CSE 260 Final Project: Multigrid using MPI
نویسنده
چکیده
In this project, we began with a serial plus OpenMP implementation of a multigrid method to solve Poisson’s equation in three dimensions. Our main goal was to parallelize this method using data-level parallelism and MPI. To this end, our primary goal was to extend the work of previous students by handling two and three dimensional data divisions. Beyond that, we sought to implement the fastest possible MPI communication, possibly through the use of OpenMP in packing and unpacking data structures, the use of the MPI Pack function or new MPI data types. We also gave secondary consideration to dynamic load reallocation, i.e. the turning on and off of processes at different levels of grid refinement in the multigrid V-cycle. We adjusted these goals as we worked in order to best serve our own learning and to allow us to complete the project within the given timeframe.
منابع مشابه
CSE 260 Project Report Parallelizing Pre-conditioned Conjugate Gradient Method with OpenMP, MPI and CUDA
متن کامل
Development of a Korean OCR System Term Project in CSE 581 - Pattern Recognition
This is the final report for the term project in CSE 581 (Pattern Recognition). The goal of this project is to develop a character recognition system that is able to recognize and classify a subset of the Korean language. Training and test samples where obtained from Korean books, additional sets created by applying a degradation model to the obtained samples. Using geometrical, statistical and...
متن کاملThe High{Level Communications Library CLIC
One important class of applications for high performance computers is based on 3{ dimensional block{structured grids. Multigrid and CFD codes using nite diierence or-nite volume discretizations fall into this class of applications. As the communication parts of these application programs are similar, the central Communications Library CLIC{3D has been developed. Because the CLIC Library is base...
متن کاملParallel Multigrid Solvers Using OpenMP/MPI Hybrid Programming Models on Multi-Core/Multi-Socket Clusters
OpenMP/MPI hybrid parallel programming models were implemented to 3D finite-volume based simulation code for groundwater flow problems through heterogeneous porous media using parallel iterative solvers with multigrid preconditioning. Performance and robustness of the developed code has been evaluated on the “T2K Open Supercomputer (Tokyo)” and “Cray-XT4” using up to 1,024 cores through both of...
متن کاملOn the Performance of an Algebraic Multigrid Solver on Multicore Clusters
Algebraic multigrid (AMG) solvers have proven to be extremely efficient on distributed-memory architectures. However, when executed on modern multicore cluster architectures, we face new challenges that can significantly harm AMG’s performance. We discuss our experiences on such an architecture and present a set of techniques that help users to overcome the associated problems, including thread...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2012