. 20 50 100 200 500 DIE SIZE(mm^) 0 HIT REDUNDANCY IS A KEY TECHNOLOGY FOR

100 1 1 . , —1 r— • -I •—

80 -

>

60 -

tys

—Q) 40

20 -

r 1^' « - • 1 1 1 1 1,— 20 40 60 80 100

Non - Redundancy Yield (%)

0 HITA WAFER SIZE AND DIE SIZE

10 - •WAFER SIZ -.<^^ y INCREASE

Wafer size / NEEDED A WITH DIE 10 - QQ INCREASE rvj Cd E < a; £ Die<;/Wafe r Cd •8 INCH M N sb- 3 UJ '^ Qj 10 KEY IN 16 ^ N ^^^ ru o; ^Q uo •WHAT AB UJ 10 - Q INCH IN 6

Die si?e

10 1 1 1 • • 1 i __ 16K 64K 256K 1M 4M 16M 64M

DRAM 0 HIT MOS MEMORY MARKET TRENDS IN 90

DIVERSIFICATION OF APPLICATION

REQUIREMENT FOR HIGH PERFORMANCE

PACKAGING FOR HIGH DENSITY ASSEMBLY

0 HIT MARKET TRENDS IN 1990'S

MEM MAIN APPLICATIONS REQUI r

PERSONALIZATION • NOTE PC iHIGH DE • PERSONAL EWS PACKAG SOLID STATE • MEMORY CARD I ULTRA L

MASS STORAGE • SOLID STATE DISK /NON V • FLAT DISPLAY >DIVERSF

I HIGH QUALITY • HDTV ASIC ME

VIDEO/AUDIO • RISC EWS (HIGH SP • LOW BIT

iHIGH PERFORMANCE v^ 0 HIT r DIVERSIFICATION OF DRAM 90'S MULTIPORT VRAM HDTV

FRAME MEMORY D DIGITAL A/V SOLIDSTATE RECORDING C MAIN 64K 256K 1M 4M 16M 64M 256M 1G- FRAME

BIT WIDTH DIVERSIFICATION PC x8/9 X16/18 X32/36- D

HIGH SPEED (BI-CMOS) EWS

PSEUDO-STATIC ULTRA-LOW NOTE RAM POWER RAM PC W SOLID STATE DISK DISK

0 HIT HIGH DENSITY PACKAGING IN SMALL SYST

MEMORY PC VOLUME (cm3) CAPACITY (MB) unit vol/bit 100,000 / SM 64KDtP DO \ M DESK TOP 256KDIP

IMDILP 10.000 10 0.1

0.01 . ^» \ Si —,\«X 4MTS0P \ 1,i)00 1 0.001

PIN TSO INSEPCION ^\ SMD •*- TA ij.. nr"- —I 1980 1990 2000 1980 1990 0 HIT r~ DRAM/PSRAM FOR SMALL SYSTEM

n^ MIGRATION

FROM STANDARD 1MDRAM

womb TO 4MDRAM/TS0P/L0W ) ) DBAtut^^^im&t GIVES 10 TIMES IMPR

IN VOLUME & 1MDRAM/S0J

[256Kx4]x16:2MB DATA RETENTION CU

WIDE BIT, HIGH DENSITY DRAM

HIGH DENSITY PACKAGE AND J L J POWER SPEC !S A KEY TO FUT

SYSTEM APPLICATIONS. 4MDRAM/TSOP/LOW POWER.

[256Kx16]x4:2MB

0 HIT HIGHER SPEED IS EVER INCREASING REQUIRE

1000

ACCESS TIME REQUIREMENT FROM CPU 8086 (5MHz)

C

68020 (12.5MHz) DRAM ACCESS TI LO 880380 6 (12,5MHz) IJO •I- y 100 (kA\{ -y^cAC1 Z JUlV 1M 80386 I (20MHz) I 4M 68030 (20MHz) "-

Bi-CMOS DRAM ACCESS TIME /

10 1980 1985 1990

^ 0 HIT MEMORY AS BUSINESS OPERATION

R&D INVESTMENT

MANUFACTURING INVESTMENT

TRADE ISSUES

0 HITA HUGE INVESTMENT INVEST

INVESTMENT M$

, 15,000-

^

10,000-

5,000 •

0 1980 '81 '82 '83 '84 '85 '86 '87 '88 '89 '90 '91 '92 SOURCE: DATAQUEST 0 HIT TRADE BALANCE BETWEEN USA, EC & JAPA

(SEMICONDUCTOR) BILLION

U.S.A E C

0 HIT COMPETITON & COOPERATION

1990'S IS THE AGE OF INTERNATIONAL BU^NESS INTER0E -NO SINGLE MANUFACTURER CAN SATISFY EVERY SEGh/fE FAST GROWING SEMICONDUCTOR DEMAND-

ALLIANCE

COOPERATION

JOINT INVESTMENT

ACTIVE PURSUIT OF " COMPETITION AND COOPERATION "

REQUIRED BY OVERALL MARKET ENVIRONMENT THROUGH

0 HIT r ISSUES

TECHNOLOGY )

MARKET OPER ( ) c

<0> HIT INCREASING COST: Bl or n

mil

10 - ^^^^=S= CD USER u Bl-rule 'l_ Q.

C Z)

n-rule MAKER

J 1 :—I L

DRAM Generation 0 HIT Dataquest nn a company of MMmm TheDun&BradstreetCorporatK

THE TREND OF NEW CONCEPT DEVICES

Dr. Hiroyoshi Komiya General Manager LSI Research and Development Laboratory Mitsubishi Electric Corporation

Dr. Hiroyoshi Komiya is a General Manager of the LSI Research and Development Laboratory at Mitsubishi Electric Corporation. Prior to this position, he was General Manager of their Saijo Factory, which mass produces VLSI memories. Previously, at Mitsubishi's Kitaitami Works, Dr. Komiya managed the development of memories, microprocessors, and wafer processing technology. He also managed the design, development, and construction of fully automated manufacturing lines for the Saijo factory. Before that, he worked for Research and Development of LSI processing technology and of II - VI compounds as a senior researcher at Mitsubishi's Central Research Laboratories. Dr. Komiya received a B.S. Degree in Physics and a Ph.D. in Science from Kyushu University.

Dataquest Incorporated JAPANESE SEMICONDUCTOR INDUSTRY CONFERENCE April 12-13, 1990 Tokyo, Japan

1290 Bidder Park Drive, San Jose, CA 95131-2398 (408) 437-8000 Telex 171973 Fax (408) 437-0292 |B8ea70oo.iMG oams/qollBrT ULSI Era: Challenges and Opportunities

The Ik'end off New Concept Devices Dr. Hiroyoshi Komlya General Manager LSI Research and Development Laboratory Mitsubishi Electric Corporation

1990 Dataquest Incorporated April 12 — Reproduction Prohibited TREND OF NEW CONCEPT DEVICES

Hiroyoshi KOMIYA LSI R&D Laboratory Mitsubishi Electric Corporation

CONTENTS 1. FUZZY CHIP 2. NEUROCHIP 3. 3 DIMENSIONAL IC 4. SUPERCONDUCTING DEVICE 5. QUANTUM EFFECT DEVICE 6. ACTIVITIES OF THE R&D ASSOCIATION FOR FUTURE ELECTRON DEVICES 7. CONCLUSION

PU:

DcoNCiii'r

]• \t 7.L y 'I'll o () r y : N u Jiie r i c a 1 [i e p r c s e n I, a L i o ii of U ii c e r L a i ii I i c s using M o III b e r s li I p r" II 11 c I. i o II

•X. % = UrM O [S "111"

O.E

Input X risinucrumc

It Input X " Lane, Input Y - Ncdlua, and then output A ' lot

Hula Z Z 71 ltul« I KUIM DQiiioalllon Untecsdent) (Conclusion) inpula outpuli Xeiberiiklp KB lb* ml) I p »n Cenlrold Function KIN Funsllon

u /

1 - p'uzrzY ciiiF>

DFEATURES

•Fast Inference

•Incorporates Knowledge of Human Expertise in Systems

•Tolerates Discrepancies in Rules or External Noise

DAPPLICATION

•Feedback Control

•FeeIing—based Control

•Classification Control

F'uz;?:^-" ci-iip (Br) DPRESENT STATUS

Digital Method Example of Fuzzy Chip Number of Rules Unlimited Membership Function Arbitarary Configurable Performance 58KFLIPS(=Fuzzy Logical Inference Per Second) Inference Time ITMS (7 Rules,2 Inputs/1 Output) 276//S (75 Rules,3 Inputs/2 Outputs) Integration Scale ~60K'Tr Clock Frequency 20Mllz (1 Instruction/2 Clock Cycles) Analog Method Table Look—up

DPROBLEM •Improving Fuzzy Chip Performance and Programmability •Developing Suitable Application Area and Constructing Fuzzy Systems •Providing System Development Tools

DFUTURE •Growing Fuzzy Market •Still need Fundamental Research on Fuzzy •Combination of Fuzzy System with Neuro—Computer NEOFtO 0 1-111=

DCONCEPT •Emulate the Human Brain and Nervous System •Execute High Speed Operation by Learning without Programming or Algorithm

nSTRUCTURE

•Neuron Model Xi : Input (=Output from Other Neuron) .— CELL 80DY Wi : Synaptic Interconnection Strength h : Threshold f : Transfer Function ex. Sigmoid Function —SYNAPSES Z : Output •_DENDRITE 'x^AXON (INPUT) (OUTPUT) Z = f (SWiXi-h) Biological Neuron

Neural Networks Model

Back Propagation Model Ilopfield Model

N E U Ft O C II I F> < H )

DFEATURE

•ParallelProcessing • Se 1 f — Le a r n i n g, Se 1 f —Or gan i z a t i o n • Fau 1 t To 1 e rant ' ;. •Fuzzy Control Processing, Pattern Processing

DAPPLICATION

•PatternMatching Re c og n i t i o n^^-Spe e c h, Character, Image etc. Synthes i s Speech

Optimization Problem • • , • • , Traveling Salesman Problem P o r t f o 1 i o Automatic Pattern Layout for ULSI Design "• NEUFtO CI-IIF> cm) nPRESENT STATUS DPROBLEM

•Neuro—like Computor •Model (Parallei Processing) •Architecture of Massive •Neuro Chip R&D Phase Parallei Operation

Digital Neuro Chip •Device Technologies (ATT, Hi tachi etc. ) DFUTURE Analog Neuro Chip (MIT, Mat su sh i ta etc.) •Still need Fundamental Research Theory and Hardware Optical Neuro Chip (Mitsubishi etc. ) •Development Phase App1 i c at i o n s

•Combination of Neuro and Fuzzy

TTI-iriEE DI]VIE]MSIONA.I_ IC

DCONCEPT

Vertical Stacking of Active IC Layers

DSTRUCTURE

Light

5th layer: Image sensor

4th layer: Pre-processing circuit (A-D) 3rd layer: Memory (RAM) cy-CWjViWAW^^L-^ id layer: Image-processing 5s eircujtjCPUl 1st layer: Post^rocessjng circuit (D-Al

Substrate (Si) A Multi-functional 3-D LSI

- 4 T^i-iFj-EE DI]VIE^srsIOI^r^^.L^ ic

Conventional 2-D LSI 3-D LSI

DAPPLICATION ,:--*-

•HighDensityLogic& Memory

•Super Compact High Performance Signal Processing Equipment

•Intelligent Image Information Processor

T^I-IFtEE DIlVIENSIOlNr^^L IC (Iff)

DPRESENT STATUS DPROBLEM

High Density Devices Still SOI 64K-Bit SRAM^^^(TI) and Other Basic Technologies PchA'ch Stacked CMOS under Development

8K-Bit SRAM (Matsushita) Long Processing Term 4K-Bit CMOS Cell/4K-Bit CNDS Cell for Production

256-Bit SRAM^^^ (Mi tsub i shi ) CMOS Peripheral/N>DS Menioi-y Cell DFUTURE

SRAM Ce 1 \-- (Mitsubishi) AppI i c a t i o n to (0.6 um Design Rule) Simple Structured Device Inter ODS High Density New Functional Devicees Lo g i c & Memo ry

Image Signal Intelligent Devices Processor TEG PHOTO DIODE One Chip Image Processing (Mitsubi shi) with Super 3—D Chip M> CONVERTER Consisting of Monolothic Multilayer and Chip Bonding Structure S U F'E FtC OND UCTT I WG D E X^ I C E (I)

DCONCEPT •Superconducting Effects in Cryogenic or Cooled Operation

DSTRUCTURE

Josephson Device

Insulator Superconductor

Superconductor _^ |A —^

0 Vg (gap voltage) Voltage (mV)

Basic Structure I—V Character!sties

S U F" E R.C OND UCTT I NG DEVICE < H >

DPEATURES •High Speed Switching ?; V X*^ "^ \. •Low Power Dissipation % X ^ - \ \^' \ x^^ - 100 \ n ^ V a ^ GaAs \^ £ • -^ ^«<^ \ \ \. Uli \ ;: \ ^"•"AumaiK KmT^,^ llim^VS' HEMT V i,. \ I.IK ^ s ^ •y J f\ \ ^^"- • »• \ ' \ 1 t k_^ , [— T 10 100 1000 lOOOO POKEBDISSIPATlON/fate (juW) DAPPLICATION •High Speed/Low Power Consumption Logic

•Huge Scale Memory

•Wiring for Wa fer Scale Integration

- 6 suF'EFt.coiNrDucT^iNG'r IDE:"VICE: CM)

DPRESENT STATUS DPROBLEM

Josephson Device •Liq. He Temperature Operation •4Kbit RAM •Incompatibility with Semiconductor 2K Gate Array

4bit CPU (Hitachi, Fujitsu) DFUTURE

8bit DSP •High Temperature Operation (Fuj i tsu) ex. Liq. N2

Superconducting Transistor •Hybrid of Semiconductor with Superconductor Quantum Flux Parametron

Q U y^ NT:^ T-J IVT DEV^ICE CI)

DCONCEPT •Functional Device Using Wave Nature of Electrons Confined in a Quantum Sized Area

Resonant Tunneling Effect DSTRUCTURE

Quantum Well, Quantum Wire, Quantum Box, by Super Lattice Structure Lattice Period Longer than That of Semiconductor Crystal Very Thin Film with Only One Hetero Interface

Quantum well Quantum wire Quantum box

ji f^./iyr ..'t-ycy/i'^j,-. ,rrcii •ciorliW, /y^^iLf'-i tj\-j.v I

DDODC S» DDonrMSi^ iiaaocWiW DcinDC (W' lannr: w -innnrl/ Potential wall Quantum well Quantum wire Quantum box

(a) Quantum Wei 1 Structure (b) State Density vs Energy ( P (e) vs e ) Q U ^^ N TT Tjr IVl DE"V"ICE (II)

DFEATURE DPRESENT STATUS

•Ultra High Speed •Development Pha s e^--Su p e r Lattice HEMT •Low Power Dissipation •Research Phase •Ve ry Sma11 Size •••Incoherent Elecrtron Waves (Tunnel ing, Resonance) Hot Electron Transistor DAPPLICATION Resonant Tunneling Hot Electron Trans i stor (RHET) • Op t o—De vice •Exploratory Research Phase •High Speed, High Density Coherent Electron Waves Lo gic Circuit (Interference, Diffraction) Quantum Interference Transistor •Ultra High Capacity Memo ry

t-: II Al.Ga,.,Aa IS.Mnnil Ji_L pi-CrAn 1 " HOT ELECTHON ^—Qtfa (J^nm) n-Urt, AI,Ga,.,A9 (.1.11 nml

AI,Gii|.,A9 (3.00iini) ii-[;»Ai J. 0.31) f-0.5] n-(;.Aj ^ <-O.I6 Layer Structure Energy Band Structure

Resonant Tunneling Hot Electron Transistor (RHET)

aUJ^TSS^T^UWl DEA/^ICE (HI)

DPROBLEM

•Fabrication Technology

•Si Based Material System Instead of I—V Compound Semiconductors

High Operating Temperature ex. L i q. Nz

Circuit Disign Techno 1ogy/CAD

DFUTURE

•New IC Technology for 0. 1--0. Olyiim '

•Brain Scale Integration & Intelligent System on Chip yVCT^I^^ITIES.OE TI-IE Ft&^ID J>^ S S O C 1 J^IT 1 O JS! EOrt EUTUrtE E EE CTrriOISr DE^/'ICES

Research and Development Schedule of Future Electron Device Project

I'iscal yeai '81 '82 '83 '84 '85 '85 '87 '88 '89 •90 '91 ~ '95 '96-

Phase 1 Phase 2 Phase 3 Superkttice Develope heteroepitaxial Propose novel Demonstrate the Devices growth of atomic layer devices using performance of superlattices device

Phase 1 Phase 2 Phase 3 Thiee- Develope SOI technol­ Progress process Fabricate proto­ Dimensional ogy and 3D-IC process tech. & fabricate type 3D-IC's ICs technologies 3D-IC TEG

Phase 1 Bio- Model biological information Demonstrate & fabricate Electronic process prototype bioelectronic Devices Develope molecular assembly devices technique

Phase 1 Phase 2 Phase 3 HiTc- Construct device Fabricate device Fabricate proto­ Superconductins concept test elements type HiTc-super Devices Develope high Develope micro- conducting quality film pre­ fabrication & devices paration and processing control technol­ technologies ogies

CONCEX-JS I ON

I t em^"^~^Pha s e Present Future

Fuzzy Fuzzy Inference Fast Inference So f twa r e •• Fuzzy Chip

N e u r o Neural Network Step by Step (General Purpose CPU) (Special Purpose CPU) Pseudo—Neuro Chip

3-D Development Phase Application to Simple (Device Technologies) Structured Device

Supe r— R&D Phase Development Phase conducting High Speed Computer Migh Tc Operation

Qu ant um Research Phase R&D Phase Device (Transistor Level) (IC Leve1) Dataquest nn a company of MMMM The Dun &Bradstreet Corporation

RAMPING UP OF THE LSI BUSINESS

Nobutsune Hirai Managing Director Kawasaki Steel Corporation

Nobutsune Hirai is Managing Director of Kawasaki Steel Corporation. He is also Assistant General Manager of their New Business Division, and General Manager of its LSI Works Construction Team. Mr. Hirai has served as General Manager of various departments and divisions including the Planning and Development Department, New Business Division; LSI Department; the New Materials Department; the Planning Department, Mizushima Uorks; the Plate Rolling and Shape and Bar Rolling Departments, Mizushima Works. He was also Staff Assistant General Manager of the Systems Planning Section of the Systems Planning and Data Processing Department at Mizushima Uorks. Mr. Hirai received a Bachelor of Engineering in Metallurgy from Kyushu University, Fukuoka, Japan.

Dataquest Incorporated JAPANESE SEMICONDUCTOR INDUSTRY CONFERENCE April 12-13, 1990 Tokyo, Japan

1290 Bidder Park Drive, San Jose, CA 95131-2398 (408) 437-8000 Telex 171973 Fax (408) 437-0292 |9aee7oio.iMG oo/oy/aojsi| ULSI Era: Challenges and Opportunities

Ramping Up of the LSI Business Nobutsune Hirai Managing Director Kawasaki Steel Corporation

1990 Dataquest Incorporated April 12 — Reproduction Prohibited ACCELERATING DEVELOPMENT OF LSI BUSINESS

NOBUTSUNE HIRAl MANAGING DIRECTOR NEW BUSINESS DIVISION KAWASAKI STEEL CORPORATION

KAWASAKI STEEL BUSINESS DIVERSIFICATION AT KAWASAKI STEEL

/IN 41 COM

KAWASAKI STEEL WHY KAWASAKI STEEL DECIDED_ TOGOINTDLSI INDUSTRY (1) MARKET VOLUIVIE & GROWTH RATE OF LSI INDUSTRY PRODUCTION ] 1 IN JAPAN (Triiiionyen) ]Q

(Data Source :IVIITI, The Bank ol Japan) KAWASAKI STEEL WHY KAWASAKI STEEL DECIDED TO GO INTO LSI INDUSTRY (2)

FAMILIARITY WITH MANUFACTURING

SIMILARITY FROM INDUSTRIAL VIEWPOINT

"RICE" OF INDUSTRY "NEW RICE" OF INDUSTRY

KAWASAKI STEEL ENTRY STRATEGY

TARGET ASICs"

Niche MsM in Japan in 1985 To Avoid Competition with Major Players

ESTABLISHED "NIHON SEMICONDUCTOR, INC." AS JOINT VENTURE WITH LSI LOGIC CORPORATION

In View of Benefit to U.S. - Japan Industriai Cooperation Parachute to the Front of "LSI IVIANUFACTURING"

KAWASAKI STEEL NIHON SEMICONDUCTOR, INC.

8/1985: Established as J/V of LSI LOGIC (55%) and KAWASAKI STEEL (45%)

10/1987: Started Commercial Production of Masterslices

8/1988: Started Commercial Production of Metalized Wafers

1989: Generated Operating Income

1990: Expected to clear off Accumulated Losses

KAWASAKI STEEL NIHON SEMICONDUCTOR, INC.

HUMAN RESOURCES—MANAGEMENT BY JAPANESE"

RECRUIT MIUHM / DISPATCHED NIHON ( FROM 218 LSI LOGIC K.K. (JA SEIVIICONDUCTOR, INCN^ 4 DISPATCHED FROM KAWASAKI 53

KAWASAKI STEEL MEANING OF NIHON SEMICONDUCTOR, INC. TO KAWASAKI STEEL

LSI MANUFACTURING KNOW-HOW

PRODUCTION CONTROL TECHNOLOGY OF STEEL MAKING

KAWASAKI OBTAINED CONFIDENCE TO DEVELOP ITS OWN LSI BUSINESS

KAWASAKI STEEL r LSI RESEARCH CENTER

10/1985: Established LSI RESEARCH CENTER in TECHNICA RESEARCH DIVISION

7/1987: LSI Pilot Line in LSI RESEARCH CENTER started operation

^Number of Researchers: 124 (3/1990)

Plan to create a new facility for Fundamental Research & Next Generation Technology Research

KAWASAKI STEEL U.'IIL; . |- TOCHIGI FACTORY

1985 1986 NIHON SEMICONDUCTOR. INC. LSI RESEARCH CENTER PARACHUTE INTO LSI —TYPICAL LSI R & 0 MANUFACTURING

1990 TOCHIGI FACTORY DEVELOPMENT OF KAWASAKI'S OWN LSI BUSINESS

KAWASAKI STEEL TOCHIGI FACTORY

A- r. J

START OF OPERATIONS : July 1990

CONSTRUCTION COST : ¥28 billion ($200 million) ''';i';'

PRODUCTION CAPACITY : 10.000wafers/month |6 inch wafer

NUMBER OF EMPLOYEES: 200 (in 1994) ,

PRODUCTS : Especially ASICs (ASSPs. ASCPs)

L KAWASAKI STEEL r ALLIANCES

CAD TECHNOLOGY PRODUCT DEVELOPMENT MANUFACTUR

CADENCE DESIGN SYSTEMS, INC.

/' COMPANY A ( LSI LOGIC VCORPORATIO COMPANY B

KAWASAKI AGGRESSIVELY SEEKS ALLIANCES WITH LEADING COIVIPANI

KAWASAKI STEEL KEY TASKS FOR KAWASAKI STEEL

ENHANCEMENT OF PRODUCT DEVELOPIVIENT CAPABILITIES

ESTABLISHIVIENT OF DISTRIBUTION CHANNELS & DESIGN CENTERS

KAWASAKI STEEL ENHANCEMENT OF PRODUCT DEVELOPIVIENT CAPABILITIES

TARGET AREAS FOR ASIC PRODUCTS o Digital Signal Processing o Image Processing

ASIC PRODUCT DEVELOPIVIENT o Self-Development o Joint Development with otiier Semiconductor Manufacturers o Joint Development with LSI Design Companies o Joint Development with Customers

FEEDBACK FROM KAWASAKI INTERNAL CONSUMPTION o Electronic Equipment Business o Information & Communication Business

KAWASAKI STEEL ESTABLISHMENT OF DISTRIBUTION CHANNEL & DESIGN CENTERS

Establishment of Distribution Channels

First Design Center is established at Makuhari Techno Garden (Chiba)

Other Design Center Locations scheduled

KAWASAKI STEEL DataQuest nn a company of MMmM The Dun & Bradstreet Corporation

HALF-MICRON LITHOGRAPEHr

Shoichiro Yoshida Senior Managing Director Nikon Corporation

Shoichiro Yoshida is Senior Managing Director at Nikon Corporation. Prior to this he was Managing Director. Mr. Yoshida joined Nippon Kogaku K.K. (Nikon Corporation after April 1988) in 1956 and has held several managerial positions including, President, Nikon Precision, Inc.; Member of the Board and Director of their Industrial Supplies & Equipment Division; General Manager, Designing Department, Industrial Supplies & Equipment Division; Mr. Yoshida was elected Director of the Semiconductor Equipment Association of Japan in May 1989. He was appointed Project Director, YOSHIDA Nano-Mechanism Project Research Development Corporation of Japan in October 1985. Mr. Yoshida graduated in Precision Engineering from the Department of Technology at Tokyo University.

Dataquest Incorporated JAPANESE SEMICONDUCTOR INDUSTRY CONFERENCE April 12-13, 1990 Tokyo, Japan

1290 Ridder Park Drive, San Jose, CA 95131-2398 (408) 437-8000 Telex 171973 Fax (408) 437-0292 |B86a701MMG Q3/06/90iJ5< p ULSI Era: Challenges and Opportunities

Haif-iUiicron Lithography Shoichiro Yoshida Senior Managing Director Nikon Corporation

1990 Dataquest Incorporated April 12 — Reproduction Prohibited Lithography in 0.5 |Lim Era

Stioichiro Yo$hida Nikon Corporation Senior Managing Director Tendensy of Micronization of VLSI Patter

1965 1970 1975 1980 1985 1990 1 1 1 1 ' - r 10//m o sz w 8//m a ^ T3 ^ ~ 0) ^i 6//m EQ- 4/im 1-5 2.5//m 1.5/im Is 1.0/im -.__ I |0.75//m o.5//m

DRAM 256 1K 4K 16K 64 K 256 K 1M 4M 16M

Grade IC LSI VLSI ^^ Contact /^ Projection X/ Wafew.c.1%^r1 steppex^i^KK^r' S lij Aligner y^ Aligner I >^ o Improvement in N oud 7i o o <:

•p 1.0 C Q)

ba •H

o •H 0.65 4-1 ;3 r-l % 0.5 O \ CO Q) 0.5 Pi 0.35 Alignment Accuracy "v. •s 0

0.1 0.13 1-0.065

0 1980 82 84 86 88 90 92 94 96 Resolution and Focus Depth

R = Ki ^ / NA (1)

D = K2 ?^ / (NAf (2) Wave NA Max Length (NA Today) 0.56 0.50 0.44 0.39 0.35 0.31 0.28 0.25 0.22 0 1 1 g (0.54) Co.r : 0.5 ) ^— L^ 436 0.65 ( 0.5 ) 1 i (0.45) C0.7:; 0.5 ) 365 0.60 (0.7 0.5)

——-*• 1 KrF (0.42) (0.7 0.5 1 ? 248 0.50 (0.7 0.5 J ArF 193 0.50 :o.7 F2 157 0.50 (0.7 1989 90 91 92 93 94 95 96 97 9

Process Coef. K1 (pm)

° TSMR-V3 ( g-line NA=0. 54 )• u. 5.0 o TSMR365iR( i-line NA=0.45 ) o Q 40

3.0

2.0

1.0

OA 0.5 0.6 0.7 0.8 0.9 1.0 L/S [jj m The Depth of Focus for Various L/S Patterns L&S (urn)

0.45

0.35

Resist Profiles of Various Line Width

TSr^R 365-1.2jumt (Center) OPTICAL SYSTEM OF LSA

RETICLE BS1 L1 K S He-Ne LA

PROJECTION —M— SPATIAL FILTE LENS PHOTOELECTR SENSOR {>SIGNAL PROCESSOR ALIGNMENT WAFER MARK Field Image Alignment

FIA Processor CCD Camera Projection Lens Optical Fibre // FIA Microscope

Wafer Laser Interferometric Alignment

A/D Memory

A/D Detector

Reference fi

Wafer UNIFORMALIZER

MA CO

^ PROJECTION LEN

DOSAGE WAFER CONTROLLER ^y V STEPPER MAIN CPU Basic specifications for the excimer lase

Spectrum \ft^dth (FWHM) <3 pm Wavelength stability ±1 pm Average output power 2'-'3 W Pulse energy 10'--20 mJ Repetition rate 100 or 20 Gas life time 10^--10^ p Window cleaning interval 10^-^10^ p Main components life time 10^^10^ p PROJECTION LENS ITV CAMERA

^

/ ^ 7 i He-Ne LASER

WAFER ;^f/ DIFFRACTION LIGHT fin nn t _ i =L DETECTOR

X-Y STAGE ALIGNMENT MARK.

DIAGRAM OF WAFER ALIGNMENT SENSING SYSTEM DataQuest nm a company of MiH TheOun&BradstreetCorporation

THE ROLE OP THIN-FILM TECHNOLOGY IN ULSI

William W. R. Elder President and Chief Executive Officer Genus, Inc.

William Elder is President and Chief Executive Officer of Genus, Inc.» a semiconductor equipment manufacturing company. He is one of the founders of Genua and has served as a director of the company since its organization in 1981. Prior to Genus, Hr, Elder was President of the Wafer Processing Division of Eaton Corporation. He was affiliated with Fairchild Semiconductor in various senior management positions before he entered the semiconductor equipment industry. Mr. Elder has an extensive background in the semiconductor industry with over nineteen years of senior level management experience in semiconductor equipment manufacturing and semiconductor manufacturing and is currently a member of the Board of Directors of SEMI/SEMATECH.

Dataquest Incorporated JAPANESE SEMICONDUCTOR INDUSTRY CONFERENCE April 12-13, 1990 Tokyo, Japan

1290 Bidder Park Drive, San Jose, CA 95131-2398 (408)437-8000 Telex 171973 Fax (408) 437-0292 |BBa87012.lMG OOWlSreolIgl ULSI Era: Challenges and Opportunities

The Roie of Thin-Fiim Technology In ULSI wniiam W. R. Elder President and Chief Executive Officer Genus, Inc.

1990 Dataquest Incorporated April 12 — Reproduction Prohibited jf-^.r^i**-:-:^' j^JhwMriHinC JfAJtMtb. •

Genus Presentation

The Role of Thin Film Technol in the ULSI Era

by William W. R. Elder Chairman and CEO Dataquest Conference, Japan April 12,13, 1990 GENUS CM I O o> CO tiii I iSSlmXlUi^ •V* i^ - t^^n TOP TEN MERCHANT SEMICONDUCT SUPPLIERS 1975 1983

^ffyilll^THZ; i^ Texas Instruments ^^ Motorola W NEC ^^ Motorola i^^ Texas Instruments H; Tosh ^^ Fairchild : #i NEC #^: Hitac

^^M National Semiconductor , #§1 Hitachi Moto

EIH Phillps/Signetics iii #p Toshiba S^^:- Fujits ill NEC l^y Philips/Signetics Texa lilli Hitachi 1^^ National Semiconductor Mitsu

Siemens Intel ^ Intel

___ RCA ;:. Fujitsu ;^i Mats

•'^'2 Toshiba : Matsushita Phili

Source: Dataquest GENUS ^•Ukrf >V\ ^MfWA *l"''CX?L/<' VWflwr- ^ ~:r>^, vted ^^*^ :^. . -.-Xf « TOP TEN SEMICONDUCTOR EQUIPM MANUFACTURERS 1980 1Si4

^^ Perkin-Elmer ^^ Perkin-Elnner # Tok ^H General Signal # Nik ^B Applied Materials # Tokyo Electron ^S App

Kij.gs.;?**""'*" ^^ a ^m Fairchild TSG ^i^:=r Eaton # Adv ^fc Varian ^ GGA # Can ^^ Teradyne m Teradyne fm Gen ^^ Eaton Nikon ^^ Var ^^ Genera! Signal Sentry Schlumberger Hita ^ Kuiicke & Soffa Varian Ter ili Takeda Riken Canon AS

^^^^ Source: VLSI Research, 1990 < o o •II CO 4i

>- a Q^ h- C/)

o z o cc h- o m _i m HI Z

QO

itO

;0 •,•: Sf.i^i SEMICONDUCTOR EQUIPMENT INDUS

1^ Worldwide Sales: $9.2 Blllior! in 1989 1^ Highly Fragmented Industry ^ 500 Equipment Comp - U.S. 68%, Japan 21%, Europe 11% > Most Companies Are Single Product Suppliers ¥ Top 10 Companies Have 41% of Market, Top 50 Hav • Largest Supplier: $634M, 7% Total Market

GENUS MAJOR TRENDS - SEMICONDUCTOR EQUIPMENT INDU

Submicron ICs Drive New Production-Worthy Equipmen Equipment Makes IC Technology Transferable Japan Assumes Leadership in Several Equipment TtGh and Surpasses U.S. Overall IC Mfrs Want Stable, Multi-Product, Technology Supplie Customer/Equip. Vendor Joint Development Efforts Business Shakeout, Consolidation Increases 5/^- Tr''^"^

a.^-s..^-'/-^^

GENUS WORLDWIDE SEMICONDUCTOR EQU MARKET Wafer Fabrication - Geographic Japan • U.S. Europe I ROW 1985 1989 ici 4

$3.4B $5.8B Source; Dataquest, March 1990 GENUS ^•^SS':;-"'-^^r^-"*:: ^Kf!] r' K:" ^mm''m^'<- -.Ji, SEMICONDUCTOR EQUIPMENT MARK Wafer Fabrication - Technology Function III Lithography • Etch and Clean I Diffusion • Deposition* • Implantation 1 Other 1984 1989 ,5=^ -^:^^^

$3.5B $5.8B Source: Dataquest, March 1990 *C GENUS WAFER FAB EQUIPMENT MARKET Thin Film Deposition

MOCVD 1984 MOCVD 1989

1994F MOCVD 5%

$632M $1,152 CVD : PVD.^,-. 49% •.

.t ^ K'4 Source: Dataquest, 1990 J* i ••• * j£^, —_ $2,375M CVD THIN FILM - WORLDWIDE MARK

Metal CVD 1984 Metal CVD 1989 10% Poly 11%

$575M 45%.a^ \ ^ ^W

Source: Dataquest, March 1990 ^^'^j^t^. GB^ $1,150M ift^ r^pe-mi/Ai^^^mMk^ ir^y^^^'-y^s^ SHRINKING DESIGN RULES DRAM ASIC Microprocessor Microns 10

0.1 75 80 85 90 95 2000 Development Year GENUS p-^^mm 3tH JK w^WwfciwAwnMy^ If ^^* DRAM COMPLEXITY FACTORS • Line Width Metal Levels Aspect Ra

4K 16K 64K 256K 1M 4M 16M 64M 256M 1 10

0.1 1974 1976 1978 1982 1985 1988 1991 1993 1997 20

Source: Dataquest, Genus GENUS GENUS TUNGSTEN TECHNOLOGY BEYOND 1G TODAY

9:1 Aspect Ratio

Optimized High Pressure/High Deposition GENUS Blanket W Process _ _ ^ ^ _ .,. „.„ _ . _ . INTERCONNECT TECHNOLOGY - A NEW FRONTIER

Scaling of Devices Has Outpaced Interconnect Capabilit Search for Low Resistivity, Ease of Deposition, Etchabilit Drives Selection of Material for M1 Interconnects ^ PVD Aluminum Displays Step Coverage and Reliability Problems at 1M DRAM P Tungsten Silicide Choice for Gate Interconnect Beginnin • CVD Blanket Tungsten Now in Production for 1-4M DRA • Selective Tungsten Not Yet Production-Worthy • CVD Aluminum, High Dielectric Films and Copper Metal in Research

GENUS '^^^^w^^*'i4tS'>^^^fet»^wj>%l^^C."'fJii^^-*"^^^MW* >;i" ^!'2>< ^'*'w«^dr^ Wi'i^ i^ ^^K'J?^^^^^^^N ^>E^£^^u}i^WzTj:;:§^^i£^j ^l^ ^^^T^'Q^^ ty. %'^-^Sii^^;zf^Ji^V * •.•.^.^:.d% ,^^B CVD EXPANSION

CVD Market Grew 126% from $254M in 1987 to $5 IM DRAM Capacity Increase Doubled Japanese CV in 1988 Japan Is Largest Market for CVD Fabrication Equip U.S. Equipment Companies Hold Largest Share of CVD Market Vertical Tubes Replacing Horizonal Tubes Non-Tube Reactors Take Over Market for Submicro Applications and Large Wafer Diameters

GENUS ROLE OF THIN FILMS IN ULSI TECHN Equipment Driven • Sputtering no Longer Technology Driver • CVD Becoming Dominant Deposition Technology D - Planarization Issues - Increasing Aspect Ratios - Particle Control • Metal Deposition (W) CVD Will Be Process Standar • Non-Tube CVD Increasing • CVD Driving Integrated Process Tool Development P Multilevel Interconnect Is Technology Bottleneck for ULSI Technology (0.35 um)

GENUS ROLE OF THIN FILMS IN ULSI TECHN Film/Process Driven • Interlayer Dielectrics - Between Poly and Metal 1: BPSG, Oxy-NlWcle, PSG, SOG - Between Metal 1 and Metal 2: Al-Si-Cu, Ti/Al-Gu Blanket Tungsten, Selective Tungsten, Gold - Between Metal 2 and Metal 3: PETEOS Oxide, ^ Barrier Metals - Ti/TiN, TiW P Contact Plugs - Blanket Tungsten, Selective Tung • First Via Plugs - Blanket Tungsten/Etchback, Selective Tungsten Source: Dataquest 1990, Genus GENUS . .: • itWTWiT' i-viv,-,^*-;. -Ci!.; INTEGRATED PROCESSING

Wave of the Future for Sub-0.5 urn Crowded Market - 14 U-S. Companies - 5 Japanese Companies - 5 European Companies Only a Few Players Will Survive Partnering Is a Must for "Best of Breed" Techn

GENUS K^^^^-*^"

^J^ii^KwS&^^WQKfe''•iuyjxniwrrju^fffja^ i,3SJ»A^J^^* THE CHALLENGE ??????99?

|)an 3 Globa Eqjuip^|nent (|||j^p 9 7 Gpoperate |||;;;|lrtne^^^ .. • Ji^l fei World Claiijte^^ atecfteystem

v>:.i>-i-:---:-.-.-:-Xr:->.--';>->^;';.;.. SuccessfuilyJojyiarK •"Js^M

? ? ^gi?&S mm ? ? 9 9 9 9 ? ?

GENUS —. V. ^^.:5r.;:^j-"^-. .'.^J«SKift^~:;:.iK,:;:.;;;•:: r-O"^ M^^^-t'-- .vj^Kiv.. m^'-'tAifvSfiiAi'^ ^i MULTI-LEVEL INTERCONNECTS - YEA

64M DRAM With 0.3 Micron Lines Is a Mature Produet X-Ray Lithography in Production DRAM and ASIC Factoriei Require Different Equipme In situ Multi-Processing Equipment with "Best of Breed ASIC/High Density Logic Use 8-10 Levels of Metal Copper CVD Interconnects with Low Dielectric Consta Tungsten Gate Structures with New Gate Dielectric M 20-30 Large, Multi-National Equipment Companies Do

GENUS (A (0 o o 3 a o DataQuest nn a company of lija The Dun &Bradstieet Corporation

ARTIFICIAL INTELLIGENCE AND H7PER MEDIA: THE SEMICONDUCTOR IMPACT ON HYPER MEDIA

Koji Yada President Hyper Media Corporation

Koji Yada is President of Hyper Media Corporation and President of CSK Research Institute. Prior to this, Mr. Yada was Director of Linguaphone Japan Ltd. He established the Knowledge Information Research Institute. From 1960 to 1983, Mr. Yada was with Electrotechnical Laboratory, Ministry of International Trade and Industry. His development activities included conversational adding machine systems, computer control systems, robot for ocean research, research information processing systems, optical communication systems, CAD system for LSI design, expert systems, and security systems. Mr. Yada has received awards from the Japan Information Center of Science and Technology, and the Information Processing Society of Japan. He has been published many times including Understanding Software, Dictionary of Microcomputers, The First Step in Electronics, AI Vill Change the Future, A Micro Computer Handbook. The Story of AI, An AI Primer, and AI Handbook. Mr. Yada graduated from the University of Electric Communication, Tokyo, Japan.

Dataquest Incorporated JAPANESE SEMICONDUCTOR INDUSTRY CONFERENCE April 12-13, 1990 Tokyo, Japan

1290 Ridder Park Drive, San Jose, CA 95131-2398 (408) 437-8000 Telex 171973 Fax (408) 437-0292 |BMe7oi3.wQ (o/oyrao^BT^ ULSI Era: Challenges and Opportunities

Artificiai inteiiigence and Hyper IMedia: Tlie Semiconductor Impact on Hyper IMedia Koji Yada President Hyper Media Corporation

1990 Dataquest Incorporated April 12 — Reproduction Prohibited AI&HYPER MEDIA

HYPER MEDIA C KOJI Y

I Trend of Al Business

11 A View of Hyper Media Business

Expect to Semiconductor Teclinolo The Trend of AI Business 1. Platform for Al

•LISP Machine •Work Station •Mainframe •Personal Computer •Game Machine •Embeded Machine 2. Al Integration

Various Platform

•Fusion to Data Processing

System Integration using Pattern Pecognitio 3. Al Application User

Al Development Division

•Data Processing Division

End User 4. Expert System Tool

Shell

DST (Domain Specific Tool)

Embeded Expert System 5. Hyper Media Technology

Audio Visual

CD-ROIV!

High Vision (HDTV) A View of Hyper IVIedia Business 1. Computer Software and Hyper Software

Computer Software Hyper Software

Text Media Image Media

Logic Base Sensitivity Base

Computer Culture Traditional Culture

Business Sense Maniac Sense 2. Technology of Hyper Software

(1) Reasoning System (3) Dispiay • Expert System. •Clear Vision •Object Oriented Programming •Higii Vision •Actor Oriented Pfogrammlhg (4) WIechatronics (2) Human interface • Robotics •image Understanding •Intelligent Se •Speecli Understanding (5) Muiti Media •CD-ROM CD-ROM XA 3. Hyper l\1edia Dedicated System

mTOWNS

CD-I

DVI 4. Market for Hyper Media

Vertical Market

PC Market

Home Use Market 5. Application for Hyper Media

Business

Industrial

Entertainment

Education Expect to Semiconductor Teciinolo 1. State of Arts

386 Micro Processor IBM 370/16 Operation 4—5 MIPS 4 MIPS Speed Memory 4 Gbyte 4 Gbyte Capacity (4X10' byte) (2X10' byte Price ¥1,000,000 ¥500,000,00 For a Many User For One Person People 2. The Growth of Micro Processo

Number of Integrated Transister

M 10' 1

25,000,000 10' 4,000,000

10^ 486 CPU 1,200,000 CPU 10- "^86 275,000 CPU 130,000

1980 1985 1990 1995 2000 3. Future Micro Proseccor

•^^^"""CPU ~^^^ Super Computer

^Neural] (Fuzzy]

[Audio] [Graphics (^Photograph)

(Real Time Video]

Application Software DataQuest nn a company of MBMI The Dun SiBradstreet Corporation

SUPERCONDUCTOR AND JOSEPHSON COMPUTER: THE ROLES AND IMPACT OF THE 1990s

Dr. Hiroshi Kashiwagi Director-General Electrotechnical Laboratory Ministry of International Trade and Industry

Dr. Hiroshi Kashiwagi is Director-General of Electrotechnical Laboratory and also serves as Expert Member of the Science and Technology Commission. Dr. Kashiwagi joined Electrotechnical Laboratory in 1963 and has been engaged in research related to laser technology. He pioneered in optical fiber and optical integrated circuit fields in Japan and started an experiment on optical LAN, the first in the world, which is partly used in the Research Information Processing System (RIPS) network at the Agency of Industrial Science and Technology in Tsukuba. As Director of the Computer Department, Dr. . Kashiwagi led the research on supercomputers, including data flow, non "Von Neumann" type parallel processing machine "SIGMA-1" and Josephson devices. Dr. Kashiwagi received his Ph.D. from Keio University.

Dataquest Incorporated JAPANESE SEMICONDUCTOR INDUSTRY CONFERENCE April 12-13, 1990 Tokyo, Japan

1290 Ridder Park Drive, San Jose, GA 95131-2398 (408) 437-8000 Telex 171973 Fax (408) 437-0292 |Bflag70H.IMG 03W7/30:JSlt ULSI Era: Challenges and Opportunities

Superconductor and Josephson Computer: The Roles and Impact of the 1990s Dr. Hiroshi Kashiwagi Director-General Electrotechnical Laboratory Ministry of International Trade and Industry

1990 Dataquest Incorporated April 12 — Reproduction Prohibited THIS PRESENTATION WAS NOT AVAILABLE AT

TIME OF PUBLICATION

Dfflaquest Incorporated, a company of The Dun & Bradstreet Corporation 1290 Hidder Park Drive. San Jose. CA 95131-2398 / (408) 437-8000 / Telex 171973 / Fax (408) 437-0292 Dataquest nn a company of If jl The Dun &Bradstreet Corporation

NEURAL NET APPLICATION IN ULSI ERA

Masaka Ogi President Fujitsu Laboratories Ltd.

Masaka Ogl is President of Fujitsu Laboratories Ltd. Mr. Ogi joined Fujitsu Limited In 1952 and has held several managerial positions including Executive Director, Fujitsu Limited; President, Fujitsu America, Inc.; Managing Director and Board Member, Fujitsu Limited; Deputy General Manager, International Operations Group; Deputy General Manager and Vice General Manager, Communications Systems Group; General Manager, Radio Communications Division; and Manager, Satellite Communications Laboratory, Fujitsu Laboratories, Ltd. Mr. Ogi received a Bachelor of Electrical Engineering degree from the University of Tokyo.

Dataquest Incorporated JAPANESE SEMICONDUCTOR INDUSTRY CONFERENCE April 12-13, 1990 Tokyo, Japan

1290 Ridder Park Drive, San Jose, CA 95131-2398 (408) 437-8000 Telex 171973 Fax (408) 437-0292 |flee870i5.iMG o3W7/gola1 ULSI Era: Challenges and Opportunities

Neural Net Application In ULSI Era Masaka Ogi President Fujitsu Laboratories Ltd.

© 1990 Dataquest Incorporated April 12 — Reproduction Prohibited THIS PRESENTATION WAS NOT AVAILABLE AT

TIME OF PUBLICATION

Data^iest Incorporatsd, a company of The Dun & Bradstreet Corporaion 1290 Ridder Park Driva San Josa CA 95131-2398 / (408) 437-8CraO / Telex 171973 / Fax (408) 437-0292 DataQuest nn aa^ompanyo company off The Dun & Biadst reel Corporation

HDTV~ITS PRESENT STATUS AND FUTURE PROSPECTS

Kenji Aoki Managing Director Planning & Administration Headquarters for Affiliated Corporations Japan Broadcasting Corporation (NHK)

Kenji Aoki is Managing Director, Planning and Administration Headquarters for Affiliated Corporations, Japan Broadcasting Corporation (NHK). Mr. Aoki joined Nippon Hoso Kyokai (Japan Broadcasting Corporation), originally assigned to their Kushiro Broadcasting station. He was then assigned to their Current Affairs Program Division, News Department, General Broadcasting Administration. He has since held various managerial positions, including Senior Director and Manager of the Current Affairs Program Division, News Department; Deputy Director, News Department; Senior Executive Vice President, NHK Enterprises, Inc. Mr. Aoki graduated from the Department of Literature, Tokyo University.

Dataquest Incorporated JAPANESE SEMICONDUCTOR INDUSTRY CONFERENCE April 12-13, 1990 Tokyo, Japan

1290 Bidder Park Drive, San Jose, CA 95131-2398 (408) 437-8000 Telex 171973 Fax (408) 437-0292 |Ba8e7016.WQ 03M6;90J51 ULSI Era: Challenges and Opportunities

HDTV -- Its Present Status and Future Prospects Kenji Aoki Managing Director Planning & Administration Headquarters for Affiliated Corporations Japan Broadcasting Corporation (NHK)

1990 Dataquest Incorporated April 12 — Reproduction Prohibited PRESENT AND FUTURE PROSPECTS OF HDTV IN JAPAN

KENJI AOKI

NHK, JAPAN BROADCASITNG CORPORATION 1. Introduction

The number of television units in Japan at present totals an estimated70 million units, and the figure climbs to some 140 million for the United States. About 200 million units are in use in China, although the majority of these are of the black and white variety. On a world- wide scale, the number of TV receivers in current use is believed to total approximately 700 million units.

These figures indicate the potential for a gigantic new industry. If all of the 700 million TV sets were replaced today by HDTV sets, the economic impact would be enormous and beyond the calculation skills of industry analysts.

While it is important to discuss the HDTV issue from a macro-economic point of view, it is also essential at this time to carry out a thorough discussion of the potential of HDTV as a medium. HDTV boasts picture quality comparable to that of 35ram film, its sound quality is CD-comparable, its pictures are easily convertible to those of other media and, above else, it has diversified application potentials for transmission and record-keeping. All these attributes point to the fact that this technology can occupy top place as a medium in an increasingly information-oriented society. HDTV represents not merely an advanced television format for the next generation of broadcasting;it opens up the door to a wide range of applications in areas other than broadcast i ng.

Following is a list of industries that have attempted to apply HDTV technology to date: motion film, CATV, printing, publishing, education, medicine, art-museum, museum, HDTV theater, video programm-ing, video game, event, fashion, commercial film, airport information system, TV conference, computer graphics, CAD/CAM, etc. And the development of HDTV for different applications has just begun.

2. Attempts at HDTV Software Production

In June 1989, a daily one-hour Hi-Vision broadcast (aired 2:00-3:00 p.m.) on an experimental basis got underway. In addition to live sports, the primary components of software packages, HDTV programs produced both in Japan and overseas are also shown. These programs

1 - are also displayed to the public at Hi-Vision exhibits set up at aboutlOO sites around the country, including train stations, department stores, post offices, broadcast stations, manufacturers' commercial exhibit sites and city halls. A full-scale Hi-Vision broadcast service, to be offered jointly by NHK and commercial broadcasters, is scheduled for the fall of 1991. A spare channel of the BS-3b, a back-up satellite for BS-3a, will be secured exclusively for Hi-Vision broadcasting.

There have been positive signs from the general public. In September 1989, NHK broadcast the Seoul Summer Olympics Games in the HDTV format for a total of 73 hours. Hi-Vision receivers were set up at 81 public exhibit sites all over Japan, attracting a total of 3.72 million people to the world's first Hi-Vision broadcast.

A survey conducted at this time revealed that 9396 of the people watching the Hi-Vision broadcast were interested in Hi-Vision, and 7596 replied that they want to buy Hi-Vision receivers. Forty-eight percent of those polled said they would buy a Hi-Vision set if the price was within the range of two to three times the retail price of current TV sets.

The impact of HDTV will be substantial, resulting in high quality video pictures and a cinema-scopic screen. But it will also result in a considerable diversification of applications creating a new video/ image system for the 21st century. It is expected that it will have a substantial influence on people's life, culture and business as a whole.

Naturally, the production of software will not be limited solely to TV programs, but will include a wide variety of software. In addition to simply broadcasting HDTV productions on TV, a series of activities that would never have been possible in the realm of the conventional TV, have become realistic all because of the fine quality of the HDTV picture. Examples include organizing an exposition or a trade show utilizing HDTV software, publishing a book using high definition pictures and converting them to a 35-mm film, and commercializing the home video. The present situation regarding software production is evidence that "broadcasting is only one of the many applications of HDTV. "

A major focus of NHK at present is the organization of large

:2 — international co-productions in HDTV. "The Ginger Tree," which NHK co-produced with BBC and WGBH in 1989 is a typical example. After its completion, the program was converted to PAL and broadcast in Britain first. It was also converted to NTSC and shown to the viewers in Japan where later it will be shown in HDTV as well. American public will also be able to see the program in NTSC.

Another joint work already completed is the production of a record- marking fifteen-hour-long HDTV video of "Nibelung Ring, " an opera in four parts written by Wagner, which was performed in November 1989. In collaboration with the Bayer National Opera Theater of West Germany, NHK recorded all of the four opera performances in HDTV, which would then be later broadcast in Japan. The plan under current discussion calls for screening the opera video in a HDTV theater where a limited number of viewers will be able to savor the art of opera in an elegant and relaxed setting.

Another possibility is the sale of HDTV-disc recordings of the opera series and/or video cassettes converted into NTSC. Such artistic records in HDTV will prove to be priceless human assets in the future.

NHK plans to broadcast about 400 hours of HDTV programs in 1989. And the number of HDTV broadcasting hours will dramatically increase during the next decade. Opinions differ regarding the type of programs most appropriate for HDTV broadcasts, but the primary ones will be sports such as baseball and football. In my personal opinion, however,I believe HDTV will also enhance news broadcasting with its full scope of features. Already at NHK, research has been conducted to determine the most effective size of video or film images, as well as the most effective display position of these pictures on the screen for news broadcasting.

Conventionally two types of cameras, NTSC and HDTV, are used for sports events coverage when broadcasting in both NTSC and the HDTV; but recently only HDTV cameras have been used for shooting with the video pictures simultaneously down-converted for broadcasting in NTSC. A plan is gradually being implemented to replace all the NTSC cameras with HDTV cameras in large studios to tape entertainment programs and down-convert for the conventional broadcast. 3. Industrial Applications of HDTV .•

In Japan HDTV has been applied to many industrial areas other than broadcasting. Medical application is one such area that is attracting a high level of attention. A microscopic brain surgery operation was recorded with a HDTV camera for the first time at the Shinshu University Hospital in December 1987. By plugging a HDTV camera into a microscope and projecting finely detailed pictures of the surgical process on a large screen, it was possible for a number of the hospital's doctors and medical students to view the surgery which would otherwise be available only to the physicians involved.

Furthermore, another brain surgery operation was videotaped again at the same hospital in February 1989, but it was taken with a 3-D camerathis time. By linking two HDTV cameras to both the microscope's right and left eye pieces, pictures were then recorded simultaneously. The educational effect far exceeded expectations; witnessing the wonder of a human, body that is offered through a vivid high definition picture of a brain was an emotional experience.

The application of HDTV in the world of art is also making progress. The Gifu Art Museum in Gifu Prefecture recently celebrated the opening of its HDTV Art Gallery in April 1989. The HDTV system there is designed to allow guests and professional researchers free access to all of the art pieces on hand at the museum; still pictures of all the museum's paintings were taken using an HDTV camera and then stored in a CD-ROM data base. In addition to descriptions and commentaries on paintings, provided both in caption and oral narration, the viewers can also magnify any portion of the picture which is projected on a large screen. All these service functions had never been possible before at any conventional museum.

As a result of the introduction of the new system, art lovers and other interested people have swarmed to the art museum. Attendance hit a record in the first month of the operation of the HDTV Art Gallery. The number of visitors reached the same level as that of the entire year of 1988. The new system is attracting a great deal of attention among art museum officials, both domestic and international.The same system has been introduced at the Saison Art Museum in Tokyo, and a new museum in the suburb of Tokyo (Machida-city) recently decided to adopt the system. HDTV has another great potential in the application of generating tour guide information and ecological video pictures. The dynamic scenery of the Grand Canyon, caught through an HDTV camera attached to the front of a small jet plane, has no comparison whatsoever to any video images produced until now.

An attempt has been made to use closed circuit television in combination with a communication satellite. In November 1989, scenes of a Japan Cup Race, held at the Fuchu Horse Race Track in Tokyo, were taken with seven HDTV cameras, and the satellite was used to transmit the images to three different race tracks in Japan, where the viewers were able to see the Japan Cup Race on huge screens. The following December was also another successful month. A lively concert scene of one of the most popular Japanese rock music groups, filmed with HDTV cameras, was transmitted via communication satellite, and shown at a charge at five live houses around Japan. The plan proved to be a big success.

A gas company in Osaka recently installed two HDTV screens at a gas processing plant. One display is for showing sharp and detailed pictures of actual work on the operation line at the plant, while the other one displays computer graphics of the work process. Operators can supervise the entire operation by keeping their eyes on these two screens. An entirely new use for HDTV linked to a computer network was thus invented.

In the publishing area, attempts have been made to produce animal and plant encyclopedias, art encyclopedias, and general encyclopedias using CD-ROM or CD-I. HDTV image signals are being converted into digital numbers, then processed by computers to create photoengraving for printing. Several HDTV photography books made with this process are already on sale in book stores.

However, the area with the greatest potential for HDTV apart from broadcasting is probably the movie picture industry. In film-making, HDTV will greatly reduce filming costs and allow for the creation of special effects with ease. This was not possible with conventional film methods. Currently, thousands of copies of a movie are made for commercial distribution, but HDTV will definitely save on the costs of copying films. More importantly, however, it will make possible the simultaneous transmission of movies through satellite and fiber optics, and bring about a revolution in the movie distribution system. HDTV technology has already been used, although partially, in at least seven movie productions in Japan, including those of Akira Kurosawa.

The plan for mini-movie theaters incorporating the HDTV system is already within the reach. While they are now seriously being studied from various angles in Japan, a plan to link 14 HDTV theaters is being realized in Florida, U.S.A.

Conclusion

We do not need to go through all the classic stories about historic innovations and discoveries to be convinced that the history of man's advancement would not have been possible without intellectual inspiration, supported by the good will and cooperation of a number of people.

In past years, there have been a number of discussions held repeatedly to determine an HDTV world standard. It appears, however, that these have only led to an escalation in hostile debate, thwarting joint efforts to develop an advanced medium. HDTV, which is not a movie or a simple TV but a new medium, requires a pooling of different experiences and technical creativity in order to develop in the future.

6 Dataquest nn a company of MMMM The Dun &Bradstreet Corporatioln

EVOLUTION OF TELECOMMUNICATIONS NETWORKS AND LSI TECHNOLOGY

Dr. Iwao Toda Executive Vice President and Senior Executive Manager NTT Research and Development Headquarters Nippon Telegraph and Telephone Corporation

Iwao Toda is an Executive Vice President of Nippon Telegraph and Telephone Corporation (NTT). Since 1958, has has been with NTT, vhere he participated in and managed many of software and hardware development projects, including DIPS and DCNA computer-communication system projects. He is currently the Senior Executive Manager of NTT Research and Development Headquarters. He is currently Senior Executive Manager of their Research and Development Headquarters. Dr. Toda received his B.S., M.S., and Ph.D. degrees in electrical engineering from the University of Tokyo in 1956, 1958, and 1964, respectively. He was a vice president of the Artificial Intelligence Society of Japan. In 1987, he chaired the Information and System group of the Institute of Electronics, Information and Communication Engineers of Japan (lEICJ). Dr. Toda is an editor-in-chief of the Japanese translation of IEEE Software Magazine and is a Vice President of the Information Processing Society of Japan. He is a Fellow of IEEE.

Dataquest Incorporated JAPANESE SEMICONDUCTOR INDUSTRY CONFERENCE April 12-13, 1990 Tokyo, Japan

1290 Ridder Park Drive, San Jose, CA 95131-2398 (408) 437-8000 Telex 171973 Fax (408) 437-0292 |B5ee7oi7.iMG ooffisrao^jsi) ULSI Era: Challenges and Opportunities

Evolution of Telecommunications Networks and LSI Technology Dr. Iwao Toda Executive Vice President and Senior Executive IVIanager NTT Research and Development Headquarters Nippon Telegraph and Telephone Corporation

© 1990 Dataquest Incorporated April 12 — Reproduction Prohibited Evolution in Telecommunicatio Network and LSI Technologies

ApriMS, 1

Iwao TODA Executive Vice President Senior Executive IManager Researcli and Development Headquarters N" Contents

-LSI's in Telecommunications Network

-Technical Trends in Telecommunications Services

-Trends in Telecommunications Systems Technology

-LSrs for Telecommunications Network Evolution

-Summary N 's Telecommunications Networks JS^. Videotex Network "T^ 30 Facsimile Communications 95,593 Network 33^ QOO Mobile ^^ '~~°^i, 1

^ 'Vc^ Teleph FAX ISDN and Telephone Network Dsts [Free Dial ^- 1Z_1 51,117 iVoice Bulletin Board ISD

Data DDX-P (Packet Switched Network) 15 Data / tittiSI V^ (^DDX-C (Circuit Switched Network) J—' Telex 3

(As in September, 1989 r Telecommunication Equipments Market in Ja

10 "Electronic Industry in Japan" Japan Electronic Industry Assosiation

CD other Industria > Electronic Equ (17%)

t5 0) Electronic Com N (53%) (7) 0) -i^ V. 05 Telecomm.Equ (30%)

1980 1985 1990 Year LSI's in Current Telecommunications Networ Terminal Systems Control & Managanent Sys -Computers -Telephone -FAX -Data -Processors -Image -Memories -Mobile Transport Systems -Circuit Switching -Frequency Synthesizer -Packet Switching -AD/DA Converter(Voice) -Wire Transmission -AD/DA Converter(Video) -Radio Transmission -CODEC(Voice) -CODEC(Video) -Time Switch -Packet Control -MUX/DEMUX -MMIC -Viterbi r Technical Trends in Telecommunications Sevic

-Multimedia Transportability: Voice, Character, Image, Video Super Digital, INS-NET 64/1500 B-ISDN(150Mb/s) -Portability : Pager, Mobile Telephone

-User-friendliness : Intelligent services controlled by NSP (Network Service Control Point), NSSP (Network Service Support Point) Multi-Media Communication Sevices r High-speed Digital Services

6,000 High-speed Digital B 5,000 Transmission Services • mi^m 13 (NTT/NOG) o (64kb/s-6Mb/s) • MMMi 4,000 o H— 2 3,000 CD -Q E 2,000 INS-NET 64/ 2 1,000 -

1984 1985 1986 1987 1988 1989 1 Year Intelligent Network

NSSP ntel Laye NSP

Tran LS LS Laye ao

NSSP : Network Service Support Point NSP : Network Service Control Point Personal Communication Services Indoor Outdoor Home /'-^ -Mobile Telephon Portable Telephone

IC Cardphone

Personal ID Service Shop/Store -Portable Telephone -Pager Subway Statio v^ r Trends in Terminal Systems

-Transportability : Voice/Character/Image/Vid

-Portability : Small Size, Light Weight

-User-friendliness : Recognition, Synthesis

V- (6>i)}L|6!8M o o H !2!

I o CD (J) 0 c o o c . /•• Q- 0) / i _g) E / CD • >/ D)

• i^^^m -Q ^ lO CC O / ^1 00 m

O "D 0)

o 00 ,v-'^ .• 05 o o o o o o o o (00)9LUn|0A J Trends in Transport Systems

-Fiber to the Home

-Coherent Optical Transmission

-Asynchronous Transfer Mode(ATM)

-Digital Mobile Communications

^ Optical Transmission Systems Trend 1 1 1 1

10G ^ /(o) X-IOG F-1.6G o 0) CO 1G r;^ • mm^ F-400M y6 0 "cS S1OOM — F-100M 0/ DQ F-32M cf

10M 1 1 1 1 1970 1975 1980 1985 1990 Year Asynchronous Transfer Mode

Originating terminal Destination te r Control and Management Syste (Functions of NSP/NSSP)

OAdvanced Network Services -Free Dial -Voice Bulletin Board -Closed User Group -Personal Number Program

OAdvanced Network Managements

-End Customer Control -Software Defined Network Free Dial Service

NSSP

03-509-3333

0 120-123456 TS u LS LS

7 0422-59-2222 03-509-3333 v_ Typical LSI Functions in Telecommunications Netwo OAnalog Information •Voice AD/DA •Audio Conversion •Video /

ODigital Information Transmis CODEC Recognition 1 mage Synthesis Written Character -Binary Data") -Coded Character AD/DA Converter

20

18 Limitation (Noise & Jitter) •^f^*M'^\ 50 Q Resistor W '^P^^

••^'«-' V (Signal Amplitu 16 «-^..^. |14 izer ^^*v

0*12 1 Video Fine BiCMOS [^(HDTV) §1° Measu O CMOS Video Instrum < 8 (NTSC) Present Status BiCMOS 6 Bip. 4 GaAs-MESFE Ik 10k 100k 1M 10M 100M 1G Sampling Rate (sps) Voice CODEC and LSI Applicat Telephone o 64k ^ Audio CD '. 64k; 0) :^ 32k B 3//m CMOS "c5 A/D,D/A DC 16k 2.5//m CMOS if) ~5M0PS DSP .52 8k E 1.5//mCM0S CO -lOMOPSDSP 03 V- 4k [- h- Submicron CMOS >50MOPS DSP 1975 1980 1985 1990 1 Year Video CODEC and LSI Applicat

1G NTSC TV HDTV "cT High Q |100M '3^ (Fu B 10M Video) CO DC (TV Conference) § 1M Deep Submic (/) BiCMO±S C/) E100K (/) Submicron ^. MK; V 05 10K CMOS h- \ Low Bit 1 I 1975 1980 1985 1990 19 Year Voice/Video CODEC and Recognition

CO DL O

"D 0) CD Q. 0) D) C "co CO 0 o O QI 10K 100K 1M 10M 100M 1G Memory Size(Byte) Optical Transmission Technology

1T

O100G

0 -I—• 10G CO DC Bipolar (SST) 1G

1985 1990 1995 Year v_ r Mobile Communication Terminal

10,000

•3;imCM0S GaAs MMIC o^ 1,000 Technology o^

10 1985 1990 1995 2 Year Summary Technical Trends in Telecommunications Serv -Multi-Media Transportability -Portability -User-friendliness

: Trends in Telecommunications Systems Techn -Transport Systems---Optical Transmission, ATM -Control and Managament Systems—NSP,NSSP -Terminal Systems—Multi-media, Personal, Complex, Intell High-speed v^

LSI Technology Evolution Toward 21st Centur -Ulta High Speed IC-LSI—Tera-Herz (1(3^HZ) Operation -Ultra Large Scale LSI——Mega Gate (10 gate/chip) Integrat -Low Power ASIC Femto Joule Devices ^^^^^^^^^^^^^^^^^^^^^^^^^^^^m^^^^^^^^^^^^^m^^^^^^^^^^^^^^^^^^ N DataQuest nn a company of MMMm TheDun&BradstreetCorporatK

THE EVOLVING ROLE OF RISC TECHNOLOGY IN ULSI

Robert C. Miller Chairman, President, and Chief Executive Officer MIPS Computer Systems

Robert Miller is Chairman of the Board, President, and Chief Executive Officer of MIPS. Prior to joining MIPS, Mr. Miller was with Data General Corporation where he was Senior Vice President of the Information Systems Group. He had responsibility for all computer systems, as well as marketing and sales development for the federal marketplace. Previously, Mr. Miller was with IBM Corporation in a series of senior technology and product development assignments, culminating in Director of IBM's Boulder, Colorado laboratory., Mr. Miller holds six U.S. patents and has authored a number of publications related to computer architectures. He is a senior member of the IEEE Society, is a registered Professional Engineer, and on the board of directors of Bucknell University. Mr. Miller received a B. S. degree in Mechanical Engineering from Bucknell University and an M.S. degree in Thermodynamics from Stanford University.

Dataquest Incorporated JAPANESE SEMICONDUCTOR INDUSTRY CONFERENCE April 12-13, 1990 Tokyo, Japan

1290 Ridder Park Drive, San Jose, CA 95131-2398 (408) 437-8000 Telex 171973 Fax (408) 437-0292 |iM»7oa).iMS auosKoMi ULSI Era: Challenges and Opportunities

The Evolving Role of RISC Technology In ULSI Robert C. Miller Chairman, President, and Chief Executive Officer l\/IIPS Computer Systems

1990 Dataquest Incorporated April 12 — Reproduction Prohibited RIS

1990s: Era of RISC

RIS Robert C. Miller, CEO MIPS Computer Systems What Do You Mean by "Era"?

Technology

Micros

Minis

Mainframes

Style of Inter­ Time Batch active sharing processing 360/370 VAX x86 „7\ Computing ^^^x^ Proprietary/vertical Integrated/plug compatible

Open systems/technology Specialists/standards-based

Business Model The Era of RISC: The Big Picture

Microprocessor-based machines dominant across th entire breadth of the market Steady 50% to 100% annual increases in performanc for the next five years Comparable reductions in cost per mips New generation applications which appeal to broade groups of end-users Order of magnitude inGfeases in microprocessor volu The emergence of ona Of possibly two dominant CP architectures The 1990s RISC Explosion

RISC

iPrice/performaneB leadership

KlbU ^ Ability to take .^»NRISC \ advantage of new Increasing / '^ semiconductor dependence --^ on networks technology /A\ Demand for Changing styles more mips of business at desktop

RISUC RISC The Era of RISC: Technology or "The Beauty of a Low Transistor Countjj Takes advantage of innovations in semiconduotor technology sooner Incorporates computer architecture innovations easte Smooth transition path to 64-bit architecture Makes better use of chip area for purposes of integr - At high-end: on chip caches, floating point processing - At low-end: (embedded controllers) I/O and peripheral interfaces Riding the 1990s Performance Explosion*

RISC

Faster clock /ratve - ECL BiCMOS RISC ^^^ Lower cycles GAS Software- ^^\ RISC ^ per instruction - I I Better / \f>^ Superscalar ^ L-* compiler ^^^ Syperpipeline optimization Dynamic scheduling Integration A Systems of system /> engineering- function Parallel processing .(as nodes) \J RISC RISC * The key to a good RISC design is balancing these forces The Era of RISC: Business or "The Revenge of the Marketplacefj Open Systems: Customers always wanted freedom but felt trapped. RISC offers enough performance to lure users away from proprietary systems - Trends are accelerated by macro changes in business climate, e.g. increased competitiveness emergence of global markets Semiconductor houses become high-end CPU vend Value-added comes from process technology CPU design returns to systems houses The Era of RISC: Applications or "Who Will Eat All Those Mips?Jj

The "Mipsburger"

At high-end: competes with minis, mainframes In middle: provides workstations with new functiona for networking, high-resolution graphics, and human responsive interfaces Underneath: innovative new embedded control applications User Interface Consumption of Mips

mips Human responsive user interface 50 - Co sp - Hi 40 Graphical user an interface, vis 3D graphics 30 Aipha-nunneric user interface 20

10

1980 1985 1990 1995 RISCmyths versus the Weil-Balanced RIS

• Aren't RISCs only good for technical workstations? No. Some are used in commercial applications 50% of MIPS servers are sold into multi-user co applications, running databases, including Pick/ • Aren't RISCs only good for C and Fortran? No. RISCs support many other languages MIPS has supplied COBOL, ADA, PL1 since 198 At least 3 third-party COBOLs are available A MIPS M/2000 is about equal an Amdahl 5860 • Aren't MIPS RISC chips limited to UNIX? No. > 50% elsewhere. Laser printers, switching machines, avionics, automobiles, high-speed controllers, etc. ^ MIPS Computer Systems: A Product of the RISC Era

Designs and supports a CPU architecture Licenses manufacturing rights to five semiconductor partners worldwide Purchases CPUs for its own systems division Licenses system design to manufacturers worldwide For Example ... MIPS Chips in

Workstations: MIPS, DEC, SGI, SONY, Sumitomo, NE Servers: MIPS, DEC, SGI, CDC, ERSO, HCL Multi-user commercial: MIPS, Bull, Nixdorf, CDC, Pyra Multi-user commercial, fault-tolerant: Tandem, RC Co Real-time: Concurrent Computer Laser Printers: Kodak and many others Avionics: Westinghouse and many others Switching: Northern Telecom and others Controllers: Array Technologies and others Automobiles: Several A small sample ... Uniprocessor CPU Performance Trends

1000

Past F im 4 - RISC microprocessors 1990

10

1 - Mainframes - S/370 1 - VAX- 2 - Superminis - VAX mips (real CISC microprocessors pgms) 1 I I L 1965 1970 1975 1980 1985 1990 1995 2 Year delivered 'An approximation Final Thoughts: Evolution vs. Revolution

How long does it take for a new idea in architecture catch on? Key metric is how it affects the user - transparent, su recompile, rewrite programs The more revolutionary the longer the time to transfe from the lab to practice ^^ ^c^V^ isO^ .^ -NA«>y^ x\v •iT V^^x \<^ •d^ ® .^

Massively parallel: just Doming into market Optical computing: in research labs Biological computing: on the way to the lab

The 21st Century: Era of ??? DataQuest nn a company of Miaf The Dun & Bradstreet Corporation

AUTOMOTIVE ELECTRONICS — TODAY AND TOMORROW

Shoji Jimbo Director and Member of the Board Toyota Motor Corporation

Shoji Jimbo is Director and a Member of the Board of Toyota Motor Corporation. His current executive duties have focused on drive train engineering, electric and electronics engineering, chassis engineering, micro electronics development, and vehicle testing and research. Previously, Mr. Jimbo was General Manager of Product Planning. In that capacity, he oversaw the development of models in the Mark II line (marketed as the Cressida outside Japan), and also supervised the early development work on the Lexus. These responsibilities took him overseas a number of times on visits to Toyota's overseas marketing facilities. Prior to this, Mr. Jimbo supervised a technical division at Toyota. Mr. Jimbo received a degree in mechanical engineering from Hokkaido University.

Dataquest Incorporated JAPANESE SEMICONDUCTOR INDUSTRY CONFERENCE April 12-13, 1990 Tokyo, Japan

1290 Bidder Park Drive, San Jose, CA 95131-2398 (408) 437-8000 Telex 171973 Fax (408) 437-0292 lBeB87oia.iMG oaw7/od!^ ULSI Era: Challenges and Opportunities

Automotive Electronics -- Today and Tomorrow Shoji Jimbo Director and Member of the Board Toyota Motor Corporation

1990 Dataquest Incorporated April 12 — Reproduction Prohibited ^yT(5)»ffl©TD¥ L ©MD TODAY AND TOMORROW

Shoji Jimbo Director

TOYOTA MOTOR CORPORATION '40 '50 '60 '65 70 '75 '80 '85 I I r I _l I i_ I- Radio / • Diode for alternator Breakerless ignition C Age of transistor } •Cruise control •Electronic instrument o Transistoransistor 'EFI cluster o MOS-transistor • EL illuminated meter

ABS z / • AT control •AT control ( Age of IC ) •EFI • Multiplex optic •^j •Ignition module •ABS / oICo MOS IC OCCD • MT c •CRT r -Spark control •Ha C Age of micrOCOn^;^ '.Trr/oS' '^ o 4-bit CPU Keyless entry •,. -,i o 8-bit CPU ol6-bltCPU Oil crisis Social Exhaust impact emissions U.S.A CK-o ( regulations JAPAN O—O O- EVOLUTION IN AUTOMOTIVE ELECTRO 00 CO CO CO (0 o 0) lO CO

CO 00 CO

00

O) h* O UJ f h* E N- s LO o O O o o o CO CD CM uoiionpojd |Bioi lo % J2 8 o Entertainment/Convenience •»• 7 0) Display/Communication 6 Chassis/Safety 0) Q. 5 Power Train X (D •o 4

-M 3 (0 o O 2 & m 1

'76 '80 '84 88 92

INCREMENT OF AUTOMOTIVE ELECT o CO OO oc OO o LU oo X

oo C/)

Csl [ oo o

< [ oo

< oo •••a _L ± ± I 1 1 (/) O o O O o o o o O o o o o CM O o CO CM rf CO (/) CQ sjlunoOOLx < '60 '70 '80 r 1 Power assist STEERING ( CONTROL 4 wheel steering 1

Heiglit I SUSPENSION Damping force 1 CONTROL

Spring rate ^ BRAKE Wheel lock ( CONTROL TRACTION Engine torque CONTROL cz

EVOLUTION OF CHASSIS ELECTRO '60 70 '80

Climate control Automatic L Micro processed L Power window l_ Automatic L

Power sheet IVIemory type I ~ Vibrating sheet L Tilt & telescopic steering Door lock Automatic C Keyless entr Remote controlled _J mirror IAutomati c Safety

EVOLUTION IN CONVENIENCE & SAF '70 '80 '90 20GG I r Convenience ( * ["^^^ ^? • Human f Level up di Safety Artificial in • Lean burn Fuel economy • Combusti pressure Emission regulation • Precison • Level up d

FUTURE AUTOMOTIVE ELECTRO Warning light Front sensor Air bag and (in steering w

EC

Front sensor Warning buzzer Floor sensor

AIR BAG SYSTEM Printed-on glass antenna

RF Distributor "^ Radio i Receiver Cancel s ca c 0) 1 (0 Door control Ignition k o relay position Oi c 3 i •—, Door locksolenoid Door cou o switch o J KEYLESS ENTRY CD-information disc Spea UHF W WW Antenna ^ ^ ^ amplifier CD player \_ with radio ^ R 0. & tape player TV tuner LC jm ^^QDD Ht^Stel 4- [=:t o Audio VTR adapter [o^ TV controller nnaa Amo I 00 Oi ODDO Door switcii 3 IR I D Audio inter jDDisplai y unit with CD-R decorder Compass sensor[ ^

DireclioFL switch Cruise Tail lanip Multi. Suspens Park brake swM display Speed sensor ECU { Tractio AT E Hands-free Climat telephone rt Engine Tranceiver MicroplionI e

TOYOTA ELECTRO MULTIVIS Direction sensor Speed sensor

>

DEAD RECKONING CRT touch switch

POWER BRIGHTNESS OFF I 1 m^'f^^^lKT VEHICLE AUDIO INFORM

Origin TV I MAP I •> MAP- input CH |RETURN| TINT COLOR

NAVIGATION SYSTEM '60 '70 '80 '90 POWER TRAIN Fuel injection Spark advance '3nr Engine Idle speed control <=^ INTEGRATED Knock control T] Transmte$fori Transmission V control : Cruise ± 1 control Traction control {CS CHASSIS

Brake ABS -iJl Power steering J Steering control TV Damping force ^ Suspension control Height control *"1

Active control [2 J

EVOLUTION OF SYSTEM INTEGRATI Fuel pump relay Air conditioner compressor Warning lamp ~n L~^ Fuel pump i (gCi— ii 5^ 1

Air flow ISC valve =^^" meter

Jifc Fuel pressure control Engine solenoid control ECU Ignition Injector •• Intake air module control solenoid Transmission shift Coolant temp, sensor position^ r switch

peed sensor

89 TOYOTA COMPUTER CONTROLLED SYSTEM Combustion pressure Inject sensorX yp -103 vo^^^ Crank angle Ignitio sensor Wide-range A/F sensor Engine -{DS ^ control r—

Accelerator pedal angle sensor

t^ Cruise control ECU |—• Engine •ff©^ Throttle Traction ECU control actuator AT & 4WD ECU y^ ECU 2 4WS ECU Suspention control ECO |—• ENGINE THROTTLE CONTROL SY Target vehicle Laser radar c^::>:((((((i- sensor

Laser Cruise eontrol Traction radar —( switch actuators ECU

Sensor & Throttle Throttle switch control control ECU signal actuator i^ Buzzer

LASER RADAR CRUISE CONTROL Wheel speed sensors Brake control actuators ©Q©0

Engine ECU-

INTEGRATED VEHICLE MANAGEMEN Diagnosis A.I. terminal monitor

ECUS

FUTURE DIAGNOSTIC CONCE '>))jnn \ ] • V yTaffic Information,

Police Traffic Traffic Control Center Information Teleterminal Processing Teleterminal S Center Co-operation Center (Posi

Direction Sensor Antennas :^ Tele Navigation terminal Receiver ECU Sign post|_ Receiver CD ROrM

Wheel Sensor

ADVANCED MOBILE TRAFFIC INFORMATION & COMMUNICATION S

^1)^-'..- ^ TRANSMIS

^ KNOCK CONTROL OTHER FUNCTIONS OTHER FUNCTIONS DIAGNOSTICS ECU EGR FUNCTIONS ISC - ESA EFI

ADC I/O NO. OF LSIs STAND-BY-RAM 8 BIT SINGLE CHIP CPU ROM ADC CPU. RAM 4 BIT SINGLE CHIP CPU 16 BIT SI

(1983 MODEQ ( 1985 MODEL) ( 1993

ECU TREND OF ENGINE CONT Engine control Transmisson control CONTROL ABS SYSTEM Traction control Air suspension control CRT display Electronic instrumental cluster ENTERTAINMENT Car audio Mobile telephone Climate control OTHERS TOTAL THE NUMBER OF LSI AND IC ON A LUXU DISCRETE

CUSTOM

iiST"

SEMICONDUCTORS USED IN AUTOMOB r Down sizing

Custom IC

Anti-Noise capability Low co

TREND OF AUTOMOTIVE SEMICONDUC 1. High reliability 2. Low cost •'t •^•'•r • 3. Highspeed : 4. Low power 5. Small size 6. Sufficient ROM/RAM 7. IVIuIti function 8. Quick turn around tim

REQUIREMENTS FOR AUTOMO SEMICONDUCTORS Manufacturer Today Fu TOYOTA (NIPPON DENSO) 8 Bit (Custom) 8 Bit, (FUJITU TEN) NISSAN 8 Bit (6301) (JECS) 16 Bit (8096) 8 Bit, (HITACHI) G M Q p!* (GMP4, (DELGO) O DII GMCM) 8 Bit, FORD 16 Bit (8061) (EED) 16 BOSCH 8 Bit (8051) 8 Bit,

CPU TREND FOR AUTOIVIOTIVE U Design rule I 5/A 3/£ 2jti )| 1.5/ 106

(0 N 0) o H16 > O O ft •D 068HCil M- 105 80960 68000 Q, O O HD6301V E 6809 Custom m 3 O 8051 6802 Q. for Toyot z O t 104 '75 '80 '85 '90 -> Year INTEGRATION TREND OF CPU HDh ik T 8bit CPU AV \ r^ A«,« DM/ VU ^ p ROM 1 V ^^ M' r ^ w -• SC]I ^ F RAM

4— .rl T / ^ -• Hi-Spee' Q Vo ^ W Parallel

%

<-| M ^ TirtKe i ^ W WDC

NEW PRODUCT CPU DIAGRA

-• J:4i*"

••%

ir - ,!l

«^i

«pe^ m^'-

•^m % 1 •5«. f'I'r, p; -.m .111^ ^ti" •.r.:Bi:;-::'1. % #.:?^V--s^^-v.-i: , •^•,

. !.• •t ^'f^:' h-'.-v••^TSf; - ti:':->''y^:"'i;irr''M< - ^ufcy^; .^-1^: . w, t'^1 ^te.s^#^^- -Ji," :•'-"•!•] V"^ ?:--d -4 IPIl mk^ 'S -..I TT ll>t II fV;J ^M 1^:11^

steady improvement with each de analysis in the market

Quality control in the upstream

(more approarch to the source)

STEPS TO ESTABLISH RELIABI ACS 6:2 ELflfi 289/298 iit l8tll88B1 OISP fi;: VSi n HODIBD C' =, 1 CI 5n Cr

(X 1) II-li

^ Q. Q. t

^ Yea ZERO PPM RELIABILITY IN THE M •'^

';^;-f

^'HMi DataQuest nn a company of 1MB The Dun & Bradstieet Corporatiol n

MULTIMEDIA IN THE 1990s

Richard A. Stauffer Marketing Manager Princeton Operation

Mr. Stauffer is the Marketing Manager for Intel's new Princeton Operation. He is responsible for marketing, application development, and customer support for Intel's evolving line of DVI Technology products. Mr. Stauffer joined intel with its acquisition of DVI Technology from GE where he had been managing GE's internal DVI Technology Venture. Mr. Stauffer was instrumental in developing the Initial DVI Technology business plan and in negotiating the transfer of the technology from GE to Intel for its continued development. Prior to his work at GE, Mr. Stauffer worked with the international management consulting firm of Booz Allen & Hamilton. Mr. Stauffer received a BSE from Princeton University and an MBA from Stanford University.

Dataquest Incorporated JAPANESE SEMICONDUCTOR INDUSTRY CONFERENCE April 12-13, 1990 Tokyo, Japan

1290 Ridder Park Drive, San Jose, CA 95131-2398 (408) 437-8000 Telex 171973 Fax (408) 437-0292 19116001 .IMG 03/SGraO-.^1 ULSI Era: Challenges and Opportunities

Multimedia in the 1990s Richard A. Stauffer Marketing Manager Princeton Operation Intel Corporation

1990 Dataquest Incorporated April 12 — Reproduction Prohibited Intel's Multimedia Roadmap

Presented by: Dave House Intel Corp. Presented at: Microsoft CD-ROM Conference February 28, 1990

Contact: Karen Andring (609) 936-7619 Paula Zimmerman (609) 936-7615 Intel's key multimedia strategies are to deliver. a clear, long-term vision of multimedia computing dramatic component performance growth state-of-the-art technical performance and video quality increasing integration and decreasing cost a flexible approach compatible with emerging industry standards a total product solution an array of authoring products targeted to a broad range of markets

We've included a glossary at the end for your reference.

-^.^

2/27/90 DVI™ Technology LdDDQg ir^orom V3S2(DIII DVI Technology Timeline V

Video Processors • Prototype 1 i750 PA/DA i7S0 PB/DB Thi VIUcv Video Video i750 Processors Processors Processors Vid

AlgoriUim s • PLV • PLV 1.0 PLV, RTV, PIC PLV, RTV, PIC CX3 1.0 1.5 & 2.0 MP 0^ Httildwtir e O • Prototype Pro750 Products For ActionMedia Products Pro (/) Systems Commercial Applications For Desktop Applications and a> Technology • New Alliances - Announced Intel Purchases DVI li.S. & International Broa • Broad Availability of Offic Loca IBM DVI Development Tools Joint Project Agreement Merg • First DVI Vide Application RtrikmH Olivetti European Announcement

1986-87 1988-89 1990-91 1992-

Intel's long term view of the future of multimedia computing - DVI Techn 2/27/90 on-going progression from an R&D project to commercial products. OVITM Technology lIoQ(£ir(eaisiiiij^ CcomipciDUKBiiafi lP(Biif(Dii^UBaiii( TS DVI Video Processors ao u 1000 a> c« 900 1750 u a> THIRD D. 800 GENERATION M 700 S2l o-a 600 500 a. O 400 c^ O 300 a o 200 100 i750A

1989-90 1991 1992 i750A 175011 1750 Third G

Multimedia computing requires tremendous amounts of processing power to m needs of motion video, audio, stills, graphics and text manipulation and display processors will provide greater power over time to meet these needs. Their pro offers full flexibility and backward compatibility.

2/27/90 DVITM Technology / N Md&iS^iii VM(gciD HDTV a Near HDTV O) TV Dl) C« s Near TV hH o a> VCR -O c Near VCR o

1989 1990 1991 1992 i750A 1750 A 1750 B 1750 Third Generation

DVI Technology currently uses two motion video compression algorith PLV (Production Level Video) and RTV (Real Time Video). Advancem in silicon technology and algorithm software will move image quality tow HDTV resolutions over time. 2/27/90 DVI™ Technology lIiiatt;Lr^(iii^ij]iai Uuategu^a DVI Delivery Produ ii» I- Fro750 •• mill 3-board set •• iiim. = •• mill r W — a> ssiinis '-t_i 1—1

PH Act ion Media •• iiiiii Scries 1 p^itf ^11111= single board _-—- •—^—i 1 cd pfi >^ AclionMcdia CQ Series 2 ••lllii single board

r 1

Q DVI Integrated on iiiiiiss Motherboard =11111I ..J = 1990 1992 1994 1996 1998

^ Since its first product introduction in 1989, Intel has been working to red and footprint of DVI products, on-a-path from multiple boards to a fraction

2/27/90 DVI™ Technology $1.00 /^ HuKgirceii^Qiiig KintegirMiicD Lower Cost Per MOPS .80

C/5 Achieving Nickel/MOP PH .60 before 1995 o u O) .40 a C/3 o U .20

0 1990 1991 1992 1993 1994

The cost of MOPS (millions of operations per second) will continue to decr next five years, making multimedia more powerful and more cost effect generation. 2/27/90 ^' 1.,

DV|TM Technology im^ iP€ Multimedia at Low Cost ,,, Esaiucfia Estiiaated Pierc

- 386™* microprocessor -DOS/Windows -40MbHD/noppydisk - VGA + Monitor $2500 -CD-ROM drive - Audio - 2 Mb system RAM - Mouse/Keyboard

(Cost of adding DVI Technology to motherboard)

-17506 video processor -1 Mb video memory } $500*

At $500 or less iim fxp^t DVI Technology to become a standard feature of ev by 1992.

2/27/90 *Sub$300kyt99S, OVITM Technology IFIl(issQ[bll(B IPirqDg]riimm3i[bIl(i ApipirdDiKell} Evolving Algorithm Alternatives <

CA i750 Third fl Generation ^ oc Sa o i750B PLV 2.0 U RTV 2.0 PIC 2.0

i?50 A PLV 1.0 • PLV 1.5 RTV 1.0 RTV 1.5 PIC 1.0 PIC 15

1989 1990 1991 Because DVI utilizes a programmable video processor, it runs a wide r compression algorithms and allows compatibility with future standards 2/27/90 evolve. DVI™ Technology T^lull IPDr(ID(flllQ(£ll ScDilQallS®!! DVI Technology

Intel's is a/o/a/solution. DVI components, compression techniques, hardwa software evolve in parallel making a wide array of exciting applications possible.

2197/90 DVI™ Technology

Am Awmj ODIF AnalltiaDirSiiig IPirdDolliiiKetb for DVI Application Development "Hyp

SABER - Pinnacle Courseware

MEDIAscript* - Network Technology

Authology*: MultiMedia - CEIT Syst

LUMENA*-Time Arts (paint)

Pro750 C-based authoring/ ActionMedia

Pro750 capture software/ ActionMedia

1989 1990 1991 Easy-to-use, advanced authoring tools are making their appearance and offer user-friendly, flexible options for creating applications. 2/27/90 NJ NJ vO O

n

Cft DVI™ Technology

Glossary of Terms

ActionMedia ™ - DVI Technology's product family, introduced in 1990 and consisting of single-board dehvery and single board capture capability for AT or Micro Channel™ architecture buses (introduced with IBM*), e.g. ActionMedia 750 ADP, etc.

CCITT - Consultative Committee of Telegraph and Telephone. International standards committee whose charter is to generate standards for netwoiking and telephony.

DCT - Discrete Cosine Transform - algorithm or mathematical technique which can be used as part of a compression algorithm. 1*750 ™ - DVI Technology programmable video processor family. ISO - International Standards Organization. The worldwide group responsible for establishing and managing the standards committees and expert groups, including those working on intemational compression standards.

JPEG - Joint Photographic Experts Group. A group within ISO whose charter is to generate standards for full color still image compression.

MIPS - Millions of instructions per second. Refers to a processor's perfor­ mance.

MOPS - Millions of operations per second. Refers to a processor's performance. In the case of DVI Technology, more MOPS translates to better video quality. Intel's video processor can perform multiple video operations per instruction, dius the MOPS rating is usually greater than the MIPS rating. 2/27/90 12 DVI™ Technology

MPEG - Motion Picture Experts Group. A group within ISO whose charter is to generate standards for motion picture video compression from digital storage media.

Px64 - informal name for CCITT video algorithm (formal name is H.261) proposed by CCITT for teleconferencing applications. Targeted for ratification as a standard for telephony within the next year. PIC - Picture Image Compression - DVI Techology's on-line still image compression algorithm. PLV - Production Level Video - DVI Technology's highest quality motion video compression algorithm. Compression is "off-line", i.e. non-real time, and playback (decompression) is real time. Independent of the technology in use, off­ line compression will produce a better image quality than real time since more time and processing power is used per frame.

Pro750 - A DVI product family introduced in 1989 consisting of Application Development Platform, boards and software. RTV - Real Time Video - DVI Technology's on-line, symmetrical, 30 frames per second motion video compression algorithm.

DVI, 1750, 386 and ActionMedia are trademarks of Intel Corp. Intel is a registered trademark of Intel Corp. Micro Chaimel is a trademark of International Business Machines Corp. IBM is a registered trademark of International Business Machines Corp. MEDIAscript is a registered trademark of Network Technology Inc. Authology is a registered trademark of CEIT Systems Inc. LUM^A is a registered trademark of Time Arts Inc. 2/27/90 13