|
"The 21st century should pave the way to a millennium that excels in
science, technology, and the way in which these disciplines interface
with society. Advanced computing and computational science will be indispensable
parts of the new ethos, and SciDAC will help lead the way."
|
DR MICHAEL STRAYER
Director, SciDAC program
|
|
| By Dr Michael Strayer |
| The end of the 20th century marked the end of the
first 50 years of the development of advanced computing
for the pure and applied sciences. As Stephen
Jay Gould argues in his monograph Questioning the
Millennium: A Rationalist's Guide to a Precisely Arbitrary
Countdown (1997, Harmony Books), it is hard to identify
when the exact beginning of this century
occurred, due to numerous - and often conflicting
- cultural and historical views. Within the context
of the Office of Science (SC) at the Department of
Energy (DOE), however, the start of the first SciDAC
program within the Advanced Scientific Computing
Research (ASCR) program was probably what
marked the beginning of the new era. Five years on,
SciDAC has added a new dimension to science and its
practice, and has driven both pure and applied sciences
to new plateaus of achievement. For scientific
discovery, the program has transformed both our
approach and our understanding of science. And in
terms of 20th century methodologies, it has proved
revolutionary in the capabilities and precision it
brings to science.
|
|
The SciDAC program has developed enabling technologies
for scientific discovery using the most
advanced software ideas available. Developed for
massively parallel high-end computers, these technologies
have been transformational drivers of
change. Part of the change has been societal. One of
the key successes of the SciDAC program has been its
ability to integrate diverse interdisciplinary groups
that are focused on scientific discovery. The individual
investigator approach has evolved into a tripartite
partnership between discipline scientists, applied
mathematicians, and computer scientists. These now
largely define the mission and function of SciDAC.
Moore's law suggests that before the end of the next
cycle of SciDAC, we shall see petaflop computers. The
priorities of the SC are clear. The 20-year facilities
plan is driven by new science. High-performance
computing is placed as one of the two highest priorities.
Our effort in leadership-class computing is just
a first step towards this goal. Clearly, computational
science at the petascale with hundreds of thousands
of processors will face enormous challenges, and
holds immense promise. Performance evaluation and
assessment will be critical to unraveling the software
technologies required.
|
|
The landscape of science, discovery and computing
does not stand still, and in this environment,
growth needs to be measured in terms of applications
and achievements. I believe it is necessary to institute
a system of science-based performance metrics to
help quantify our progress towards our goals and
measure growth. Such performance metrics could,
for example, track accountability and performance
in terms of processor hours. Since particular problems
and disciplines differ greatly in their science variables
and scales, metrics would need to be
normalized across disciplines. Quality, validation,
and relevance must be considered when tuning the
metrics, and the discovery factor should perhaps provide
the leading edge.
|
|
New challenges, new possibilities
|
|
In computational science, as in the other sciences, the
arena of international collaboration is the forum for
science for the 21st century. The SC is heavily committed
to multiple international collaborations -
including the International Thermonuclear Energy
Reactor, the International Linear Collider, and the
Large Hadron Collider, to mention a few. SciDAC,
together with ASCR facilities, could provide powerful
resources and the nexus of a new global village for
computing that could take computational science
and scientific discovery to wholly new levels.
Expanding SciDAC-like partnerships that foster collaborative
research with the applied sciences is of
great importance. SciDAC can harness the power of
petascale computing for simulations and modeling
to provide dynamic insight into the areas of environmental
and waste management, and developing virtual
technologies and prototyping for nuclear, fossil,
and solar energy systems.
|
|
Basic and applied sciences influence and empower
society in multiple ways. The 21st century should
pave the way to a millennium that excels in science,
technology, and the way in which these disciplines
interface with society. Advanced computing and
computational science will be indispensable parts of
the new ethos, and SciDAC will help lead the way.
|
|
Dr Michael Strayer is Associate Director for Advanced
Scientific Computing Research (ASCR), Office of Science,
US Department of Energy; Director, Scientific Discovery
through Advanced Computing (SciDAC); and Acting
Director, Mathematics, Information, and Computational
Sciences (MICS).
|
|
