Publication:
Implementation of 2D Domain Decomposition in the UCAN Gyrokinetic Particle-in-Cell Code and Resulting Performance of UCAN2

Simple item page
dc.affiliation.dptoUC3M. Departamento de Ingeniería Telemáticaes
dc.affiliation.grupoinvUC3M. Grupo de Investigación: Aplicaciones y Servicios Telemáticos (GAST)es
dc.contributor.authorLeboeuf, J.-N.
dc.contributor.authorDecyk, V. K.
dc.contributor.authorNewman, D.E.
dc.contributor.authorSánchez Fernández, Luis Raúl
dc.contributor.funderMinisterio de Ciencia e Innovación (España)es
dc.date.accessioned2022-09-13T08:03:45Z
dc.date.available2022-09-13T08:03:45Z
dc.date.issued2016-01
dc.description.abstractThe massively parallel, nonlinear, three-dimensional (3D), toroidal, electrostatic, gyrokinetic, particle-in-cell (PIC), Cartesian geometry UCAN code, with particle ions and adiabatic electrons, has been successfully exercised to identify non-diffusive transport characteristics in present day tokamak discharges. The limitation in applying UCAN to larger scale discharges is the 1D domain decomposition in the toroidal (or z-) direction for massively parallel implementation using MPI which has restricted the calculations to a few hundred ion Larmor radii or gyroradii per plasma minor radius. To exceed these sizes, we have implemented 2D domain decomposition in UCAN with the addition of the y-direction to the processor mix. This has been facilitated by use of relevant components in the P2LIB library of field and particle management routines developed for UCLA's UPIC Framework of conventional PIC codes. The gyro-averaging specific to gyrokinetic codes is simplified by the use of replicated arrays for efficient charge accumulation and force deposition. The 2D domain-decomposed UCAN2 code reproduces the original 1D domain nonlinear results within round-off. Benchmarks of UCAN2 on the Cray XC30 Edison at NERSC demonstrate ideal scaling when problem size is increased along with processor number up to the largest power of 2 available, namely 131,072 processors. These particle weak scaling benchmarks also indicate that the 1 nanosecond per particle per time step and 1 TFlops barriers are easily broken by UCAN2 with 1 billion particles or more and 2000 or more processors.en
dc.description.sponsorshipThis work was supported in part in the USA by Grant No. DE-FG02-04ER54741 to the University of Alaska, Fairbanks, AK, from the Office of Fusion Energy Sciences, Office of Science, United States Department of Energy. It was also supported in part at Universidad Carlos III, Madrid, Spain, by Spanish National Project No. ENE2009-12213-C03-03. This research used resources of the National Energy Research Scientific Computing Center (NERSC), which is supported by the Office of Science of the U.S. Department of Energy under Contract No. DE-AC02-05CH11231. It also took advantage of resources at the Barcelona Supercomputing Center (BSC), Centro Nacional de Supercomputación, Barcelona, Spain. One of us (Leboeuf) would particularly like to thank David Vicente from BSC and Zhengji Zhao from NERSC for their help in the porting, debugging, and optimization of UCAN2 on the mainframes at their respective centers.en
dc.format.extent20
dc.identifier.bibliographicCitationLeboeuf, J., Decyk, V., Newman, D., & Sanchez, R. (2016). Implementation of 2D Domain Decomposition in the UCAN Gyrokinetic Particle-in-Cell Code and Resulting Performance of UCAN2. Communications in Computational Physics, 19(1), 205-225.en
dc.identifier.doihttps://doi.org/10.4208/cicp.070115.030715a
dc.identifier.issn1815-2406
dc.identifier.publicationfirstpage205
dc.identifier.publicationissue1
dc.identifier.publicationlastpage225
dc.identifier.publicationtitleCommunications in Computational Physicsen
dc.identifier.publicationvolume19
dc.identifier.urihttps://hdl.handle.net/10016/35674
dc.identifier.uxxiAR/0000017681
dc.language.isoengen
dc.publisherGlobal-Science Pressen
dc.relation.projectIDGobierno de España. ENE2009-12213-C03-03es
dc.relation.publisherversionhttps://www.cambridge.org/core/journals/communications-in-computational-physics/article/abs/implementation-of-2d-domain-decomposition-in-the-ucan-gyrokinetic-particleincell-code-and-resulting-performance-of-ucan2/387D6EA7A8EB4365B42CB21831647463#articleen
dc.rights© Global-Science Press 2016en
dc.rights.accessRightsopen accessen
dc.subject.ecienciaFísicaes
dc.subject.otherGyrokinetic simulationen
dc.subject.otherParticle-in-cell methoden
dc.subject.otherMassive parallelizationen
dc.subject.otherGradient-driven turbulenceen
dc.subject.otherSimulation-modelen
dc.subject.otherPic codeen
dc.subject.otherTransporten
dc.subject.otherPlasmasen
dc.subject.otherFrameworken
dc.subject.otherFlowsen
dc.titleImplementation of 2D Domain Decomposition in the UCAN Gyrokinetic Particle-in-Cell Code and Resulting Performance of UCAN2en
dc.typeresearch article*
dc.type.hasVersionAM*
dspace.entity.typePublication
Files
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
Implementation_CCP_2016.ps.pdf
Size:
1.08 MB
Format:
Adobe Portable Document Format
Download
Collections
DIT - GAST - Artículos de Revistas