Morningstar - Fall 2019 - 14

Dispatches

come from, he picked Chippewa Falls, Wisc.
(population 14,000). I went there last summer, and
most of Chippewa Falls does look like the kind
of boring place that would drive a bright girl to New
York. I was there, however, to visit Cray Research,
one of Acorn's important holdings, a company
which has made Chippewa Falls a major center of
the computer world. Seymour Cray is the best
computer designer in the world. He made Control
Data the leader in scientific computers in the
1960s and then went out on his own, to the woods,
to build the fastest scientific computer in the
world. He did it, with very few people and in a short
time, to the surprise of the most observers. It was
a surprise because IBM, Burroughs, and Texas
Instruments had tried to build ultrafast computers
and failed. The Cray-1 works well and can
execute 80 million instructions per second on the
right kind of problem, 10 to 100 times as
fast as competitive machines. Cray Research is
a fascinating, high potential company, and
it will be exciting to see whether Seymour Cray's
inventive genius will also create a financially
successful company. The signs are favorable so far.
There were lots of reasons not to buy Cray stock.
One was that there were a number of companies
that had tried to build a supercomputer,
including IBM and Burroughs.2 New pieces of
hardware and software were under development
in many places, but no one had been able to
produce a useful system. Daniel Slotnick, the head
of the Burroughs ILLIAC supercomputer project,
said after he delivered the machine to the NASA
Ames Research Center in what is now Silicon
Valley that he was "dismayed it cost too much,
took too long, doesn't do enough, and not enough
people are using it."3
Bet on a Genius
Plus, how could Seymour Cray run a startup
company in the Wisconsin woods and succeed?
But as I studied Cray's expertise in designing
many excellent computer systems, I concluded
that he was a genius on the same level as
Thomas Edison. If the company failed, it would be
just one more loss in a diversified portfolio,

but if I didn't buy it, and Cray lived up to his
reputation and built a highly successful company,
I would spend the rest of my life regretting
that I was too chicken to write the buy ticket.
One of my rules of life was to insist that any stock
recommendation have a "reason to own."
For Cray Research, the reason was "bet on
Seymour Cray."
One of Cray's rules of life was to never talk to an
investor. When I drove to Chippewa Falls, passed
the Leinenkugel's brewery and the excitement
of downtown, I arrived at the Cray offices.
I was able to talk to the CEO and tour the plant,
but no Seymour. He had retreated to a cabin
on an unnamed road on an unnamed lake, where
he and a couple of buddies designed the
world's best computer. He was not a nice man,
but neither was Beethoven.
One of the technical problems with building large
electronic devices in the 1970s was cooling
them. The engineers wanted to build circuits that
were close together because they would run
faster, but then the machine would overheat. Cray
was an expert on getting rid of the excess heat.
Not many people remember this, but in the 1960s
and '70s, Minneapolis/St. Paul was the computer
capital of the world. Great engineers such
as Cray and Slotnick started their careers at Univac
and produced the first usable computer systems.
Then, Control Data and Minneapolis Honeywell
started up and triggered a lively technology stock
boom in the Twin Cities.
Stay Downstream
In general, in those days, I did not like companies
that made computer hardware. The rapidity of
advances of computer technology meant that most
of the companies that made hardware
components found their product leapfrogged in
an embarrassingly short time. Fine companies
such as General Electric GE, RCA, and NCR tried
to compete in computer systems but failed.
Companies such as Digital Equipment had success

2 In 1978, Burroughs was considered a major player in the computer industry with an operating system superior to IBM's.
3 Hord, M.R. 1990. Parallel Supercomputing in SIMD Architectures. CRC Press.

Morningstar Fall 2019

for a while but eventually failed. Cray was a
founder of Control Data in 1957. That company was
an important maker of large computer systems,
but it broke apart in 1986. Even mighty IBM IBM,
possessor of an unassailable monopoly in 1978,
failed to adapt to the PC revolution, and today
is a pale reflection of its former glory.
I much preferred to be "downstream from
technology." Motorola MSI might make a highly
engineered microprocessor chip and sell it
to International Game Technology IGT for $10. IGT
would add some silk-screened glass displays,
a chip-handling device, and a cabinet and sell
it as an electronic slot machine for $15,000.
Harrah's would buy the slot machine and make
$60 per day once it was mounted in its
Las Vegas casino. The further downstream, the
more profit.
Today, the best companies are downstream-fromtechnology plays. Facebook FB is a great
example. The technological wonder of the Internet
is enabling the instantaneous transmission of
billions of pictures of kids and pets. Cray Research
was a profitable exception to the downstream
rule. But by the end of 1991, we had sold all of our
Cray Research stock, and Acorn had bought
a lot of International Game Technology, a much
better stock in the 1990s.
The Cray company continued in business, but
Seymour Cray died in a car crash in 1996.
The remnants of the company carried on, but
in May, Hewlett Packard HPE bought Cray
for $1.4 billion, a small amount of money for a
company that had dominated the supercomputer
market for 47 years. K
Ralph Wanger, CFA, ran Columbia Acorn Funds from 1970
to 2003.

Morningstar - Fall 2019

Table of Contents for the Digital Edition of Morningstar - Fall 2019