The Systolic Reconfigurable Mesh

Abstract

In this paper, we introduce the Systolic Reconfigurable Mesh (SRM), which combines aspects of the reconfigurable mesh with that of systolic arrays. Every processor controls a local switch that can be reconfigured during every clock cycle in order to control the physical connections between its four bi-directional bus lines. Data is input on one side of the systolic reconfigurable mesh and output from another side, one row/column per unit time. Efficient algorithms are presented for intermediate-level vision tasks, including histograming, connectivity, convexity, and proximity.

DNA Sequence Matching at Nanoscale Level

“…This chapter, demonstrated several techniques for forming graphs representing partial-order multiple-sequence alignment of a given set of N-aligned sequences, using two types of reconfigurable mesh architectures (the spin-wave version [10] and the VLSI version [11,12]). It showed that for a constant number of variables, the run-times of both architectures are the same, O(1).…”

Nanoscale Image Processing

“…">IMAGE PROCESSING ALGORITHMSOver the past few decades, several mesh-based parallel architectures such as meshconnected computer, mesh-of-trees, pyramid, mesh with multiple buses, reconfigurable meshes, systolic meshes, and optical meshes have been considered for performing low and intermediate-level computer vision tasks [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20]. The first section concentrates on image processing applications of nanoscale spin-wave architectures.…”

“…In a systolic reconfigurable mesh (SRM), as described in [19], data is input into the reconfigurable mesh and eventually output from the other side, one column of data per unit of time. A systolic reconfigurable mesh (SRM) of size 16 is shown in Figure 19.1.…”

“…Each generic processor contains a switch that is under local control and can be used to configure the four bus lines in any of 15 possible combinations. 19.1. A systolic reconfigurable mesh.…”

“…In this algorithm, as explained in [23] similar to the algorithm presented in [19], we represent the component label will as oCL,CR,TW, where CL is the index of the leftmost column containing a pixel of the component; CR is the index of the rightmost column containing a pixel of the component; and T is used to break ties in the case of distinct figures with the same CL and CR values. CL is computed for every pixel during the input and preprocessing phase of the algorithm, while CR is computed for every pixel during the output and postprocessing phase of the algorithm.…”