2,433 research outputs found
A discontinuous finite element method for solving a multi-well problem
No full textGobbert, Matthias K.. (1998). A discontinuous finite element method for solving a multi-well problem. Retrieved from the University Digital Conservancy, https://hdl.handle.net/11299/3185
Design of an effective numerical method for a reaction-diffusion system with internal and transient layers
No full textSoane, Ana Maria; Gobbert, Matthias K.; Seidman, Thomas I.. (2004). Design of an effective numerical method for a reaction-diffusion system with internal and transient layers. Retrieved from the University Digital Conservancy, https://hdl.handle.net/11299/4065
Improvement of Student Learning by Automatic Grading Systems in Linear Algebra
No full textSeventh Annual Provost's Teaching and Learning Symposium April 21, 2023https://userpages.umbc.edu/~gobbert/papers/Gobbert_TeachingSymposium230421.pd
Performance comparison of Intel Xeon Phi Knights Landing
No full textThe Intel Xeon Phi is a many-core processor with a theoretical peak performance of over 3 TFLOP/s of double precision. We contrast the performance of the second-generation Intel Xeon Phi, code-named Knights Landing (KNL), to the first-generation Intel Xeon Phi, codenamed Knights Corner (KNC), as well as to a node with two multi-core CPUs as baseline reference. The test code solves the classical elliptic test problem of the Poisson equation whose performance is prototypical for the computational kernel in many numerical methods for partial differential equations. The results show that the KNL can perform approximately four times faster than the KNC or than two CPUs, provided the problem fits into the 16 GB of on-chip MCDRAM memory of the KNL. The studies also confirm the nominal five times faster speed of the new high-performance MCDRAM memory in the KNL compared to the DDR4 memory of the node. We demonstrate the ease of porting code to the KNL by focusing on performance
that was achieved by only re-compiling hybrid MPI+OpenMP code with a KNL flag.These results were obtained as part of the REU Site: Interdisciplinary Program in High Performance Computing (hpcreu.umbc.edu) in the Department of Mathematics and Statistics at the University of Maryland, Baltimore County (UMBC) in Summer 2016. This program is funded by the National Science Foundation (NSF), the National Security Agency (NSA), and the Department of Defense (DOD), with additional support from UMBC, the Department of Mathematics and Statistics, the Center for Interdisciplinary Research and Consulting (CIRC), and the UMBC High Performance Computing Facility (HPCF). HPCF is supported by the U.S. National Science Foundation through the MRI program (grant nos. CNS–0821258 and CNS–1228778) and the SCREMS program (grant no. DMS–0821311), with additional substantial support from UMBC. Co-author Ishmail Jabbie
was supported, in part, by the UMBC National Security Agency (NSA) Scholars Program through a contract with the NSA. The authors thank both our team’s graduate assistant Jonathan Graf and faculty mentor Dr. Matthias K. Gobbert for their support throughout the program and beyond. Graduate assistant Jonathan Graf was supported by UMBC. The authors would like to thank the project client and collaborator Samuel Khuvis of ParaTools, Inc. for his support and connection with the Performance Research Laboratory, University of Oregon, that provided access to the KNL Hardware. This work used the Extreme Science and Engineering Discovery Environment (XSEDE), which is supported by National Science Foundation grant number ACI–1053575. We acknowledge the Texas Advanced Computing Center (TACC) at The University of Texas at Austin for providing HPC resources that have contributed to the research results reported within this paper.http://www.siam.org/Portals/0/Publications/SIURO/Volume%2010/Performance_comparison_Intel_Xeon_Phi_Knights_Landing.pdf?ver=2018-02-28-145959-11
Comparison of Performance Analysis Tools for Parallel Programs Applied to CombBLAS
No full textPerformance analysis tools are powerful tools for high performance computing. By breaking down a program into how long the CPUs are taking on each process (pro- filing) or showing when events take place on a timeline over the course of running a program (tracing), a performance analysis tool can tell the programmer exactly, where the computer is running slowly. With this information, the programmer can focus on these performance "hotspots," and the code can be optimized to run faster. We com- pared the performance analysis tools TAU, ParaTools ThreadSpotter, Intel VTune, Scalasca, HPCToolkit, and Score-P to the example code CombBLAS (combinatorial BLAS) which is a C++ implemenation of the GraphBLAS, a set of graph algorithms using BLAS (Basic Linear Algebra Subroutines). Using these performance analysis tools on CombBLAS, we found three major "hotspots" and attempted to improve the code. We were unsuccessful in improving these "hotspots" due to a time limitation but still gave suggestions on improving the OpenMP calls in the CombBLAS code.https://userpages.umbc.edu/~gobbert/papers/REU2015Team8.pd
A Technique for the Quantitative Assessment of the Solution Quality on Particular Finite Elements in COMSOL Multiphysics
No full textProceedings of the COMSOL Conference 2007, BostonTo validate the reasonableness of a numerical solution to a partial differential equation on a given mesh, a common approach is to refine the mesh, compute a solution on the finer mesh, and compare the solutions on the two meshes. Comparing graphical representations of the two solutions gives a qualitative assessment of the solution quality. In many cases though, a priori error estimates from the theory of the finite element method are available that provide quantitative predictions of the expected solution quality in terms of the mesh spacing. This note shows how to use tools available in COMSOL Multiphysics to compute numerical estimates that can confirm if the finite element method performs as predicted by the theory. The technique presented does not assume that the true solution of the PDE is known. It is applied to linear Lagrange elements here, and extensions and limitations of the technique are discussed.https://userpages.umbc.edu/~gobbert/papers/GobbertCOMSOL2007.pd
Parallelization for Fast Image Reconstruction using the Stochastic Origin Ensemble Method for Proton Beam Therapy
No full textProton beam therapy is becoming increasingly common in the eld of cancer treatment because of the advantages over other forms of radiation therapy. These advantages arise from the nite range of the proton beams, the relatively low dosage of radiation upon entering a patient, and the large spike in dose at the end of the beam range known as the Bragg peak. A new computer code has been developed that uses the stochastic origin ensemble method to reconstruct an image of the gamma radiation produced by the proton beam. The objective of this research is to signi cantly improve the run time of the given computer code. For the reconstruction algorithm to be useful in medicine, it must be fast and precise, since it is impractical to ask that a patient lie completely still for several minutes. The original C++ code using OpenMP multi-threading on CPUs was ported to hybrid CPU/GPU code using CUDA. It shows very good speedup on the GPU up to the maximum possible number of threads, achieving a 5x speedup over the serial CPU run.https://userpages.umbc.edu/~gobbert/papers/REU2015Team7.pd
Online Training in Team-Based Multidisciplinary Research on Big Data + High-Performance Computing + Atmospheric Sciences
No full texthttps://userpages.umbc.edu/~gobbert/papers/Poster_OnlineLearning_SIAM_CSE21_highlighted.pd
Matthias Wagner : author profile
No full textThe author presented on this page has published his 10. article in Angewandte Chemie in the last 10 years
IDL: A Possible Alternative to Matlab?
No full textCreated by Exelis Visual Information Solutions, IDL (Interactive Data Language) is a commercial package used for data analysis. We compared the usability and efficiency of IDL to that of Matlab to determine if IDL is a viable substitute. Two studies were performed for this analysis. The first, a basic test inspired by the CIRC Tutorial for Basic Matlab, consisted of solving a system of linear equations using basic operations, computing eigenvalues and eigenvectors, and creating two-dimensional plots. It showed
identical results between the two packages, though it is important to note that the syntax and display of output between IDL and Matlab differs greatly. The second test focused on direct and iterative solutions of a large sparse system of linear equations.
This system arises from the finite difference discretization of the Poisson equation with homogeneous Dirichlet boundary conditions and is prototypical for linear systems in many related contexts. In Matlab, Gaussian elimination was used as the direct
method of solving the Poisson test problem. Unfortunately, such a method was not available with our license of IDL to solve sparse systems as it required a more expensive IDL Analyst License. Originally, we aimed to solve the problem iteratively using the conjugate gradient method, but, though a function was available in Matlab for solving a sparse system this way, none existed in IDL. Instead, we turned to the biconjugate gradient method. The numerical results of this method in IDL are identical to those in
Matlab, but IDL runs the code slightly faster for finer meshes. Those looking to make a switch from Matlab to IDL might have a difficult time encountering a different syntax, output display, and the need for a more expensive license to run a larger breadth of
functions or procedures, but if efficiency is of concern, IDL can potentially be faster than Matlab.The facility is supported by the U.S. National Science Foundation through the MRI program (grant no. CNS–0821258) and
the SCREMS program (grant no. DMS–0821311), with additional substantial support from the University of Maryland, Baltimore County (UMBC). See www.umbc.edu/hpcf for more information on HPCF and the projects using its resources.https://userpages.umbc.edu/~gobbert/papers/ComanThesis2012.pd
- …
