TITLE: Turing Award History: Fran Allen
AUTHOR: Eugene Wallingford
DATE: March 21, 2007 2:03 PM
DESC:
-----
BODY:
For the last month or so, I have been meaning to write about a
historic announcement
in the computing world, that Frances Allen receiving the Turing
Award. This is the closest thing to a Nobel Prize in computing
as there is. The reason that this announcement is historic?
Allen is the first women to win a Turing. Women have long made
seminal contributions to computer science, something that most
computer scientists know and appreciate. But it's a milestone
when these contributions receive prestigious and
public acknowledgment.
You can also listen to a short
interview with John Hennessy,
the well-know computer scientist who is now
president of Stanford,
in which he puts Allen's award into context.
In one of her interviews since receiving the award, Allen reminds
us that women these days are underrepresented in computing compared
to when she began her career in the 1950s. Women were better
represented in computing than in most other science and technology
disciplines up through 1985 or 1990, at which time their numbers
began to plummet -- precipitously. This is a trend that most of
us in computing would like to reverse, and one that Allen has
working to correct since retiring from IBM five years ago.
Let's not allow the social implication of Allen's award to steal
attention from the quality and importance of the technical work
she did. The chair of the Turing Award committee said, "Fran
Allen's work has led to remarkable advances in compiler design
and machine architecture that are at the foundation of modern
high-performance computing." Her primary contributions came
in the areas of compiler design and program optimization,
techniques for making compilers that produce better executable
code. Her early work is the theoretical foundation for modern
optimizers that work independent of particular source languages
and target machines. Her later work focused on optimization of
programs for parallel computers, which contributed to high-performance
computing for weather simulation and bioinformatics. I found
one of her seminal papers in the ACM digital library:
Control Flow Analysis;
check it out.
Thinking about this award helps us to remember the "applied"
value that derives from basic scientific research in an area
as theoretical as compiler optimization can be. By making it
possible to write really good compilers, we make it possible
to create higher-level programming languages. This makes
programmers more productive and also widens the potential
population of programmers. By advancing parallel high-speed
computing, we make it possible to study much larger problems
and thus address important social and scientific questions.
This latter point is an important one in the context of trying
to make computing more attractive to women, who seem to be
more interested in careers that "advance the public good" in
obvious ways. Allen herself has stated her hope that
high-performance computing's role in medical and scientific
research will attract women back into our profession.
-----