rick@cs.arizona.edu (Rick Schlichting) (02/16/91)
[Dr. David Kahaner is a numerical analyst visiting Japan for two-years
under the auspices of the Office of Naval Research-Asia (ONR/Asia).
The following is the professional opinion of David Kahaner and in no
way has the blessing of the US Government or any agency of it. All
information is dated and of limited life time. This disclaimer should
be noted on ANY attribution.]
[Copies of previous reports written by Kahaner can be obtained from
host cs.arizona.edu using anonymous FTP.]
To: Distribution
From: David Kahaner ONR Asia [kahaner@xroads.cc.u-tokyo.ac.jp]
Re: Timely announcements
15 Feb 1991
What follows is a potpourri of recent items that were of interest to me,
sometimes because of their relation to a report I did in the past, or
because they provide a glimpse of the future. I will be following up on
a few of these in more detail but some readers might find these previews
useful. I will do more of this when warranted. I apologize for the random
nature of these short notes; none are long enough to justify a separate
report; a one-line summary of each is given below.
MITI REORGANIZATION OF NATIONAL RESEARCH INSTITUTES.
MITI PROGRAMS--IPS, LANGUAGE TRANSLATION, LEARNING.
NEW SOFTWARE STRUCTURING PROGRAM.
MAJOR NEW RESEARCH PARK.
SPACE ROBOTS R&D JAPAN, IN NEW TECHNOLOGY JAPAN.
JAPANESE MICROCOMPUTERS-WHO IS AHEAD, US OR JAPAN?
JAPANESE COMPUTER SCIENCE LABS OUTSIDE JAPAN.
SUPERCOMPUTERS.
SALES--LCD, WORKSTATIONS.
FUZZY.
MONBUSHO (Ministry of Education, Science and Culture) UPDATE AND NEW PROGRAMS.
-----------------------------------------------------------------------
MITI REORGANIZATION OF NATIONAL RESEARCH INSTITUTES
MITI has completed a plan to totally reorganize the national institutes under
its jurisdiction. The new plan will focus on
(1) international joint research centers for frontier technologies, and
(2) changing the focus from application to basic research by consolidating
and/or abolishing the research institutes in Tsukuba Science City.
The reorganization is based on Japan's reflection that, whereas Japan has
grown up to be an advanced high-tech nation, her strength is in
manufacturing technologies and she has not made significant contributions
to the world in basic technologies. One specific example of this change
is at the Tsukuba Mechanical Engineering Lab (MEL) which is administered
under MITI's AIST (Agency for Industrial Science and Technology). MEL is
the core research institute in Japan on mechanical research. It has been
involved in the development of electronics and new materials and is now
focusing on the development of "wise and strong" technologies, such as
robots and artificial bones, which can adjust themselves according to
environmental changes. MEL once emphasized developing technologies which
could be used immediately at the factories (in 1959 MEL developed a
numerical control manufacturing machine with one micron precision).
However, it is now officially going to terminate its role in the applied
research fields because of recent rapid technological advancements, and
will transfer its emphasis to basic research which will provide seeds for
creating new industries in the 21st century.
-----------------------------------------------------------------------
MITI PROGRAMS--IPS, LANGUAGE TRANSLATION, LEARNING.
MITI's Intelligent Production System (IPS) (10 year project, $1.1Bill US) is
going to have participation from DEC, Rockwell, United Tech, Boeing, GM, IBM,
Kodak, TI. Feasibility studies start now.
MITI will expand its multi-lang machine translation system development
project to include European languages. This is a six year project
($150mil) begun in '87 with participation of Chinese, Thai, Malaysian,
Indonesian governments. Uses an intermediate language. The project will
hold its first seminar with US and EC govts in July. By '95 electronic
dictionary and prototype is to be developed. Translation centers to be
established in Asian countries in '93-'95.
Basic Technology Center (MITI) will conduct research on network-based
learning, and develop a system for providing interactive learning
environments to network users. Some new companies will be funded to do
this (about $30million US).
-----------------------------------------------------------------------
NEW SOFTWARE STRUCTURING PROGRAM.
MITI has upgraded the New Software Structure Model to the status of an
international research project and set up two special research groups in
the Japanese Information Processing Association which began activities in
Oct 1990. The purpose of this project is to develop a system in which
software will be flexible enough to deal with errors and other changing
external conditions. The project is an industry-university cooperative
program. Overseas participants include Prof Stanley Peters (Stanford) and
Prof Joseph Gogian (Oxford). Remark: I do not yet know the relation of
this program to the emerging NIPT program (6th generation).
-----------------------------------------------------------------------
MAJOR NEW RESEARCH PARK.
Hitachi, MITI, Tokyo Univ and others will construct a video research park
in Tokide City. About 20 companies and various government agencies will
invest $15million US in feasibility studies beginning in 1991. The
University will move into the park this fall and more than 300 public and
private research organizations will also have labs there. Construction
costs are estimated at $1.14billion US.
-----------------------------------------------------------------------
SPACE ROBOTS R&D JAPAN, IN NEW TECHNOLOGY JAPAN.
Volume 18 No 11 of New Technology Japan, published by the Japan External
Trade Organization (JETRO), Machinery and Technology Department, 2-5
Toranomon 2-chome, Minato-ku, Tokyo 105, Japan, +81 3 3582-5579 contains
a very complete (16 page) survey of space robot research in Japan. It
describes programs in place and in planning, specific activities
organized by research organization such as ETL, MEL, Inst of Space and
Astron Sci, Nat Aerospace Lab, etc., and also R&D at private companies
such as Kawasaki, Shimizu, Mitsubishi, NEC, etc.
-----------------------------------------------------------------------
JAPANESE MICROCOMPUTERS--WHO IS AHEAD, US OR JAPAN?
TechnoJapan, Vol 23 Nos8-9, Aug and Sept 1990, contains two lengthy
articles surveying microcomputers that were designed and built entirely
in Japan. It describes in detail products from NEC, Hitachi, TRON, etc.
The article is a summary of about ten papers that appeared in a special
issue of the Information Processing Society of Japan's "Journal of
Information Processing".
One of the most interesting remarks in the article is their claim that
Japan has not caught up with the US in this technology. The only area in
which the authors claim that Japan is on a par with the US is in DRAM
(dynamic random access memory). They state the following. "Given then
that the Japanese industry is leading in the manufacture of DRAMs, one
should note that the most important technology in the field is not that
of DRAMs but rather of MPUs (microprocessors). The world market of MPUs
is now dominated by Motorola and Intel, and all Japanese computer
manufactures employ these chips in their machines. Furthermore, US
manufactures have amassed a considerable amount of MPU-related software
upon which Japanese manufactures rely heavily. In addition, licence fees
must be paid to US firms by Japanese companies manufacturing
semiconductor products, which also involves a considerable amount of
money. Experts predict that it will take at least another ten years for
Japan to catch up with American MPU technology." TechoJapan is published
monthly by Fuji Technology Press Ltd., 7F Daini Bunsei Bldg, 11-7
Toranomon 1-chome, Minato-ku, Tokyo 105 Japan. Fax +81-3-3592-0648.
-----------------------------------------------------------------------
JAPANESE COMPUTER SCIENCE LABS OUTSIDE JAPAN.
Matsushita is going to set up an Information Technology Lab in Princeton
specializing in advanced computer graphics. (NEC already has such a
facility in Princeton.) Startup will be 20 researchers increasing to 60
by 3/93.
Mitsubishi will set up computer R&D centers in US and Europe for studying
new computer architectures as well as information processing by human
beings. The first lab will be in Boston, with about 20 researchers,
mostly Americans, increasing to 50 in five years. The focus will be on
the development of super-parallel computers (10,000 processors).
Sony will beef up its Unix development operations in the US. Currently in
Japan Sony has about 170 engineers on the News Workstation, 30 working on
system software. In the US Sony has about 35 engineers, mostly working on
software. Sony claims it is difficult to hire Unix engineers in Japan and
so will strengthen Unix development in the US where is is easier to find
skilled Unix engineers.
-----------------------------------------------------------------------
SUPERCOMPUTERS.
MITI plans set up a supercomputer center in the Tokyo metro area. Access
will be free to researchers (even foreign). Funding is from New Energy
Development Org, regional governments, and private companies ($30
Million US). If the feasibility study in JFY91 (Japanese Fiscal Year)
indicates need for the proposed center it will go forward in JFY92. The
center would be part of a proposal for joint US-Japan R&D that would join
this center to supercomputer networks in the US. At this time it is not
clear what kind of computer would be purchased, what services would be
offered, or how the large Japanese computer companies would be involved.
Remark: Sounds to me like a good place to put Japan's best supercomputers
so as to make them accessible to Western researchers via network.
NEC's largest mainframe below SX/3 will be shipped by late '91. CPU speed
will be 2.8 times, and 100% more memory than current (ACOS3800). I assume
that NEC's claim to be using SX/3 technology is the big change here.
Hydrodynamics Lab will build a supercomputer facility in Osaka housing 4
supercomputers by June '92. Total budget is about $75million US. They
already have a Tokyo facility with 4 machines.
-----------------------------------------------------------------------
SALES--LCD, WORKSTATIONS.
LCD sales are booming. Sharp, Sanyo, Matsushita, Toshiba and others are
projecting very large sales boosts. Color LC TV and projectors are
expected to reach 1.8million and 150,000 units resp. LCD displays are
also being developed by Alps Electric, NEC, Mitsubishi, Asahi Glass, etc.
Remark: This is a key technology that will play an important role in
future computers. The best (largest and brightest) LCDs are produced in
Japan. Also, consumer electronics technology feeds into computers and
firms known in the West for their consumer electronics (Sony, Sharp,
Seiko, etc.) are also heavily into computer activities.
Seiko will set up an LCD R&D center in Italy this year with 20-30
researchers. Seiko can already produce 10inch PC LCDs. Omron also has a
new LCD with 320x160 dots, a contrast ratio of 8:1 and view angles of 45
degrees in all directions. They are planning to produce an 8-color
panel. Matsuzaki Vacuum will build a $75million US plant for liquid
glass substrates. Citizen Watch will build a LCD plant at $75million in
1991. The plant will eventually be able to produce 100,000 ten-inch
panels per month.
Japan minicomputer shipments in 1989 were 53% greater than '88 (48,000
units), and engineering workstation station sales were 87% greater. In
1990 engineering workstation sales jumped 53% to about 98,000 units. Of
these Sun has about 25%, HP 23%, Sony 15%, NEC 10%. X-window terminals
are estimated at about 5,000 units. In 1991 another 30% increase is
projected, with X-terminals doubling to about 10,000 units.
Toshiba has revised upward by 100% its sales goals for laptop
workstations. The new LCD production plant allows them to produce at the
Fuchu plant about 1000 units/month.
Fujitsu will completely remodel some of its workstation series to be
compatible with Sun.
-----------------------------------------------------------------------
FUZZY.
This week the 3rd International Congress of Biomedical Fuzzy Systems was
held in Tokyo. (A more complete report will be forthcoming.) Invited
speakers included Elie Sanchez (Marseille), George Klir (Binghamton), and
Lotfi Zadeh (Berkeley). The conference coincided with the appearance of
the Japanese popular science magazine Quark (March '91) describing
research in fuzzy logic including some projects such as dental
applications, and the Sendai subway, showing more than 20 fuzzy consumer
products, lists of textbooks, and questions and answers of general
interest such as the relation of fuzzy to neural. On TV at 10PM there was
a one hour variety show on a commercial channel about fuzzy, with
humorous shorts, live audience participation, a quiz show format, music,
and interviews with Zadeh at Berkeley as well as some Japanese
researchers. Very entertaining.
Sharp will license Electrolux fuzzy control technology and consumer ICs.
Will work on techniques for controlling vibration and noise in large
electric appliances.
-----------------------------------------------------------------------
MONBUSHO (Ministry of Education, Science and Culture) Update.
In 1990 MONBUSHO gave subsidies to private and public universities and
colleges of 130billion Yen, and another 56billion Yen as Grants-in-aid
for scientific research. Total MONBUSHO budget is about 900billion Yen
($6.9billion US). In the area of computers one specific program is of
interest.
Large scale computer centers used as inter-university research
facilities, and general information processing centers used as intra-
university shared educational and research facilities, have been engaged
not only in scientific and engineering computation and data processing
for research and education, etc., but also in developing databases and
providing scientific information by using databases on campus. They serve
as "nodes" in the Scientific Information System. (The National Center for
Science Information System is the center of MONBUSHO's science
information. It provides planning, coordination, scientific information
research, collection services, reference, database, and education
services.) Already seven large-scale computer centers, ten general
information processing centers and 43 information processing centers have
been made available for these purposes. In 1990 MONBUSHO will created
general information processing centers at Kanazawa and Kumamoto
Universities, and information processing centers at seven other
universities, Aichi, Shiga, Nara, Kobe, Kagawa, Fukuoka, and Oita.
Japan Society for Promotion of Science--new programs (6.1billion Yen).
* Expand "Fellowships for Japanese Junior Scientists"
* Upgrade "Post-doctoral Fellowships for research abroad for Japanese
Junior scientists."
* Upgrade "Post-doctoral Fellowships for Foreign Researchers" (mostly
young researchers).
* Begin commissioning Japanese language training courses for foreigners
conducting long-term research at Japanese universities.
* Upgrade the Liaison Center for Research in Overseas Areas, designed to
acquire information about foreign scientific research trends.
* Begin an international industry-academia cooperative program to make
available research information outside of Japan.
* Expand exchange and other programs with ASEAN nations.
International Cooperative Programs.
* Regional environment in the Arctic (new)
* Radioisotope beam science (new)
* Geosphere/Biosphere program
* Chinese Univ of Sci and Tech / U Tokyo Engineering
* Experiments in elementary particle physics CERN / U Tokyo
* World climate changes
* Nuclear fusion research
* Immunology research
* Ocean drilling
* Ground-air interaction in Hei-he valley (China)
* Global environment and other UNESCO science-related programs.
* 206million Yen to send Japanese researchers to international scientific
research meetings abroad.
* 131million Yen to hold international symposiums participated in by
first-rate researchers from abroad.
* 43million Yen to hold specific international conferences, such as
International Congress of Mathematicians.
* 2billion Yen for US/Japan research in nuclear fusion, solar energy
conversion, high energy physics.
* 67million Yen for US/Japan basic research in space, neutron scattering,
test animal science, DNA splicing.
-----------------END OF REPORT-------------------------------------------