last byte

DOI: 10.1145/1435417.1435440

Q&a
the upper limit

Jeannette Wing talks about clusters, creativity, and the power of computational thinking.

a proFeSSor oF

computer science at

Carnegie Mellon University, Jeannette Wing has long been interested in the power of abstraction. She is a leading researcher in the area of formal methods, and has worked on diverse problems in software specification and verification, security, concurrent and distributed systems, and programming languages. Wing is also deeply engaged with a concept she calls computational thinking: the way computer scientists use decomposition, recursion, and algorithms to tackle difficult problems. As the current assistant director for computer and information science and engineering at the National Science Foundation (NSF), she now has the opportunity to put her talent for abstraction to use and help shape the future of the field.

you’re about a year and a half into
your tenure at nsf. What’s it been
like?

It’s been an eye-opening experience. It gives me a much better appreciation of the university/industry/gov-ernment ecosystem, and how we all need to work together to advance the frontiers of science and engineering.

What role do you play within the
organization?

I wear two hats. One is to act as a spokesperson for the field, a scientist who represents the whole computing community—and let me just use the word “computing” to stand for all of computer and information science and engineering. At the same time, I

also have the opportunity to help guide and shape the frontiers of our field.

how’s that?

NSF is really a bottom-up organization. We listen to the computing com-

“computational
thinking encompasses
all the benefits
of computational
models and methods
and tools and the
abstractions we use.”

Leah Hoffmann

munity to find out what the trends are, where they’re going, and what their research interests are, and we try to respond to their needs.

let’s talk about some of your most
recent initiatives, such as the clus-
ter Exploratory program.

The Cluster Exploratory program enables us to offer software and services running on a Google-IBM cluster to the entire computing community. For a long time, we could see that there was something very exciting happening in industry with respect to these large data centers. But no university can really afford to build its own center and run it and pay for the power. Now, finally, researchers can access these resources and run their experiments at scale.

there’s also a second cluster at
the university of illinois at urbana-
champaign (uiuc).

Yes, thanks to HP, Intel, and Yahoo!. It’s hosted at UIUC, but it’s being made available to all of academia. And this time, it’s not just the software and services that we’re providing, it’s access to the bare machine, so researchers can run experiments that go down to the processing level and below.

PhotograPh by richarD Kelly PhotograPhy

has your involvement with nsf
altered your own research inter-
ests?

Being here has increased my awareness of the need to do more fundamental research into privacy, and to understand things like information flow and use [coNtINued oN p. 111]