Abstract | ||
---|---|---|
This paper compares three similar loop-grouping methods. All methods are based on projecting the n-dimensional iteration space J n onto a k-dimensional one, called the projected space, using (n-k) linear independent vectors. The dimension k is selected differently in each method giving various results. The projected space is divided into discrete groups of related iterations, which are assigned to different processors. Two of the methods preserve optimal time completion, by scheduling loop iterations according to the hyperplane method. The theoretical analysis of the experimental results indicates the appropriate method, for specific iteration spaces and target architectures. |
Year | DOI | Venue |
---|---|---|
2000 | 10.1109/ICPP.2000.876163 | ICPP |
Keywords | Field | DocType |
parallel architectures,parallel programming,communication overhead,hyperplane method,loop grouping methods,mesh-connected architectures,n-dimensional iteration space,optimal time completion,orthogonal projection,orthogonal projection spaces | Linear independence,Orthographic projection,Scheduling (computing),Computer science,Hyperplane,Iteration space,Distributed computing | Conference |
ISSN | ISBN | Citations |
0190-3918 | 0-7695-0768-9 | 9 |
PageRank | References | Authors |
0.59 | 9 | 5 |
Name | Order | Citations | PageRank |
---|---|---|---|
Ioannis Drositis | 1 | 14 | 1.73 |
Giorgos Goumas | 2 | 9 | 0.59 |
N. Koziris | 3 | 1015 | 107.53 |
Panayiotis Tsanakas | 4 | 115 | 13.91 |
George Papakonstantinou | 5 | 134 | 14.52 |