Back to Table of contents

Primeur weekly 2017-12-18

Exascale supercomputing

Mont-Blanc 2020 project will pave the way to a European scalable, modular and power efficient High Performance Computing processor ...

A supercomputer will discover our future medicines ...

Quantum computing

Jülich Supercomputing Centre to achieve world record: Quantum computer with 46 qubits simulated ...

DENSO and Toyota Tsusho to conduct a test applying a quantum computer to analyze IoT data with a commercial application ...

IBM announces collaboration with Fortune 500 companies, academic institutions and national research labs to accelerate quantum computing ...

Complete design of a silicon quantum computer chip unveiled ...

Error-free into the quantum computer age ...

Quantum memory with record-breaking capacity based on laser-cooled atoms ...

Physicists from Konstanz, Princeton and Maryland create a stable quantum gate as a basic element for the quantum computer ...

Physicists say rare earth metals could help quantum computers communicate ...

Focus on Europe

Lenovo and Intel to deliver powerful, energy-efficient SuperMUC-NG, next generation supercomputer at Leibniz Supercomputing Centre ...

North Rhine-Westphalia supports expansion of German national supercomputer infrastructure ...

ISC 2018 and PASC18 Conferences to announce partnership ...

Hardware

Intersect360 Research invites to participate in tenth HPC budget map survey ...

BP supercomputer now world's most powerful for commercial research ...

MareNostrum 4, chosen as the most beautiful data centre in the world ...

NCI welcomes $70 million investment in HPC capability ...

Applications

Physicists win supercomputing time to study fusion and the cosmos ...

Artificial intelligence helps accelerate progress toward efficient fusion reactions ...

Johns Hopkins scientists chart how brain signals connect to neurons ...

Artificial Intelligence and supercomputers to help alleviate urban traffic problems ...

ESnet's Petascale DTN project speeds up data transfers between leading HPC centres ...

A computer system needing less time and memory to simulate mechanical systems ...

Drug discovery could accelerate hugely with Machine Learning ...

The Cloud

Murex to offer MX.3 risk, trading and post-trade solutions on the AWS Cloud ...

Fujitsu develops WAN acceleration technology utilizing FPGA accelerators ...

Meituan.com selects Mellanox interconnect solutions to accelerate its artificial intelligence, Big Data and Cloud data centres ...

Mellanox interconnect solutions accelerate Tencent Cloud high-performance computing and artificial intelligence infrastructure ...

Physicists win supercomputing time to study fusion and the cosmos


Physicists Amitava Bhattacharjee and C.S. Chang. Photo Credit: Elle Starkman/PPPL Office of Communications.)
6 Dec 2017 Plainsboro - More than 210 million core hours on two of the most powerful supercomputers in the nation have been won by two teams led by researchers at the U.S. Department of Energy's (DOE) Princeton Plasma Physics Laboratory (PPPL). The highly competitive awards from the DOE Office of Science's INCITE - Innovative and Novel Impact on Computational Theory and Experiment - programme will accelerate the development of nuclear fusion as a clean and abundant source of energy for generating electricity and will advance understanding of the high-energy-density (HED) plasmas found in stars and other astrophysical objects.

A single core hour represents the use of one computer core, or processor, for one hour. A laptop computer with only one processor would take some 24,000 years to run 210 million core hours.

"These awards are extremely important and beneficial", stated Michael Zarnstorff, deputy director for research at PPPL. "They give us access to leadership-class highest-performance computers for highly complex calculations. This is key for advancing our theoretical modeling and understanding." Leadership-class computing systems are high-end computers that are among the most advanced in the world for solving scientific and engineering problems.

The allocations include more than 160 million core hours for physicist C.S. Chang and his team, marking the first year of a renewable three-year award. The first-year hours are distributed over two machines: 100-million core hours on Titan, the most powerful U.S supercomputer, which can perform some 27 quadrillion (1015) calculations per second at the Oak Ridge Leadership Computing Facility (OLCF); and 61.5 million core hours on Theta, which completes some 10 quadrillion calculations a second at the Argonne Leadership Computing Facility (ALCF). Both sites are DOE Office of Science User Facilities.

Also received are 50 million core hours on Titan for Amitava Bhattacharjee, head of the Theory Department at PPPL, and William Fox and their team to study HED plasmas produced by lasers.

C.S. Chang's group consists of colleagues at PPPL and other institutions and will use the time to run the XGC code developed by PPPL and nationwide partners. The team is exploring the dazzlingly complex edge of fusion plasmas with Chang as lead principal investigator of the partnership center for High-fidelity Boundary Plasma Simulation - a programme supported by the DOE Office of Science's Scientific Discovery through Advanced Computing (SciDAC). The edge is critical to the performance of plasma that fuels fusion reactions.

Fusion is the fusing of light elements that most stars use to generate massive amounts of energy - and that scientists are trying to replicate on Earth for a virtually inexhaustible supply of energy. Plasma - the fourth state of matter that makes up nearly all the visible universe - is the fuel they would use to create fusion reactions.

The XGC code will perform double-duty to investigate developments at the edge of hot, charged fusion plasma. The programme will simulate the transition from low- to high-confinement of the edge of fusion plasmas contained inside magnetic fields in doughnut-shaped fusion devices called tokamaks. Also simulated will be the width of the heat load that will strike the divertor, the component of the tokamak that will expel waste heat and particles from future fusion reactors based on magnetic confinement such as ITER, the international tokamak under construction in France to demonstrate the practicality of fusion power.

The simulations will build on knowledge that C.S. Chang has achieved in the previous-cycle SciDAC project. "We're just getting started", C.S. Chang stated. "In the new SciDAC project we need to understand the different types of transition that are thought to occur in the plasma, and the physics behind the width of the heat load, which can damage the divertor in future facilities such as ITER if the load is too narrow and concentrated."

The Bhattacharjee-Fox award, the second and final part of a two-year project, will advance progress in the team's understanding of the dynamics of magnetic fields in HED plasmas. "The simulations will be immensely beneficial in designing and understanding the results of experiments carried out at the University of Rochester and the National Ignition Facility at Lawrence Livermore National Laboratory", Amitava Bhattacharjee stated.

The project explores the magnetic reconnection and shocks that occur in HED plasmas, producing enormous energy in processes such as solar flares, cosmic rays and geomagnetic storms. Magnetic reconnection takes place when the magnetic field lines in plasma converge and break apart, converting magnetic energy into explosive particle energy. Shocks appear when the flows in the plasma exceed the speed of sound, and are a powerful process for accelerating charged particles.

To study the process, the team fires high-power lasers at tiny spots of foil, creating plasma bubbles with magnetic fields that collide to form shocks and come together to create reconnection. "Our group has recently made important progress on the properties of shocks and novel mechanisms of magnetic reconnection in laser-driven HED plasmas", Amitava Bhattacharjee stated. "This could not be done without INCITE support."

Source: Princeton Plasma Physics Laboratory - PPPL

Back to Table of contents

Primeur weekly 2017-12-18

Exascale supercomputing

Mont-Blanc 2020 project will pave the way to a European scalable, modular and power efficient High Performance Computing processor ...

A supercomputer will discover our future medicines ...

Quantum computing

Jülich Supercomputing Centre to achieve world record: Quantum computer with 46 qubits simulated ...

DENSO and Toyota Tsusho to conduct a test applying a quantum computer to analyze IoT data with a commercial application ...

IBM announces collaboration with Fortune 500 companies, academic institutions and national research labs to accelerate quantum computing ...

Complete design of a silicon quantum computer chip unveiled ...

Error-free into the quantum computer age ...

Quantum memory with record-breaking capacity based on laser-cooled atoms ...

Physicists from Konstanz, Princeton and Maryland create a stable quantum gate as a basic element for the quantum computer ...

Physicists say rare earth metals could help quantum computers communicate ...

Focus on Europe

Lenovo and Intel to deliver powerful, energy-efficient SuperMUC-NG, next generation supercomputer at Leibniz Supercomputing Centre ...

North Rhine-Westphalia supports expansion of German national supercomputer infrastructure ...

ISC 2018 and PASC18 Conferences to announce partnership ...

Hardware

Intersect360 Research invites to participate in tenth HPC budget map survey ...

BP supercomputer now world's most powerful for commercial research ...

MareNostrum 4, chosen as the most beautiful data centre in the world ...

NCI welcomes $70 million investment in HPC capability ...

Applications

Physicists win supercomputing time to study fusion and the cosmos ...

Artificial intelligence helps accelerate progress toward efficient fusion reactions ...

Johns Hopkins scientists chart how brain signals connect to neurons ...

Artificial Intelligence and supercomputers to help alleviate urban traffic problems ...

ESnet's Petascale DTN project speeds up data transfers between leading HPC centres ...

A computer system needing less time and memory to simulate mechanical systems ...

Drug discovery could accelerate hugely with Machine Learning ...

The Cloud

Murex to offer MX.3 risk, trading and post-trade solutions on the AWS Cloud ...

Fujitsu develops WAN acceleration technology utilizing FPGA accelerators ...

Meituan.com selects Mellanox interconnect solutions to accelerate its artificial intelligence, Big Data and Cloud data centres ...

Mellanox interconnect solutions accelerate Tencent Cloud high-performance computing and artificial intelligence infrastructure ...