Building Effective Parallel Unstructured Adaptive Simulations by In-memory Integration of Existing Software Components

There is an increasing demand for mesh-based PDE simulations to provide reliable solutions for complex problems. A key step in ensuring simulation reliability is the application of unstructured adaptive meshing driven by error estimation procedures. Typically, interfacing of unstructured adaptive procedures with existing analysis components is done through files. File I/O is currently the critical bottleneck in parallel adaptive analysis, even when advanced parallel I/O methods are used. This paper presents an approach to eliminate this bottleneck for unstructured parallel adaptive simulations using in-memory integration of existing analysis components with adaptive meshing components. First, the design of APIs that transfer information between simulation components is provided, followed by an approach to address the technical challenges faced. Then, the design of the interoperable mesh adaptation and analysis components is given. The paper concludes with performance results obtained to date.

[1]  Mark S. Shephard,et al.  3D anisotropic mesh adaptation by mesh modification , 2005 .

[2]  Joe Walsh,et al.  A comparison of techniques for geometry access related to mesh generation , 2004, Engineering with Computers.

[3]  Onkar Sahni,et al.  Cardiovascular flow simulation at extreme scale , 2010 .

[4]  Q A Koziol,et al.  Smart Libraries: Best SQE Practices for Libraries with an Emphasis on Scientific Computing , 2004 .

[5]  Carsten Burstedde,et al.  p4est: Scalable Algorithms for Parallel Adaptive Mesh Refinement on Forests of Octrees , 2011, SIAM J. Sci. Comput..

[6]  Robert M. O'Bara,et al.  Attribute Management System for Engineering Analysis , 2002, Engineering with Computers.

[7]  Mark S. Shephard,et al.  Efficient distributed mesh data structure for parallel automated adaptive analysis , 2006, Engineering with Computers.

[8]  D. Drew,et al.  PARALLEL ADAPTIVE SIMULATION OF A PLUNGING LIQUID JET ∗ Dedicated to Professor James Glimm on the occasion of his 75th birthday , 2010 .

[9]  Timothy J. Tautges,et al.  An Interoperable, Data-Structure-Neutral Component for Mesh Query and Manipulation , 2010, ACM Trans. Math. Softw..

[10]  Onkar Sahni,et al.  Scalable implicit finite element solver for massively parallel processing with demonstration to 160K cores , 2009, Proceedings of the Conference on High Performance Computing Networking, Storage and Analysis.

[11]  Onkar Sahni,et al.  A parallel adaptive mesh method for the numerical simulation of multiphase flows , 2013 .