A hybrid parallel Delaunay image-to-mesh conversion algorithm scalable on distributed-memory clusters

In this paper, we present a scalable three dimensional parallel Delaunay image-to-mesh conversion algorithm. A nested master–worker model is used to simultaneously explore process- and thread-level parallelization. The mesh generation includes two stages: coarse and fine meshing. First, a coarse mes...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Computer aided design Ročník 103; s. 34 - 46
Hlavní autoři: Feng, Daming, Chernikov, Andrey N., Chrisochoides, Nikos P.
Médium: Journal Article
Jazyk:angličtina
Vydáno: Amsterdam Elsevier Ltd 01.10.2018
Elsevier BV
Témata:
3-D technology
Algorithms
Cluster analysis
Clusters
Computer memory
Concurrency
Conversion
Distributed memory
Distributed shared memory
Finite element method
Grid refinement (mathematics)
Hybrid programming
Image-to-mesh conversion
Mesh generation
Parallel mesh generation
Parallel processing
Three dimensional models
Two-level parallelization
Parallel mesh generation
Two-level parallelization
Hybrid programming
Image-to-mesh conversion
ISSN:0010-4485, 1879-2685
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:In this paper, we present a scalable three dimensional parallel Delaunay image-to-mesh conversion algorithm. A nested master–worker model is used to simultaneously explore process- and thread-level parallelization. The mesh generation includes two stages: coarse and fine meshing. First, a coarse mesh is constructed in parallel by the threads of the master process. Then the coarse mesh is partitioned. Finally, the fine mesh refinement procedure is executed until all the elements in the mesh satisfy the quality and fidelity criteria. The communication and computation are separated during the fine mesh refinement procedure. The master thread of each process that initializes the MPI environment is in charge of the inter-node MPI communication for data (submesh) movement while the worker threads of each process are responsible for the local mesh refinement within the node. We conducted a set of experiments to test the performance of the algorithm on distributed memory clusters and observed that the granularity of coarse level data decomposition, which affects the coarse level concurrency, has a significant influence on the performance of the algorithm. With the proper value of granularity, the algorithm is scalable to 45 distributed memory compute nodes (900 cores). •The Hybrid MPI+Threads method is a hybrid programming image-to-mesh conversion algorithm.•The Hybrid MPI+Threads method creates meshes with quality and fidelity guarantees.•The Hybrid MPI+Threads method supports parallel non-uniform mesh generation for multi-material objects.•The Hybrid MPI+Threads method is a scalable parallel meshing algorithm on distributed multi-core node clusters.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0010-4485
1879-2685
DOI:10.1016/j.cad.2017.11.006