[EAS]Seymour Cray

[pjk]

Subject:   Seymour Cray

(from NewsScan Daily, 12 July 2002) [comment at the end, as usual.--PJK]

HONORARY SUBSCRIBER: SEYMOUR CRAY
      Today's Honorary Subscriber is the American electronics engineer
Seymour R. Cray (1925-1996), whose name has become synonymous with 
large-scale, high-performance computing because of the Cray
supercomputer he designed and built in 1976.
      Cray was born in Chippewa Falls, Wisconsin. His father was the
town's  civil engineer, and Cray grew up fascinated with electronics,
spending much  of his time in his father's electrical engineering
laboratory, toying with  radios, electrical motors and other kinds of
electrical equipment. He  graduated from high school in 1943 before
getting shipped off to war, where  he spent time in both the European
and Pacific theatres before returning  home and resuming his
education. He earned a Bachelor of Science degree in  electrical
engineering from the University of Minnesota in 1950 and a  Master of
Science degree in applied mathematics a year later. He began his 
career as a computer scientist working on UNIVAC I, a landmark 
first-generation electronic digital computer that became the first 
commercially available computer.
      In 1957 Cray helped found Control Data Corp., which became a
major computer manufacturer. There Cray designed the CDC 6600 and the
CDC 7600, large-scale computers notable for their high processing
speeds.
      In 1972 he left Control Data to found his own firm, Cray
Research  Inc., with the intention of building the world's highest
performance  general-purpose supercomputers. He realized his goal
through a multiprocessor design that employed computer processors
working simultaneously in parallel. His company's first supercomputer,
the Cray-1, which came out in 1976, could perform 240,000,000
calculations per second.  It was used for large-scale scientific
applications, such as simulating  complex physical phenomena, and was
sold to government and university laboratories. Further supercomputers
followed, each with increased computing speed: the Cray 1-M and the
Cray X-MP.
      Cray resigned as chairman of his growing firm in 1981 and became
an independent contractor to the company, designing ever-faster
machines at  his laboratory in Chippewa Falls. In 1985 the Cray-2 was
introduced to the  market. This machine, which was cooled by
Fluorinert, could perform  1,200,000,000 calculations per second.
      In 1989 Cray founded the Cray Computer Corporation, but with the
end  of the Cold War the military demand for supercomputers
disappeared, and by  1995 Cray Computers was forced to file for
bankruptcy without having sold a  single computer. Nonetheless, in
1996 Seymour Cray undertook a new computer  venture that was just
barely off the ground when his life ended in a tragic  auto accident.
      Seymour Cray's admirers speak of him "the Thomas Edison of the 
supercomputing industry," and they like to recall the time when upon
being told that Steve Jobs bought a CRAY to help design the next
Apple, he said, "Funny, I am using an Apple to simulate the CRAY-3."

See
http://www.amazon.com/exec/obidos/ASIN/0471048852/newsscan/ref=nosim/ 
for Charles J. Murray's "The Supermen: Seymour Cray and the Technical 
Wizards Behind the Supercomputer" -- or look for it in your favorite 
library. (We donate all revenue from our book recommendations to adult
 literacy action programs.)

---------------------------------------------------------------------
Mention of Cray growing up fascinated with electronics reminds me how
different it was to educate students majoring in EE in the days when
many of them had some previous personal exposure to electronics, e.g.
ham radio. And I mean more than their invariably greater comfort in
the lab.
The educator's always critical responsibility of communicating _why_
the subject matters, not just _what_ it consists of, took the much
easier form of building on, of enlarging, a sense of importance they
felt already in starting the EE major. 
These days I often feel I have to create that sense of importantance
from the much more abstract or diffuse vocabulary of "solving
open-ended problems", of "social impact," of "economic significance,"
of "personal convenience."
In few of these, except the personal convenience, do they have much
previous personal investment. And the pervasive and convenient
presence of EE, and much other technology, as motivation for studying
it, is like trying to interest a "fish" in majoring in "water."  --PJK