rick@cs.arizona.edu (Rick Schlichting) (02/03/91)
--------------------------------------------------------------------------
Trip Report -- Fujitsu Numazu Complex (April 25, 1990)
Richard D. Schlichting (rick@cs.arizona.edu)
Associate Professor
Department of Computer Science
The University of Arizona
Tucson, AZ 86721, USA
[This report represents the personal opinion of the author, who was on
sabbatical in Japan from Dec. 1989 through July 1990. The sabbatical
was supported in part by grant INT-8910818 from the NSF U.S.-Japan
Cooperative Science Program.]
---------------------------------------------------------------------------
On April 25, I visited Fujitsu's Numazu complex, which is at
the base of Mt. Fuji near the city of Mishima. The trip was arranged
by Prof. Yoshihiro Tohma of the Dept. of Computer Science at Tokyo
Institute of Technology where I was staying on sabbatical. He had been
invited by Fujitsu to present a lecture on fault-tolerant computing at
the plant, and came up with the idea of having me accompany him to
present a complementary lecture on fault-tolerant software techniques.
We started out from the institute at about 8:55 AM and headed to
Shin-Yokohama station. There, we met Yohichi Suzuki, a former student
of Yoshi's (the first, it later came out) who was to be our host for
the visit; his current position is that of deputy general manager of
Fujitsu's mainframe division. After introductions, we caught the 10:05AM
Kodama shinkansen for Mishima station, where we arrived at approximately
10:50. Another 20 minutes by taxi brought us to the Numazu complex, which
is situated in a wonderful location not far from the coast. We were
escorted to a meeting room, where we settled into a short discussion with
Suzuki, Masahito Date, a software manager who was familiar with my work
on fault-tolerant computing, and Makoto Kimura, a hardware specialist who is
also a graduate of Tokyo Institute of Technology.
The talks began at 1:15PM, and we got to the room right about on
time. We had been informed during lunch that there were going to
be about 180 people at the presentations, and they were all there when
we arrived. After introducing us, Yoshi began his talk (in
Japanese, of course.) His topic was a general introduction to
fault-tolerance, during which he gave motivations, overviewed some
current companies and architectures, etc. Afterwards, he showed
showed a 20 minute videotape of the SAFE system that they had
worked on at TIT.
After a short break, it was my turn. I gave a general overview
of fault-tolerant software, which included talking about various
useful abstractions. It went quite well, I think, and I later got
good feedback on the talk over dinner. There were a couple of
good questions after the presentation, which were given in Japanese
and translated for me by a fellow in the front row.
Following the talks, we were shown around the plant, with a
tour guide leading the way. By that time, we had been joined by
Katuhei Oshio, who is manager of the TMP Development Dept. of the
Product Assurance Division. The Numazu complex is the manufacturing
site for the high-end mainframes and supercomputers, and we got to see
how they were assembled. It was fascinating, especially for a hardware
novice like me. I thought it especially interesting that most
of the assembly for such complex machines is done by hand, with
limited automation only at a very few stages. We were also
shown the rooms in which they were tested, the main software labs
and system test rooms, a little history museum with an old
relay computer, and a piece of natural language translation software
called ATLAS II.
We left around 8:30 by taxi and made it to Mishima station
just as the train was pulling up. In addition to Mr. Suzuki, we
were also joined by Mr. Oshio since he lives in Yokohama City.
We talked about a number of general things on the way back,
including plans for the Pacific Rim Fault-Tolerant Computing conference
to be held the following year. (Mr. Oshio is the local arrangements
chair.) We arrived at Shin-Yokohama station around 9:30, where we caught
a taxi back to Ookayama.
Addresses
----------
Yohichi Suzuki
Deputy General Manager
Main Frame Division
Fujitsu Limited
1015 Kamikodanaka, Nakahara-ku
Kawasaki 211, Japan
Masahiro Date
Manager, 3rd Section
Development Planning Dept. I
Software Division, Computer Systems Group
Fujitsu Limited
140 Miyamoto, Numazu-shi
Shizuoka 410-03, Japan
Katsuhei Oshio
Manager, TMP Development Dept.
Product Assurance Division
Fujitsu Limited
1405 Ohmaru, Inagi
Tokyo 206, Japan
Matkoto Kimura
Design Sect. II
Computer Engineering Dept. IV
Main Frames Division
Computer Systems Group
Fujitsu Limited
1015 Kamikodanaka, Nakahara-ku
Kawasaki 211, Japan
Prof. Yoshihiro Tohma
Department of Computer Science
Tokyo Institute of Technology
2-12-1 Ookayama, Meguro-ku
Tokyo 152, Japan