Recession, S-Curves and Digital Equipment Corporation - History of Computing: Learning from the Past - page 182

in the US. This was a forced reaction to the sudden decline in the stock market value
of the company in 1983 when the value of the stock dropped by 29% in three weeks
due to reporting problems within the company. They finally imposed a hiring freeze,
retrained 4000 new manufacturing personnel and only had to make 600 redundant as
illustrated by Rifkin and Harrar [3]. This was undertaken over a three year period
from 1983.
The theory of the S-curve suggests that all businesses follow an S-curve in their
development taking a certain amount of time to get to 10% market share and a similar
amount of time to reach 90%. At which point, the company will fade away unless
they can re-invent their product/industry and begin a new S-curve. Modus [4] states
that:

“ The projected life cycle of consumer products and the rate at which substi-
tute products will gain market share is of vital interest to any company ”. He
suggests that “ Business, in the form of products, companies and entire indus-
tries, goes through five cycles which align with the S-curve ”.

He also suggests that the S-shaped curve also shows up in other life cycles. For ex-
ample he states that :

• A product S-curve may typically have a life cycle of 6 quarters.

• A product family S-curve, consisting of a set of related products, will typi-

cally have a life cycle of around 5 years.

Basic technologies or industry S-curves, consisting of a number of product
families and associated companies, typically have a life cycle of approx. 15
years.
DEC followed this cycle successfully for many years. For product life cycles, they
released a new major product almost every year from 1965, thus having overlapping
S-curves. For family product lifecycles they released the PDP-8 in 1965. The last
model in the family was produced in 1979 and they sold over 50,000 systems. There
were 10 different models released in the 15 years. This is accepted as the first real
minicomputer and heralded the start of affordable computing. Gordon Bell and Ed de
Castro are credited as being the main designers of the PDP-8. Five years later in 1970
they followed it with the PDP-11, with the last product in the family being released in
1990. The PDP-11 family, excluding the 32 bit extensions, consisted of at least 23
models and was the leader in the minicomputer market for many years. In 1975 DEC
released the 11/70, this was meant to be a stop-gap machine with 1000 planned. Even-
tually 10,000 were sold. Then, in 1977 the 11/780, DEC's first 32-bit machine, was
released. In 1984 the VAX 8600 came out, in 1990 they released the VAX 9000 and
finally in 1992 DEC released the 64 bit Alpha family. By the time DEC was taken
over by Compaq, the VAX family consisted of around 135 models with the final
VAX, in the Alpha range, being manufactured in 2005. These cycles fed DEC's in-
credible growth over the years.
Christensen [5] looks at technology S-curves and asks how value networks and the
concept of S-curves relate to each other. He postulates that disruptive technology does
not fall into the normal S-curve as it gets its commercial start in emerging value net-
works before invading established networks. Clearly DEC had disruptive technology
with its minicomputer products, taking the mainframe makers by surprise and creating

•