Graph Partitioning in Scientific Simulations: Multilevel Schemes versus Space-Filling Curves

Schamberger, Stefan; Wierum, Jens-Michael

doi:10.1007/978-3-540-45145-7_14

Cited by 17 publications

(5 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, minimizing the edge cuts, that is, the size of partition boundary requiring information exchange, is not an objective of the space-filling curve method. On the other hand, graph partitioning packages, such as ParMETIS which uses a k-way multilevel heuristic approach usually provide satisfactory load balance and fewer edges cut compared with the space-filling curves method 28 . In this study, the ParMETIS is implemented as the partitioning tool.…”

Section: A Eulerian Domain Decompositionmentioning

confidence: 99%

Parallel Eulerian-Lagrangian Method with Adaptive Mesh Refinement for Moving Boundary Computation

Kuan

Sim

Hassan

et al. 2013

51st AIAA Aerospace Sciences Meeting Including the New Horizons Forum and Aerospace Exposition

View full text Add to dashboard Cite

In this study, we present a parallelized adaptive moving boundary computation technique on distributed memory multi-processor systems for multi-scale multiphase flow simulations. The solver utilizes the Eulerian-Lagrangian method to track moving (Lagrangian) interfaces explicitly on the stationary (Eulerian) Cartesian grid where the flow fields are computed. Since there exists strong data and task dependency between two distinct Eulerian and Lagrangian domain, we address the decomposition strategies for each domain separately. We then propose a trade-off approach aiming for parallel scalability. Spatial domain decomposition is adopted for both Eulerian and Lagrangian domains for load balancing and data locality to minimize inter-processor communication. In addition, a cellbased unstructured parallel adaptive mesh refinement (AMR) technique is implemented for the flexible local refinement with efficient grid usage and even-distributed computational workload among processors. The parallel performance is evaluated independently for the Cartesian grid solver and sub-procedures in cell-based unstructured AMR. The capability and the overall performance of the parallel adaptive Eulerian-Lagrangian method including moving boundary and topological change is demonstrated by modeling binary droplet collisions. With the aid of the present techniques, large scale moving boundary problems can be effectively computed.

show abstract

Section: A Eulerian Domain Decompositionmentioning

confidence: 99%

Parallel Eulerian-Lagrangian Method with Adaptive Mesh Refinement for Moving Boundary Computation

Kuan

Sim

Hassan

et al. 2013

51st AIAA Aerospace Sciences Meeting Including the New Horizons Forum and Aerospace Exposition

View full text Add to dashboard Cite

show abstract

“…Consequently, the load balance is satisfied but total edges cut, that is, the size of partition boundary requiring information exchange is not an objective of space-filling curve method. On the other hand, graph partitioning packages, such as METIS using k-way multilevel heuristic approach as partition algorithm usually provide satisfactory load balance and fewer edges cut compared with space-filling curves method 21 . Moreover, it is easy to implement and provide flexible weighting function to manipulate preferred partition scenario.…”

Section: A Domain Decomposition Of Eulerian Domainmentioning

confidence: 99%

Parallel, Adaptive Grid Computing of Multiphase Flows in Spacecraft Fuel Tanks

Kuan

Sim

Shyy

2012

50th AIAA Aerospace Sciences Meeting Including the New Horizons Forum and Aerospace Exposition

View full text Add to dashboard Cite

A parallel adaptive Eulerian-Lagrangian method is developed for effective large scale multiphase flow computation. The intricate complexity of parallelism for the Eulerian-Lagrangian method is discussed and spatial domain decomposition strategies are proposed. An efficient communication approach is presented with MPI standard communication and the performance of the parallel Eulerian-Lagrangian method is evaluated. The speedup on problems having million of cells shows super-linear behavior due to the effect of cache hit rate, and the parallel efficiency peaks at the trade-off point of the cache hit rate and the computation-to-communication ratio. A first-principle-based phase change model is introduced by computing the rate of phase change in each phase as well as across liquid/vapor interface. The self-pressurization in a liquid hydrogen fuel tank is simulated and it is shown that the phase change in a liquid phase has a dominant contribution comparing phase change on the liquid/vapor surface, and the rate decreases as saturation temperature increases. The effects of heat flux amount and liquid fuel fill-level are investigated, and the pressurization accelerates showing longer transition period with higher pressure rise rate as heat flux and fill level increases.

show abstract

“…It is a general problem when working with coordinates and occurs more heavily for example in space-filling-curve based partitionings [20]. The experiments made in [3] also reveal that the selection mechanism, though improved by preferring under-weighted partitions, does still not lead to sufficiently balanced domains.…”

Section: The Bubble Frameworkmentioning

confidence: 99%

Accelerating shape optimizing load balancing for parallel FEM simulations by algebraic multigrid

Meyerhenke¹,

Monien²,

Schamberger³

2006

Proceedings 20th IEEE International Parallel &Amp; Distributed Processing Symposium

View full text Add to dashboard Cite

show abstract

Graph Partitioning in Scientific Simulations: Multilevel Schemes versus Space-Filling Curves

Cited by 17 publications

References 12 publications

Parallel Eulerian-Lagrangian Method with Adaptive Mesh Refinement for Moving Boundary Computation

Parallel Eulerian-Lagrangian Method with Adaptive Mesh Refinement for Moving Boundary Computation

Parallel, Adaptive Grid Computing of Multiphase Flows in Spacecraft Fuel Tanks

Accelerating shape optimizing load balancing for parallel FEM simulations by algebraic multigrid

Contact Info

Product

Resources

About