1997
DOI: 10.1007/bfb0017245
|View full text |Cite
|
Sign up to set email alerts
|

Data localization using loop aligned decomposition for macro-dataflow processing

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1

Citation Types

0
2
0

Year Published

2003
2003
2008
2008

Publication Types

Select...
3
2

Relationship

2
3

Authors

Journals

citations
Cited by 6 publications
(2 citation statements)
references
References 26 publications
0
2
0
Order By: Relevance
“…To improve a data locality among the macro-tasks, Loop Aligned Decomposition (LAD) [17] is applied to the loops that access data larger than a local memory or a cache memory size. LAD divides the loops, or macro-tasks, into partial loops with the smaller number of iterations so that data size accessed by the divided loops is smaller than the local memory or the cache memory size.…”
Section: Data Localization Schemementioning
confidence: 99%
“…To improve a data locality among the macro-tasks, Loop Aligned Decomposition (LAD) [17] is applied to the loops that access data larger than a local memory or a cache memory size. LAD divides the loops, or macro-tasks, into partial loops with the smaller number of iterations so that data size accessed by the divided loops is smaller than the local memory or the cache memory size.…”
Section: Data Localization Schemementioning
confidence: 99%
“…Also, PE local data memory (LDM), which is used for storing PE local data and can have twice larger memory than DSM since local memory only have a single port, can be used effectively [1,5]. Furthermore, global registers among processors allow us to reduce data and synchronization overhead with good register allocation technique in the compiler.…”
Section: Architecture For Near Fine Grain Parallel Processingmentioning
confidence: 99%