Results 1 -
3 of
3
Task Parallel Implementation of the Bowyer-Watson Algorithm
- Proceedings of Fifth International Conference on Numerical Grid Generation in Computational Fluid Dynamics and Related Fields
, 1996
"... In this paper we present a parallel implementation of the Bowyer-Watson (BW) algorithm using the task-parallel programming model. The BW algorithm constitutes an ideal mesh refinement strategy for implementing a large class of unstructured mesh generation techniques on both sequential and parallel c ..."
Abstract
-
Cited by 14 (9 self)
- Add to MetaCart
In this paper we present a parallel implementation of the Bowyer-Watson (BW) algorithm using the task-parallel programming model. The BW algorithm constitutes an ideal mesh refinement strategy for implementing a large class of unstructured mesh generation techniques on both sequential and parallel computers, by preventing the need for global mesh refinement. Its implementation on distributed memory multicomputes using the traditional data-parallel model has been proven very inefficient due to excessive synchronization needed among processors. In this paper we demonstrate that with the task-parallel model we can tolerate synchronization costs inherent to data-parallel methods by exploring concurrency in the processor level. Our preliminary performance data indicate that the taskparallel approach: (i) is almost four times faster than the existing data-parallel methods, (ii) scales linearly, and (iii) introduces minimum overheads compared to the "best" sequential implementation of the BW...
Discovery and Application of Network Information
, 2000
"... USAF, under agreement number F30602-96-1-0287. The U.S. Government is authorized to reproduce and distribute reprints for Governmental purposes notwithstanding any copyright annotation thereon. The views and conclusions contained herein are those of the authors and should not be interpreted as neces ..."
Abstract
-
Cited by 1 (0 self)
- Add to MetaCart
USAF, under agreement number F30602-96-1-0287. The U.S. Government is authorized to reproduce and distribute reprints for Governmental purposes notwithstanding any copyright annotation thereon. The views and conclusions contained herein are those of the authors and should not be interpreted as necessarily representing the official policies or endorsements, either expressed or implied, of the Advanced Research
One-Sided Communication Over MPI-1
, 2001
"... MPI-1 has been accepted as the standard parallel messaging substrate. However, as parallel applications become more complex and communication patterns more difficult to predict, there is a dire need for one-sided communication that allows asynchronous message passing. In this paper we present a one- ..."
Abstract
- Add to MetaCart
MPI-1 has been accepted as the standard parallel messaging substrate. However, as parallel applications become more complex and communication patterns more difficult to predict, there is a dire need for one-sided communication that allows asynchronous message passing. In this paper we present a one-sided communication system that provides users with the interface to send asynchronous messages whenever the need arises. On the receiving side, it allows the user to receive messages without knowing the intent or size of those messages. Portability can be inherited by building this system on top of MPI in ANSI C. By providing only the bare minimum in terms of API, the user has to suffer minimal overhead. The bottom line is that with minimal addition of code, a portable, one-sided, asynchronous parallel runtime system can be created that greatly enhances the user's environment.

