Lapas attēli
PDF
ePub
[blocks in formation]
[blocks in formation]

• Construction of new atomic reactor Fifth International Fuzzy Systems Asso

Remark: Of the HAN projects mentioned, the one most directly linked to Kahaner's background is the Intelligent Computer Project. The project's organizers see this as related to many other projects, such as CIM, materials, B-ISDN, HDTV, DRAMs, human interface, multiprocessors, and electric car. Major technology required will be flatpanel display, wireless communication, high speed networks, large scale databases, Korean language processing, object-oriented technology, applicationspecific integrated circuits (ASICs), virtual reality, etc. The intelligent computer planning team's road map for this project contains the banner "One Computer Per Person" and divides the major subprojects into three parts, associated with neural computer, knowledge processing, and multi media. Other subprojects include a voice typewriter and a personal assistant server. As most aspects of this project are still in the planning stage, it is too early to make an assessment; however, it looks like an excellent opportunity for enhanced international cooperation.

Computer researchers should make note of the following meetings to be held in Korea.

15-18 September 1992, Seoul, Korea

Second Pacific Rim International Conference on Artificial Intelligence

For information:

PRICAI Conference Secretariat
Center for AI Research
KAIST

373-1 Kusong-dong
Yusung-gu, Taejon 305-701, Korea
Tel: +82-42-829-2711, -2712

Fax: +82-42-829-8700

E-mail: pricai92@cair.kaist.ac.kr

ciation World Congress

For information:

Prof. Z. Bien

Department of Electrical Engineering

KAIST

373-1 Kusong-dong

the largess of the Korean Government but also to the connections Korean scientists have with the United States, which makes them familiar with Western software products. Application fields are represented quite uniformly, with physics, electrical engineering, and chemistry each covering about 20% of the users. During the machine's first year in operation it was free to users. Currently a charging algorithm is in place. The scientists at Seoul National

Yusung-gu, Taejon 305-701, Korea University are not heavily on this

Tel: +82-42-829-3419
Fax: +82-42-829-3410
E-mail: zbien@kumgang.kaist.ac.kr

VISITS OF THE ONRASIAAROFE TEAM

Systems Engineering Research Institute (SERI)

Keeping Korean science organization names straight is difficult, especially so as some of their affiliations have changed over time. SERI began as have changed over time. SERI began as the Computation Research Laboratory in the Korea Institute of Science and Technology (KIST), then became affiliated with the Korean Advanced Institute of Science and Technology (KAIST), and in 1989 moved back (organizationally) to KIST. It is located in the Taeduk Science Town, about 90 minutes by train from Seoul. Taeduk is the sight of several Korean science agencies as well as industrial laboratories. It differs from Japan's Tsukuba in that it was already a "real" small city when the scientists began moving in. At the moment a large EXPO is being constructed that will focus attention on Korean science and Taeduk in particular. SERI has 472 staff, 22 with Ph.D. degrees.

The main interest here is the Cray-2, which was discussed in an earlier report. which was discussed in an earlier report. Its Cray-2 was installed almost 4 years ago and was the first supercomputer in Korea. It has a very rich application software base; this is due not only to

machine because of the high cost and also because until now a high-speed link was not available.

In addition to the supercomputer center, SERI also has divisions of Artificial Intelligence, System and Application Software, and Software Engineering. Very few of the actual research activities were shown, although some that looked interesting to Kahaner were automatic license plate recognition, automatic personal identification via image analysis of fingerprints, and remote sensing/image processing work. The facilities at SERI are among the very best in Korea; essentially any kind of advanced workstation-type equipment is installed. The Cray is heavily used but not completely flooded. Parallel computing is centered around a transputer-based system. To the best of Kahaner's knowledge there is no commercial parallel system available, and this could be a serious weakness for anyone wanting to do work on applications, but might be a challenge for systems researchers.

SERI is the central information center for a collection of databases and on-line services jointly called KRISTAL (Korea Research Information on Science and Technology Access Line). Databases include the Union catalog for monographs, bibliographic information compiled by specialized information centers, scientific equipment, software registration, scientific specialists, as well as other specialized databases for specific scientific disciplines.

Information moves to/from industry, government agencies, libraries, etc. and is available via public networks and a new backbone network, KREONET (Korea Research Environment Open Network). KREONET includes a 56-kbps line to the United States through the San Diego Supercomputer Center, a T1 line to SERI's Seoul facility, and a number of 56-kbps lines from SERI in Taeduk to other Korean cities including Pohang, Kwangju, Pusan, Chanwon, etc. KREONET provides the usual network services such as file transfer protocol (FTP), mail, rlogin, etc.

Korea Advanced Institute of
Science and Technology
(KAIST)

KAIST is a research oriented graduate school with 1,000 Ph.D. students (70% full time) administratively supported under MOST (all other federal universities are administered by the Ministry of Education). Associated with KAIST is the Korea Institute of Technology (KIT), which is an undergraduate school with 3,000 students. Since KAIST's inception in 1971 it has awarded over 1,000 Ph.D. and 6,000 M.S. degrees. A new KAIST campus is mostly complete in Taeduk (the analogy with Tsukuba University is very strong). KAIST has departments in virtually all standard fields of science and engineering as well as in management science, humanities, and social sciences. As a federal university, KAIST and KIT provide various kinds of financial support; in practice it seems that most full time students are fully supported. In the area of computing, there is a large Computer Science Department with more than 100 Ph.D. and 100 master's students that has already conferred more than 40 Ph.D. degrees.

The mainstay of the research activities in computing appears to be in the Center for Artificial Intelligence Research (CAIR), which is subdivided into six laboratories (Database and

Knowledge Engineering, Language Engineering, Visual Information ProEngineering, Visual Information Processing, Speech Information Processing, Computer Systems, and Human Computer Interaction). There are plans to install a 64-transputer parallel system, but the research projects Kahaner tem, but the research projects Kahaner read about are all nonnumerical. The facilities appeared to be excellent.

In the physics area, one laboratory is conducting research in subpicosec

ond transient behavior of semiconducond transient behavior of semiconductor optical devices. Samples are prepared by the Electronics and Telecompared by the Electronics and Telecommunications Research Institute and measurements are made at KAIST, using lasers with pulse widths as short as 60 fs.

Korea Research Institute of Standards and Science (KRISS)

This was recently renamed from the Korea Standards Research Institute (KSRI) by incorporating the Korea Basic Science Center (KBSC) and the Institute of Space Science and Astronomy (ISSA). KBSC's main function was to provide research facilities and information on basic sciences to universities in Korea. ISSA was responsible for optical and radio observation and research in positional astronomy, astrophysics, and space science.

KRISS is essentially a standards organization that also conducts some R&D in precision measurement technology and supports industry with training in measurement technologies as well as development and repair of highprecision instruments. It has divisions precision instruments. It has divisions of Mechanical Metrology, Electromagnetic Metrology, Quantum Metrology, Analytical Chemistry, Precision Instrumentation, and a Materials Evaluation Center. In the physics area, research is carried out on the quantum Hall effect, direct current superconducting quantum interference devices (DC SQUIDS), laser spectroscopy, and other solid-state phenomena of interest for standards.

The staff of almost 500 (about 100 with Ph.D. degrees) is located on a campus setting in the Taeduk Science Town. Most of KRISS' scientists have degrees in physics or engineering and there are no Ph.D. holders in computer science among the staff. KRISS' 1992 budget is about $30M. As with other international standards organizations, KRISS has a large number of international cooperative agreements and joint research projects with similar organizations in the United States, Japan, the United Kingdom, Italy, Canada, Germany, the Commonwealth of Independent States, and Australia.

Electronics and Telecommunications Research Institute (ETRI)

Also located in Taeduk Science Town, ETRI is the R&D arm of the Ministry of Science and Technology (MST), focusing on information technology, telecommunications, computers, automation, and semiconductors. It has a staff of about 1,700 of which about 1,400 are technical and/or professional. Only about 5% of ETRI's work is viewed as basic science (6070 man-years per year), although the plan is to grow this to about 15% during the next decade. The current budget is about $130M, which is planned to grow to $175M by 2001. By that time ETRI hopes to have a total staff of about 2,350, of which 30% will be Ph.D. holders.

Most of the R&D work is clearly oriented toward telecommunications such as digital switches, ISDN, transmission technology, and networks. Semiconductor design and development such as ultra large scale integration (ULSI), HDTV, etc. is next. Other computer work is much more narrowly based and supports the HANP Intelligent Computer Project mentioned above.

ETRI has a small office in San Jose to support the dozen or more staff that are in the United States at any given

time and also to act as a base for obtaining information about advanced technology. This year they will open a liaison office in Europe to support information exchange, joint research, etc.

In electronics research, ETRI has developed a 4-Mbit DRAM memory using 0.8-micron technology. Work is underway currently to develop 16- and 64-Mbit DRAMs using 0.5-micron technology. In applied optoelectronics research, ETRI has developed a prototype 155-Mbps optical transmitter and receiver using InGaAs/InP materials at a 1.3-micron wavelength. Current work is directed toward 565-Mbps optical transmission for use in cable television networks. More advanced systems envisioned will use InGaAs/ InP positive-intrinsic-negative field effect transistor (PIN-FET) optoelectronic integrated circuits, fiber amplifiers, and distributed feedback lasers for coherent transmission at 2 or 10 Gbps. Other notable research includes the development of quantum devices, based on heterostructure materials grown by organometallic chemical vapor epitaxy and liquid phase epitaxy. These heterostructural materials and devices are the subject of the subpicosecond optical response research conducted jointly with KAIST (see above). Semiconductor device research is directed toward emitter-controlled logic (ECL) and complementary metal oxide semiconductor (CMOS) arrays. Device delay times in ECL devices as low as 51 ps have been achieved using 1.0-micron bipolar process technology.

The President of ETRI, Dr. Kyong, emphasized his interest in international collaboration (see Summary).

Korea Science and Engineering Foundation (KŌSEF)

KOSEF is essentially equivalent to the U.S. National Science Foundation. A brief conversation with KOSEF's Director, Dr. Kang, revealed that, unfortunately, there are very few programs to support visiting Western researchers to Korea.

Pohang Institute of Science and Technology (POSTECH)

POSTECH is Korea's newest private university, established by the Pohang Iron and Steel Company (POSCO) in the city of the same name, slightly more than 200 miles southeast of Seoul. Its first entering class was 1987. POSTECH was discussed at length in an earlier report, so that information will not be repeated here except to say that the goal is to make POSTECH Korea's equivalent of Cal Tech. Even after just a few years there is now agreement that it is one of three most important science universities in Korea (KAIST and Seoul National University being the other two), although there is friendly disagreement about its ranking between the three.

Dr. Kim, the President of POSTECH, explained that they are very satisfied with the level of students they have enrolled. By 1995, when the university will be operating at its full capacity, there will be about 1,200 undergraduates, 1,000 graduate students, and 300 faculty. Support from POSCO is generous, but Kim is hoping to reduce the university's dependence on the company as their endowment builds up. He would also like to get additional government support that is now going to other national (nonprivate) universities.

On the topic of visitors, Kim explained that international collaboration is considered very important and

that more than a dozen Nobel Prize winners have already lectured at POSTECH. There are a few Western visiting professors who are able to work in English, but lecturing in Korean is an unofficial requirement for any permanent position. POSTECH provides an attractive research situation for visitors, but there might be some lifestyle difficulties. Pohang is definitely not Seoul and access to Seoul is only fair. While English is widely spoken at POSTECH, it is unlikely to be understood outside in the city of Pohang. Further, there are no schooling facilities in the area for non-Korean speaking children.

One of the most important projects at POSTECH is the construction of a 2-GeV synchrotron-radiation source, the first in Korea. POSTECH serves as the host organization for the Pohang Light Source (PLS), Korea's entry into the synchrotron-radiation research arena. Starting from a planning task force in 1988, PLS will be a 2-GeV electron storage ring, similar in design to the Advanced Light Source currently under construction at Berkeley, CA. The injection system is very different, however. The PLS injector will be a full-energy linear accelerator (linac) designed by the High Energy Laboratory of Beijing, China. The initial section of the linac was built in Beijing and installed at Pohang by the Chinese. Subsequent sections are being built in Korea to the Chinese design, which is closely related to the design of the 2-milelong Stanford Linear Accelerator but reduced to 2 GeV instead of 50 GeV. When finished, PLS will serve as a "user facility" for all qualified Korean scientists. Of the capability for 12 beam lines, only two are currently funded, and one of those will be for machine monitoring. Hence, the ambitious plans for research in materials science, surface science, molecular biology, chemical engineering, atomic and molecular

physics, x-ray lithography, etc. will develop later as additional beam lines are added.

Research Institute of Industrial Science and Technology (RIST)

RIST is the research arm of Pohang Iron and Steel Company. It is located in Pohang within the complex of the POSTECH campus. Nearly 900 staff are supported by a budget of about $80M; there are also about 100 POSTECH faculty that have a research appointment at RIST, and many of their students are working on research projects there. Approximately half of RIST's work is concerned with steelmaking, although Kahaner did meet two students working on Bezier curve and related computer graphics applications. In the materials area, large programs are in progress in the area of advanced materials including metal matrix composites, polymers, and semiconducting materials and devices.

Facilities look very good, and the team was told that RIST is very interested in international cooperation, especially in the areas of optical fibers and infrared (IR) sensors. However, potential visitors will want to clarify to what extent they will have access to potentially proprietary information and technology.

According to Kahaner, the visit to RIST was too brief to get an accurate reading about the place. However, he came away feeling that the organization did not have its own clear vision of where it wanted to go and depended on POSCO for spirit and direction as well as for funding. By other names, POSCO has had an affiliated laboratory since 1977; it is only very recently that the organization has moved close to a university, so it is probably natural that it will take time for it to evolve into a real research institution. RIST's goal to become a world-class hi-tech research center by the year 2000 will mean that

it will have to establish its own identity, independent of faculty advisors from POSTECH.

Korea Institute of Science and Technology (KIST)

KIST is Korea's first science research institute, formed in 1966. It was merged with KAIST (essentially a science and technology university) during the 1980s but was separated again in 1989. Its main facility is in Seoul, although there are two affiliated institutes, SERI (see above) and the Genetic Engineering Research Institute (GERI), both located in Taeduk Science Town. Total Seoul staff is almost 900, with more than 500 researchers. KIST's total budget last year was about $260M, of which almost 80% is from the Korean Government.

In Seoul there are divisions of Fine Chemicals, Process Technology, Mechano-Electronics, Metals, Ceramics, Polymers, and Semiconductors. The ceramics laboratories were well equipped and were engaged in research on ceramics for turbine engines, composites, sol gel processes for ceramics, etc. There is a doping control center (used to test athletes during the 1987 Seoul Olympic Games), an optics and electronics technology center, an environment research center (focused on pollution and waste control), and a fluorocarbons alternatives center. Kahaner saw no significant work relating to computing except to assist in physical laboratory experiments. The link to the SERI's Cray is new and is link to the SERI's Cray is new and is not used much. Kahaner didn't get a sense that much modeling was going on, but admittedly the visit was very brief. There is a remote sensing and image processing group, but he did not get to see any of their work.

International cooperation is an important ingredient of KIST's programs and, in fact, there is a KoreaSoviet Scientific and Technical Cooperation Center (KSSTCC), established in 1991.

Rehn observed that KIST's Semiconductor Materials Laboratory is one of the most advanced electronic materials research groups in Korea. Among the key research activities are:

• Growth of high quality bulk GaAs crystals by the vertical-gradient freeze method, as well as other, more standard methods.

• Heteroepitaxial growth of III-V compounds by molecular beam epitaxy, organometallic chemical vapor epitaxy, and hydride vapor phase epitaxy.

• Deposition of low resistivity W and W2N layers by plasma-enhanced chemical vapor deposition.

• Recrystallization of Si-on-insulator structures with application to threedimensional semiconductor devices.

• Electrical and optical characterization: Hall effect, x-ray analysis, C-V, deep level transient spectroscopy (DLTS), PICTS, photoluminescence (PL), photoluminescence excitation (PLE), and other techniques.

Device fabrication and characterization: double heterostructure laser diode, two-dimensional electron gas devices, resonant tunneling diode, and multiple quantum well devices.

Seoul National University (SNU)

SNU is the largest, oldest, and one of the most prestigious institutions in Korea. The team's visit to SNU was only at night, to have dinner with their host, Dr. Sang Joo Kim, Vice President. Kim commented that there is competition between SNU, KAIST, and POSTECH for government support. He admitted that recently POSTECH had been very successful, but he felt

« iepriekšējāTurpināt »