Article
Mathematics, Applied
Santiago Badia, Alberto F. Martin, Eric Neiva, Francesc Verdugo
Summary: This work introduces an adaptive unfitted finite element scheme on locally adapted Cartesian forest-of-trees meshes, combining aggregated finite element method with parallel adaptive mesh refinement. By proposing a two-step algorithm to construct finite element space, linear constraints on nonconforming meshes are achieved. The resulting scheme demonstrates optimal mesh adaptation capability and parallel efficiency through numerical experiments.
SIAM JOURNAL ON SCIENTIFIC COMPUTING
(2021)
Review
Computer Science, Information Systems
Amanda Calatrava, Hernan Asorey, Jan Astalos, Alberto Azevedo, Francesco Benincasa, Ignacio Blanquer, Martin Bobak, Francisco Brasileiro, Laia Codo, Laura del Cano, Borja Esteban, Meritxell Ferret, Josef Handl, Tobias Kerzenmacher, Valentin Kozlov, Ales Krenek, Ricardo Martins, Manuel Pavesio, Antonio Juan Rubio-Montero, Juan Sanchez-Ferrero
Summary: Open Science is a paradigm that promotes transparent sharing and reuse of scientific data, procedures, tools, and results. The European Open Science Cloud (EOSC) initiative aims to provide a trusted and federated computing environment for executing scientific applications and sharing research data across borders and disciplines. The article analyzes the adaptations made in ten thematic services of EOSC-Synergy and identifies commonalities and best practices that can be applied to new services. The article also makes recommendations for integrating thematic services into the EOSC ecosystem, including authentication and authorization solutions, FAIR data and metadata preservation solutions, and cloud platform-agnostic resource management and workload management solutions.
COMPUTER SCIENCE REVIEW
(2023)
Article
Engineering, Multidisciplinary
Junqi Zhang, Ankit Ankit, Hauke Gravenkamp, Sascha Eisentraeger, Chongmin Song
Summary: This paper introduces a parallel explicit solver utilizing the advantages of balanced octree meshes and employing the scaled boundary finite element method (SBFEM). By pre-computing the stiffness and mass matrices of unique cell patterns, the hanging nodes problem in standard finite element analysis is avoided. The proposed scheme is implemented in a distributed computing environment and its performance is evaluated through various numerical benchmark examples.
COMPUTER METHODS IN APPLIED MECHANICS AND ENGINEERING
(2021)
Article
Computer Science, Information Systems
George Charitopoulos, Ioannis Papaefstathiou, Dionisios N. Pnevmatikatos
Summary: Executing complex scientific applications on Coarse Grain Reconfigurable Arrays (CGRAs) can improve execution time and energy consumption compared to software implementations or fully customized hardware solutions. This work explores the potential of application analysis methods in customized hardware solutions and presents the MC-Def, a Mixed-CGRA Definition Framework targeting a Mixed-CGRA architecture that leverages the advantages of CGRAs and FPGAs. Results show improved performance in executing various scientific applications with VHDL-created hardware implementations of CGRA cells.
Article
Mathematics, Applied
Johannes Holke, David Knapp, Carsten Burstedde
Summary: In this work, we discuss parallel algorithms for computing the ghost layer in computational, distributed memory, recursively adapted meshes. The ghost layer is necessary for executing parallel, element-based computer simulations. We present a design with modularity of algorithms and data structures, targeting adaptive, nonconforming forest-of-trees meshes with mixed element shapes like cubes, prisms, and tetrahedra.
SIAM JOURNAL ON SCIENTIFIC COMPUTING
(2021)
Article
Computer Science, Interdisciplinary Applications
David Herrero-Perez, Sebastian Gines Pico-Vicente
Summary: This work presents an efficient parallel geometric multigrid (GMG) implementation for preconditioning Krylov subspace methods solving differential equations using non-conforming meshes for discretization. The approach calculates the restriction and interpolation operators for grid transferring between the non-conforming hierarchical meshes. Using non-Cartesian grids in topology optimization, it reduces the mesh size by discretizing only the design domain. The performance of the proposed method is evaluated using topology optimization problems, showing its computational advantages.
STRUCTURAL AND MULTIDISCIPLINARY OPTIMIZATION
(2023)
Article
Engineering, Industrial
Chao Dang, Pengfei Wei, Matthias G. R. Faes, Marcos A. Valdebenito, Michael Beer
Summary: This study proposes a new method called "Parallel Adaptive Bayesian Quadrature" (PABQ) for quantifying and reducing numerical uncertainty in reliability analysis. The method uses an importance ball sampling technique and a multi-point selection criterion to effectively assess small failure probabilities with a minimum number of iterations, taking advantage of parallel computing.
RELIABILITY ENGINEERING & SYSTEM SAFETY
(2022)
Article
Engineering, Multidisciplinary
Haoju Lin, Hui Liu, Peng Wei
Summary: This work proposes a parallel parameterized level set topology optimization framework for large-scale structures with unstructured meshes, which addresses the adaptability to structures with arbitrary geometries and complex boundary conditions. The framework combines distributed memory parallel computing technology and parameterized level set topology optimization using unstructured meshes. Several means, including shape functions, directed acyclic graph data structure, direct imposition of passive domain and boundary conditions on geometry entities, and multiple averaging filter, are utilized. Computing tests demonstrate the stability, efficiency, scalability, and potential for discovering new structure styles of the framework.
COMPUTER METHODS IN APPLIED MECHANICS AND ENGINEERING
(2022)
Article
Computer Science, Hardware & Architecture
Hao Fan, Song Wu, Xinyu Zhao, Zhenjiang Xie, Sheng Di, Jiang Xiao, Chen Yu, Hai Jin
Summary: This paper proposes an Adaptive Time-slice Control (ATC) mechanism that can effectively optimize the performance of parallel applications by shortening time-slices during communication phases, prolonging time-slices during computation phases, and setting a uniform time-slice for non-parallel applications. Experimental results demonstrate that ATC achieves significant performance gains for running parallel applications, outperforming state-of-the-art solutions.
IEEE TRANSACTIONS ON COMPUTERS
(2021)
Article
Computer Science, Software Engineering
Maurin Lopez, Suzanne M. Shontz, Weizhang Huang
Summary: In this paper, a parallel variational mesh quality improvement algorithm for distributed memory machines is proposed. The algorithm adapts the mesh by using the Moving Mesh PDE method and solving a system of ordinary differential equations (ODEs) to determine the movement of interior mesh nodes. Strong and weak scaling experiments show excellent results for meshes with up to 160M elements on up to 128 cores.
COMPUTER-AIDED DESIGN
(2022)
Article
Mathematics, Applied
Kumar Saurabh, Boshun Gao, Milinda Fernando, Songzhe Xu, Makrand A. Khanwale, Biswajit Khara, Ming-Chen Hsu, Adarsh Krishnamurthy, Hari Sundar, Baskar Ganapathysubramanian
Summary: A variant of the immersed boundary method integrated with octree meshes is introduced for highly efficient and accurate Large Eddy Simulations (LES) of flows around complex geometries. Scalability of the proposed method up to theta(32K) processors is demonstrated through rapid in-out tests, adaptive quadrature for accurate force evaluation, and tensorized evaluation during matrix assembly. The method is successfully showcased in accurately computing the drag coefficient of a sphere across Reynolds numbers 1 to 10^6 and simulating flow features across a semi-truck for investigating the effect of platooning on efficiency.
COMPUTERS & MATHEMATICS WITH APPLICATIONS
(2021)
Article
Mathematics, Applied
M. H. M. Sulman, T. B. Nguyen, R. D. Haynes, Weizhang Huang
Summary: This paper presents a fast method for adaptive moving mesh generation in multi-dimensions using a domain decomposition parabolic Monge-Ampere approach. By computing an adaptive mesh on each subdomain and mapping the results to the solution of the L-2 optimal mass transfer problem, significant reduction in computational time and increased efficiency are achieved.
COMPUTERS & MATHEMATICS WITH APPLICATIONS
(2021)
Article
Computer Science, Software Engineering
Maciej Malawski, Bartosz Balis
Summary: This article provides insights into the current landscape of serverless computing for scientific applications. It discusses the advantages and challenges of serverless computing in scientific applications and proposes a science-oriented architecture for a serverless computing framework based on existing designs. The article also presents an outlook on current trends and future directions.
IEEE INTERNET COMPUTING
(2022)
Article
Computer Science, Information Systems
Yunjae Lee, Heon Y. Yeom, Hyuck Han
Summary: This article proposes a new page replacement scheme called Adaptive Page Replacement (APR) for looping access patterns in scientific applications. The experimental results show that APR outperforms prior schemes in terms of performance.
CLUSTER COMPUTING-THE JOURNAL OF NETWORKS SOFTWARE TOOLS AND APPLICATIONS
(2023)
Article
Computer Science, Interdisciplinary Applications
Polykarpos Thomadakis, Christos Tsolakis, Nikos Chrisochoides
Summary: The PREMA framework introduces a new design for large-scale applications, enabling communication, remote method invocations, and object migrations for load balancing and improved performance. It includes multi-threading support and monitoring interfaces for system load, ensuring task correctness and concurrent migrations.
ENGINEERING WITH COMPUTERS
(2022)