Follow us on Instagram
Try our latest crossword

PPPL researchers receive grant to use supercomputer

Researchers from the Princeton Plasma Physics Laboratory won a grant to use 80 million processor hours on the nation’s fastest supercomputer to conduct simulations of various phenomena.


The researchers include University Astrophysical Sciences Professor Amitava Bhattacharjee GS ’81, PPPL physicist William Fox, University research scholar Yi-Min Huang and PPPL graduate student Jonathan Ng.

In addition to the PPPL team, University Astrophysical Sciences Professor James Stone also won 47 million processor hours to pursue a project titled "Magnetohydrodynamic Models of Accretion Including Radiation Transport," and Geology Professor Jeroen Tromp GS ’90 GS '92 received 80 million processor hours to pursue a project titled "Global Adjoint Tomography."

Bhattacharjee said that he plans to use the supercomputer to conduct simulations of high-energy-density plasmas. He explained that plasma is the fourth state of matter consisting of freely-moving electrons and protons and said that some of these high-energy-density plasmas, created in the laboratory, have the unique property of having self-generating large magnetic fields.

“These magnetic fields are a major source of energy,” Bhattacharjee said.

Fox added that since plasmas possess electrical and magnetic properties, it is possible to control and confine the behavior of the plasmas by modifying the surrounding electrical and magnetic fields. He noted that such plasmas comprise some large astrophysical objects. He added that such plasmas are very important for controlled nuclear fusion research since they can generate large amounts of energy from seawater that can be used as an alternative source of power.

Bhattacharjee explained that these plasmas have been created in the laboratory, specifically at the University of Rochester Laboratory for Laser Energetics and the National Ignition Facility at Lawrence Livermore National Laboratory in California.Researchers have studied the properties of these plasmas in the laboratory, and what they might teach us about plasma processes in larger astrophysical objects.Bhattacharjee said that experimental measurements and high-performance computer simulations are necessary in understanding the plasma processes.


“Our goal in the experiment is to understand the processes that might go on in astrophysical objects,” Bhattacharjee said. “Our ability to simulate them on the computer will give us confidence in understanding the laboratory experiments and how these processes will play out in the context of the universe.”

“Specifically, we’re trying to understand what is the environment in the plasma around the Earth, and how does the Earth interact with the wind of plasma blowing from the sun?” Fox said.

He noted that solar flares are very relevant today, since a large solar flare has the potential to knock out communication satellites and GPS satellites. Understanding this phenomenon can help researchers prevent disruptions to Earth’s equipment.

Another research interest involves finding out where the highest energy particles in the universe, called cosmic rays, get their energy from. Fox said that the rays have energies billions of times larger than particles used in accelerators on Earth.

Get the best of ‘the Prince’ delivered straight to your inbox. Subscribe now »

In order to use the supercomputer, researchers must be allotted time in the form of processor hours. Bhattacharjee’s group won 35 million processor hours from the INCITE (Innovative and Novel Impact on Computational Theory and Experiment) program and 45 million processor hours from the Advanced Scientific Computing Research (ASCR) Advanced Scientific Computing Research Leadership Computing Challenge program. Both of these grants are awarded by the United States Department of Energy’s Office of Science.

A processor hour is an hour when a CPU of the computer takes to process a single program.

INCITE Program Manager Judy Hill said they receive applications from professors, government researchers and industry researchers from around the world to obtain processor time on the supercomputer. She noted that the purpose of using the computer does not have to relate to the Department of Energy’s mission.

“The competition to use the computer is targeted at any open science problem that can efficiently use the computing resources that we have,” she said.

The supercomputer that will be used by Bhattacharjee’s group is the Titan Cray XK7 located in Oak Ridge National Laboratory in Tennessee. The Titan computer is the nation’s fastest supercomputer.

Hill noted that the proposals they receive are judged primarily on whether they will have the highest scientific impact in their fields and secondarily on whether they will efficiently use the computer. She said that it is very difficult to obtain time in the INCITE program, since only 55 to 60 projects are granted access each year.

“We want projects that utilize a substantial fraction of the machine,” Hill said.

All prospective users must submit a 15-page proposal to the INCITE program that describes the significance of the research, the importance it has to the general public and the specific simulations the researchers plan to run. Hill noted that researchers can ask for discretionary time, on the order of a few million processor hours, to demonstrate their simulations.

One process Bhattacharjee and his collaborators hope to simulate on the computer is magnetic reconnection, which is when magnetic field lines break and reconnect, resulting in a liberation of energy in the form of highly energetic particles.

“The magnetic field lines can be thought of as rubber strings,” he said. “They break and reconnect and release a lot of energy that shows up in particles and heat.”

Bhattacharjee noted that researchers at the PPPL have made important discoveries in this field in the last couple of years. He explained that this research explains new phenomena observed in the universe.

Other processes Bhattacharjee hopes to run include the “Biermann battery,” which is the creation of a large magnetic field with a laser in a material that has little or almost no magnetic field, and Weibel instabilities. He explained that he and his team require large amounts of time because he plans to run the processes in both two and three dimensions and that the objects being simulated are very large.

“Using up all of the 80 million processor hours will not be hard,” Bhattacharjee said. “Some of it will be spent on massive or heroic simulations, but building up to that will be many small simulations that will make sure everything is working right.”