High-Performance Computing
U Victoria, U Michigan Partner on HPC for High Energy Physics Research
Physicists at the University of Michigan and the University of Victoria in Canada have
implemented a multi-site supercomputing project to help them share massive
volumes of data with the CERN Laboratory in Geneva, Switzerland and 100
computing centers around the world.
In 2015, the Large Hadron Collider at CERN will begin "colliding elements at
the highest highest energies ever achieved in a particle accelerator," according
to a news release. Physicists at the University of Michigan and the University
of Victoria will use the data from CERN to help discover new particles and
forces. They need to be able to transfer approximately 170 petabyte datasets
from CERN at speeds of 100-gigabits per second, so they can access the data
quickly, analyze it and speed up discoveries.
To achieve those speeds, the researchers created a single server data
transfer architecture using SanDisk's Fusion ioMemory,
By implementing a single server rather than a multi-server architecture, they
were able to reduce the server's footprint, complexity, cost and
points-of-failure, according to a news release from SanDisk.
"The ATLAS and CMS supercomputing projects are very large international
projects, each involving approximately 3,000 researchers and most of the world's
countries," said Randall Sobie, a scientist at the Institute of Particle Physics
Research and professor at the University of Victoria, in a prepared statement.
"These are long term projects — they started 20 years ago and will continue for
another 20-plus years."
The physicists demonstrated the data transfer at the SuperComputing 2014 conference in
New Orleans in November.
About the Author
Leila Meyer is a technology writer based in British Columbia. She can be reached at [email protected].
