Communications of the ACM
.S. Collaborators to Make Higgs-Hunting Tech Available
The Atlas detector can detect a wide range of energies, and creates an enormous data flow.
Credit: CERN
The University of Texas at Arlington is teaming with the U.S. Department of Energy's Brookhaven and Argonne national laboratories to develop a universal version of PanDA, a workload management system built to process huge volumes of data from experiments at the Large Hadron Collider in Geneva, Switzerland.
The new project will bolster science and engineering research that relies on "big data," a priority recently promoted by The White House. The U.S. Department of Energy's Office of Advanced Scientific Computing Research has awarded a combined $1.7 million to Brookhaven and UT Arlington to fund the PanDA work over the next three years.
"PanDA has been an extremely useful piece of software. We could not have found the Higgs without it," says Kaushik De, a physics professor and director of UT Arlington's Center of Excellence for High Energy Physics. "It's been used by thousands of physicists around the world. We thought, 'Wouldn't it be nice if we repackaged it so others could use it too?' So, we proposed generalizing PanDA as a meta-application."
UT Arlington and Brookhaven developed PanDA for use by the ATLAS collaboration, a particle physics experiment at the Large Hadron Collider at CERN, the European Organization for Nuclear Research. ATLAS includes 3,000 physicists from UT Arlington and more than 170 other institutions, 40 of which are based in the U.S. In July, it was one of the groups that made headlines by announcing the discovery of a new particle that scientists said could be the Higgs boson, also known as the "God particle."
The computing hardware associated with ATLAS is located at 100 computing centers around the world that manage more than 50 petabytes, or 50 million gigabytes, of data. PanDA links the computing centers and allows scientists to efficiently analyze the tens of millions of particle collisions taking place at the Large Hadron Collider each day. In addition to De's work developing PanDA, UT Arlington is home to the ATLAS Southwest Tier 2 grid computing center.
"The collaboration between UT Arlington and Brookhaven National Laboratory was established a long time ago and it is a very successful story," says Alexei Klimentov, Physics Applications Software Group Leader at Brookhaven. "We are physicists, and we developed this software to be used by a high-energy physics experiment. But very soon we realized that it could be used in other areas, such as astro-particle physics, biomedicine, and others."
As an example of the possible applications, officials at the Alpha-Magnetic Spectrometer, or AMS, are considering adopting PanDA software, De says. AMS is an International Space Station experiment looking for dark matter in the universe.
In March, the Obama administration announced the "Big Data Research and Development Initiative," a $200 million investment in tools to handle huge volumes of digital data needed to spur science and engineering discoveries. At that announcement, White House officials highlighted the Energy Department's Office of Advanced Scientific Computing Research as key to that ongoing effort.
PanDA — which was created with funding from the U.S. National Science Foundation and the Department of Energy's Office of High Energy Physics — was cited as an example of successful technology already in place.
Klimentov is the principal investigator on the new computing grants. De and Gergely Zaruba, an associate professor of computer science and engineering at UT Arlington, are co-principal investigators for UT Arlington's portion, which totals $704,488. The grant to Brookhaven will total $997,000 over three years. Other key scientists include: Alexandre Vaniachine, of Argonne National Lab; and Torre Wenaus, Dantog Yu, and Sergei Panitkin, all of Brookhaven.
Klimentov said one of the major challenges of the new project would be to extend the grid-based PanDA system to a cloud-computing platform. He and De agree that work on the project would undoubtedly improve the capacity of the current PanDA system as it's used by ATLAS.
No entries found