One Family of Computers Early 1964

You are advising Thomas J. Watson Jr., chief executive of IBM.

IBM sells several successful computer lines that cannot run each other’s software. Its engineers have designed a single compatible family of computers, from small to large, that would replace almost all of the existing lines at once.

Known information

Developing the new family is costing billions of dollars, an enormous commitment for the company.
Customers who move to a bigger IBM computer today usually have to rewrite their programs.
Announcing the new family could make customers stop ordering IBM’s current machines.
Competitors are already targeting IBM’s customers with cheaper machines that can run IBM programs.
Several models, new circuit technology and a large operating system must all be ready at roughly the same time.

Reveal what happened (spoilers: decide first)

Fact

What actually happened

IBM announced the System/360 family. (Announce the whole family)

IBM announced the System/360 on 7 April 1964: six processor models covering a wide range of performance, sharing one architecture so that software written for one could run on others. Orders exceeded 1,000 in the first month. Its operating system software ran late and over budget, but the System/360 family went on to dominate business computing.

Important

Possible consequences · Analysis and speculation, not fact

Continuing with separate lines might have protected short-term sales but left IBM more exposed to competitors offering compatible machines. Business historians disagree on how quickly such a shift would have come without IBM leading it.

History does not reveal what would definitely have happened under the other choice. Many factors could have changed the outcome.

Context · Why it was decided this way

IBM’s own history describes the project as a “bet-the-business” move costing about $5 billion over four years. It established the idea of a compatible computer platform, and its 8-bit byte became an industry standard.

The software difficulties inspired Fred Brooks, who managed the project, to write The Mythical Man-Month, a classic book on why adding people to a late software project can make it later.