Japanese Computer Pioneers

Watanabe HitoshiWatanabe Hitoshi

Watanabe Hitoshi was born in Shimane prefecture on December 26, 1930. He graduate from E.E. Department, School of Engineering, Kyoto University in 1953 and joined NEC Co.. In September 1954, he was assigned in designing electrical wave filters. Even though "filter" was one of the most critical devices for FDM (Frequency Multiplex) communication systems, it was designed by an empirical approximation method, and therefore, hardly satisfied stringent design requirements of FDM systems. Watanabe developed an exact design method of filters based on the circuit theory, and applied it to practical filter design. This design process involved enormous amount of numerical calculations, and Watanabe wrestled with it using a motor-driven calculator through months. This difficulty was like climbing up the steepest precipitous cliff. To overcome the difficulty, he was firmly determined to develop an automatic computation machine.

Although there was neither a single working computer in Japan nor a technical paper on computers in the year of 1954, Watanabe found the fundamental architecture that such a machine ought to have, through his own experience of manually conducting the numerical calculations using a motor-driven calculator. In his manual calculation procedure, he viewed the motor-driven calculator as "arithmetic registers", the paper sheet (memorandum) as the "memory device", and the function keys of the calculator as the "control unit". In other words, his procedure was merely a repeated sequence of operations of (1) reading numeric data from the paper sheet (memory device), (2) putting the data on the calculator (arithmetic registers), (3) pushing appropriate keys of the calculator according to his calculation procedure (programming), and (4) writing the calculated result to the paper sheet (memory device), while the calculator was doing the addition by counting the number of rotations of gears. As a natural consequence, the fundamental architecture and everything about the design diagram of an automatic computation machine (i.e. the computer) popped up into his brain. It was the fall of 1954.

In spite of his eager entreaty, his proposal of developing a digital automatic computer was not approved by the company as an official project. Therefore, Watanabe worked as a circuit design engineer at the company during the day time, while he devoted himself to the development of a digital computer at his lodging late at night. Continuing this double working mode for more than two years, he finally completed the fundamental design of his computer at early 1957, using parametron as its logical element. This computer had several very unique features. One of the most salient was the dual arithmetic structure. Namely, it provided two floating-point arithmetic units with the 40bit mantissa and the 8bit exponent. When the two units were concatenated, this computer worked as a double precision floating-point machine with the 80bit mantissa and two sets of 8bit exponents. In April, 1957, TOHOKU University and NEC contracted a joint computer project, and the computer designed by Watanabe was fully accepted as the prototype. After serious assessment and enhancement of the prototype, the machine was completed in November 1958 and was named SENAC-1 (NEAC1102). In January 1960, Watanabe accomplished its improved machine as the computation machine for circuit designs. This long desired machine was named NEAC1103, and was successfully used for practical circuit design project through 1972. Following this project, Watanabe had been engaged in broad areas of CAD research, as well as development of small-scale computers.

From 1991 to 2005, he had been a professor of SOKA University.

He received IEEE Gustav Robert Kirchhoff Award for "Pioneering Contributions to filter design theory and computer-aided circuit design (so-called CAD)" in June 2010.

(As of Nov. 5, 2010)