Vviewpoints
DOI: 10.1145/1743546.1743560
The Profession of iT
The resurgence
of Parallelism

MULTi-Core ChiPs Are a new paradigm!” “We are entering the age of parallelism!” These are today’s faddish rallying cries for new lines of research and commercial development. Is this really the first time when computing professionals seriously engaged with parallel computation? Is parallelism new? Is parallelism a new paradigm?

 

Déjà Vu all over again Parallel computation has always been a means to satisfy our never-ending hunger for ever-faster and ever-cheaper computation. 4 In the 1960s and 1970s, parallel computation was extensively researched as a means to high-performance computing. But the commercial world stuck with a quest for faster CPUs and, assisted by Moore’s Law, made it to the 2000s without having to seriously engage with parallel computation except for supercomputers. The parallel architecture research of the 1960s and 1970s solved many problems that are being encountered today. Our objective in this column is to recall the most important of these results and urge their resurrection.

shared memory multiprocessing
The very first multiprocessor archi-
tecture was the Burroughs B5000, de-
signed beginning in 1961 by a team led
by Robert Barton. It was followed by
the B5500 and B6700, along with a de-
fense version, the D850. The architec-
ture survives today in the reverse polish
notation HP calculators and in the Uni-
Sys ClearPath MCP machines.