2012 **Abstract:** Abstract. Reconfigurable Architecture (RA), which provides extremely high energy efficiency for certain domains of applications, have one problem that current mapping algorithms for it do not scale well with the number of cores. One approach to this problem is using SIMD (Single Instruction Multiple Data) paradigm. However, SIMD can complicate the mapping problem by adding an additional dimension, i.e., iteration mapping, to the already inter-dependent problems of data mapping and operation mapping, and can si…

Help me understand this report

Search citation statements

Paper Sections

Select...

3

1

1

Citation Types

0

8

0

Year Published

2014

2014

Publication Types

Select...

1

Relationship

0

1

Authors

Journals

(8 citation statements)

(8 reference statements)

0

8

0

“…When exploiting the data parallelism, the effect is terrible. The 4-dot product has no loop-carried dependence, and there is no conflict with the method of [8]. Similarly, the same goal can be obtained by our conflict-free iteration-data mapping algorithm with no space cost.…”

confidence: 53%

“…When exploiting the data parallelism, the effect is terrible. The 4-dot product has no loop-carried dependence, and there is no conflict with the method of [8]. Similarly, the same goal can be obtained by our conflict-free iteration-data mapping algorithm with no space cost.…”

confidence: 53%

“…Moreover, we can copy several RPs to exploit the data parallelism. In [8] an SIMD (Single Instruction Multiple Data) reconfigurable architecture is proposed to map some iterations of a kernel to one array in SIMD style.…”

confidence: 99%