Reconfigurable mesh-connected processor arrays using row-column bypassing and direct replacement

Enhanced cluster k-Ary n-cube, a fault-tolerant multiprocessor

“…Even though, compared with the published results in the literature [6], [7], [8], [9], [10], [13], [14], [15], [16], [17], [19], [18], [20], [24], the ECKN under Theorem 1 has an attractive fault-tolerant capability, only a small fraction of spare nodes and spare links are utilized to tolerate 2n þ 1 faulty nodes. Hence, with the majority of the spare nodes and spare links left intact, one should expect that the actual fault tolerance of the ECKN would be much higher.…”

“…A non-nodeminimal f-fault-tolerant two-dimensional mesh [19] has a node degree of 13, but requires 8f 3 þ 16f 2 spare nodes. A second group of global reconfiguration schemes utilizes rows and columns of spare nodes to tolerate the faulty nodes [21], [22], [23], [24]. The main objective of rowcolumn bypass schemes is to increase yield (defect tolerance) in integrated microelectronic systems.…”

An efficient reconfiguration scheme for fault-tolerant meshes

“…Global reconfiguration [9][10][11][12][13][14][15][16][17] replaces a faulty node with a spare one regardless of the distance between them. Due to extensive utilization of spare nodes, global reconfiguration is able to achieve the maximum 100% reconfiguration rate.…”

“…For these schemes, desirable reconfiguration rates are often achieved at the cost of involving large node degrees (e.g., [11]) or a large number of spare nodes and links (e.g., [16,17]). …”

“…Similar researches have been conducted for various architectures, such as hypercubes, trees, meshes and k-ary ncubes [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20]. These designs employ either local or global reconfiguration in their replacement operations.…”