B Associate OB W Director EBSTER Weapons Physics
for The new associate director discusses his
viewpoints regarding the nature of the
science and scientists at the Laboratory...
and why he misses using his slide rule.
Recently, Bob Webster was named the Laboratory’s
Associate Director for Weapons Physics. He oversees the
Computational Physics and Theoretical Design divisions,
as well as the Laboratory’s Advanced Simulation and
Computing (ASC) program. The Laboratory’s ASC
capabilities are inextricably woven into the work of weapons
physics and design. Webster recently spoke with National
Security Science (NSS) about his new role at the Laboratory.
Webster: I think there was some value in the way that we
had to think about the problems when we were still using
slide rules, a way that we could reintroduce into the system
right now. At the same time, though, there is a tremendous
opportunity presented by leaving slide rules behind.
NSS: Where in the evolution of computing did you start your
career? If you look at the last 20 to 25 years, there’s a fundamental shift
in how supercomputing underwrites our evaluation of scientif-
ic problems. In the ’70s, supercomputing, or high-performance
computing, which wasn’t very “high performance” by today’s
standards, was sometimes viewed as a crutch.
Webster: I think I was in the last class at Case Western
Reserve University that used slide rules in the exams. Slide
rules were abandoned between my freshman first semester
and the spring semester that year; the university finally let
us use calculators. So initially many of us were still carrying
“slip sticks” to engineering classes. That was a different era
in terms of how we thought about solving physics problems
because we didn’t have computers the way we have them now.
Today it’s an integral part of synthesizing theories—we can
evaluate very complex scenarios that we can’t actually test.
For economic, political, and risk factors, we can’t always
employ the classic, direct scientific experimentation that we
were taught to do. I think that’s something we need to get out
to folks—supercomputing is integral. We can’t separate
it from doing the experiments and doing the analytic
NSS: So do you think anything was lost by leaving the slide
But there was loss there with leaving the slide rule behind.
We started to leave experiments behind more than we should
have. Experiments got very expensive, so there’s a tendency
52 Los Alamos National Laboratory