Dolphin Interconnect and RackSaver install AMD Athlon processor-based Supercomputer at the University of California, Santa CruzWestlake Village 21 May 2001 Dolphin Interconnect and RackSaver completed installation of a Beowulf cluster supercomputer at University of California, Santa Cruz (UCSC) using 32 1.2 GHz AMD Athlon processors. This cluster, funded primarily by the National Aeronautics and Space Administration (NASA), will be used to study collision processes in the solar system, such as the origin of Earth's Moon by a giant impact, the disruption of meteors in the atmospheres of planets, the formation of impact craters, and the fragmentation of asteroids by impacts and explosions. It will also be used to run simulations of planetary dynamos, such as the one responsible for Earth's magnetic field.
The 32-processor supercluster, a planned 256-processor
supercluster funded primarily by the National Science Foundation (NSF)
will stand alongside it this summer, will be used to model an
even wider range of astrophysical problems, including supernova explosions
and the universe's origins.
"The
UCSC Earth Sciences, Astronomy and Physics departments are very excited
to have access to this new technology, because it enables us to solve
complex research problems in-house," said Erik Asphaug, UCSC
principal investigator of the new 32-node research supercluster and
co-investigator on the proposed enhanced supercluster. "In the
past, computations requiring high speed processing and extremely fast
inter-processor communication would have to be solved piece by piece,
whenever time was available on mainframe supercomputers. Now we can
run a job 24 hours a day, 7 days a week on our own machine."
He also emphasized the educational value of having a research cluster
on campus stating, "Researchers and graduate students are beginning
to recognize that massively parallel computing is part of the modern
scientist's toolkit."
Dolphin
Interconnect technology provides the high-speed inter-communication
pathway required by the 1.2GHz AMD Athlon processors in the cluster
to match the speed of the AMD Athlon processor power. The system employs
1GB bandwidth between nodes using Dolphin Interconnects SCI (Scaleable
Coherent Interface) Wulfkit, offering an efficient combination of
fast network hardware and corresponding software. The AMD Athlon processors
employed are loaded with a software program that breaks the scientific
equations into sub-problems to be solved on each processor. RackSaver
Inc., which specializes in designing flexible, yet powerful systems
that can fit into specific space limitations, successfully integrated
the supercluster at UCSC, meeting the university's space and performance
requirements.
In
the Beowulf cluster design, a series of processors are interconnected
and fine-tuned to speed communication between the processors. Each
processor performs the time-consuming computations concurrently on
a subset of the large data set and sends its result to the processor
acting as the manager. The processors then share their information
and, if the communications link can handle the load, arrive at the
answer much faster than a single processor computer.
Ad Emmen |
 |