Page 17 - Kotov

Crucial Dilemma: To Copy or To Invent
Time Frame
By mid-60s computers started their rapid expansion from their birthplac-
es - science, high-education, and defense institutions - to banking, com-
merce, and industrial design and manufacturing. Large-scale production,
maintenance, and use of computers required also to move from a variety of
non-compatible machine and devices to system families designed to cover
the complete range of applications, from small to large, both commercial
and scientific. One could expect that in the Soviet Union, with its state econ-
omy, the variety of computer models was small and some unification was
already in progress by that time. That was not the case. Dozens of organi-
zations were developing their proprietary architectures without any serious
attempts to unify interfaces, software, and applications. When IBM created The IBM S/360
an entire series of computers System/360 (S/360) of the “third generation”
from small configurations to large ones, all using the same instruction set,
customers were able to buy cheaper models and then upgrade them to larger systems as their needs increased
without the time and expense of rewriting software.

Dilemma: to copy or not to copy

In 1966 the Soviet government requested that new computers for the use in the growing national economy should
be built on a unified architecture and microelectronic technology and their system software should be compat-
ible. However, there was no consensus how to reach this goal quickly. Some proposed to start new proprietary
compatible computer family from scratch or developing some elements of hardware compatibility that had been
demonstrated by some domestic families, for example “Ural” (Ural-11, Ural-14, Ural-16) . Others argued that
was better to copy some well-tested and widely popular western families. The acquisition of licenses from
Siemens (Germany) and ICL (UK) that produced their own compatible computer families was considered. How-
ever, the required amount of hard currencies was not available. And it was not possible to get a license from the
American government to reproduce the IBM S/360 system.What was possible was to create the Soviet computers
programmably compatible with the IBM S/360 system.

Arguments against copying Arguments for copying
Copying IBM S/360 will put the Soviet Union sev- It was impossible to say with certainty how long it
eral years behind the USA and other countries (the would take to build the domestic third-generation
IBM S/360 was launched in 1964). computers.
The IBM software was the most rich, well-document-
Switching to copying would destroy the creativity ed, and advanced by that time. Creating IBM-com-
and innovation in the domestic computer communi- patible machines promised quick computerization of
ty and kill just recently emerged domestic computer the USSR economy.
R&D institutions.

Copying the American system will make the Russian
computers vulnerable to the “cyber attacks” of that
time by the American military or CIA.

After a long and heated debate, the Soviet government followed the Solomon judgment: the defense industry
would mainly rely on the development of proprietary domestic computer architectures (with a support from the
research institutions) and the national economy would embrace the computer families patterned on the most pop-
ular American architectures. The first thread was represented by the Elbrus and specialized military computers.
The second one was represented by the Unified System and the System of Small Computers.

– 15 –