The Knowledge Discovery Toolbox (KDT) enables domain experts to perform complex analyses of huge datasets on supercomputers using a high-level language without grappling with the difficulties of writing parallel code, calling parallel libraries, or becoming a graph expert. KDT provides a flexible Python interface to a small set of high-level graph operations; composing a few of these operations is often sufficient for a specific analysis. Scalability and performance are delivered by linking to a state-of-the-art back-end compute engine that scales from laptops to large HPC clusters. KDT delivers very competitive performance from a generalpurpose, reusable library for graphs on the order of 10 billion edges and greater. We demonstrate speedup of 1 and 2 orders of magnitude over PBGL and Pegasus, respectively, on some tasks. Examples from simple use cases and key graphanalytic benchmarks illustrate the productivity and performance realized by KDT users. Semantic graph abstractions provide both flexibility and high performance for real-world use cases. Graph-algorithm researchers benefit from the ability to develop algorithms quickly using KDT's graph and underlying matrix abstractions for distributed memory. KDT is available as open-source code to foster experimentation.
SUMMARYAlgebraic multigrid (AMG) is a powerful linear solver with attractive parallel properties. A parallel AMG method depends on efficient, parallel implementations of the coarse-grid selection algorithms and the restriction and prolongation operator construction algorithms. In the effort to effectively and quickly select the coarse grid, a number of parallel coarsening algorithms have been developed. This paper examines the behaviour of these algorithms in depth by studying the results of several numerical experiments. In addition, new parallel coarse-grid selection algorithms are introduced and tested.
SUMMARYAlgebraic multigrid (AMG) is an e cient algorithm for solving certain types of large, sparse linear systems. For solving very large problems with AMG it becomes necessary to use parallel algorithms. Coarse grid selection algorithms such as CLJP were created to parallelize the setup phase of AMG. For some problems, such as those discretized on structured meshes, CLJP tends to select coarse grids with more nodes than alternative coarsening algorithms. In this paper, the cause for the selection of too many coarse nodes by CLJP is examined, and a new technique which lowers the operator complexities generated by CLJP is introduced. To validate the new method, the modiÿed CLJP is compared to other coarsening algorithms for large-scale problems.
NREL's early demonstration of "systems biochemistry" links traditional and emerging biochemistry with modern computing and the mathematics of complex systems. National Renewable Energy Laboratory (NREL) scientists, supported by the Department of Energy (DOE) Scientific Discovery through Advanced Computing (SciDAC) Program, have assembled and simulated a model of key eukaryotic carbon metabolism that intends to move biochemical simulations into new realms of chemical fidelity. Achieving truly predictive metabolic engineering requires moving beyond small canonical models to realistic, large-scale, dynamic systems that describe cellular biochemical networks. NREL researchers assembled a multi-compartmental metabolic model comprising 149 compounds, 65 enzymes, 688 parameters, and 114 reactions that describe the key pathways of glycolysis, tricarboxylic acid cycle, reductive pentose phosphate, oxidative phosphorylation, starch degradation, and fermentation, including the full complexity of kinetic equations. Ongoing high-performance simulation and analysis of this model using NREL's super-computing capability-including a parallel software simulation and optimization suite developed at NREL under the same DOE program-will yield a new understanding of how biochemical systems respond to physical changes in conditions. NREL envisions this work to be an early demonstration of systems biochemistry that can join together traditional and emerging biochemistry with modern computing and the mathematics of complex systems. As our understanding of cellular metabolism advances through partnership among experiment, computation, and theory, cellular design will become as routine and reliable as automobile design is today. Achievement New simulation model coupled with NREL's supercomputer advances biochemical modeling reliability. Key Result NREL created the largest and most detailed model to date related to central carbon flow. Potential Impact This research will lead to a greater understanding of biochemical dynamics with applications to metabolic engineering and production of biofuels. NREL is a national laboratory of the U.S. Department of Energy, Office of Energy Efficiency and Renewable Energy, operated by the Alliance for Sustainable Energy, LLC. 1617 Cole Boulevard | Golden, Colorado 80401-3305 | 303-275-3000 | www.nrel.gov NREL/FS-2C00-50850 • February 2011 Printed with a renewable-source ink on paper containing at least 50% wastepaper, including 10% post consumer waste.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.