A parallel algorithm for prefix computation on N data elements mapped on a multimesh (MM) network of N = n(4) processing elements is presented here. The time required by the proposed algorithm is significantly less t...
详细信息
A parallel algorithm for prefix computation on N data elements mapped on a multimesh (MM) network of N = n(4) processing elements is presented here. The time required by the proposed algorithm is significantly less than that by any of the existing algorithms for prefix computation on mesh-like architectures due to the specific interconnection pattern used in the MM network. The proposed technique requires O(N-1/4) time for data communication and O(logN(1/4)) time for computation, when mapped on a MM network constituted by N-1/2 meshes, each of size N-1/4 x N-1/4. The data communication time in the proposed algorithm is less than the prefix sum algorithm proposed in extended multimesh. To be precise, instead of (13N(1/4) - 5) tau communication time the proposed algorithm requires a data communication time of 7.5N(1/4) t only. Moreover, the proposed parallel algorithm does not need any extra inter block links as used in the extended multimesh.
A parallel algorithm for prefix computation of N = n(4) elements on an n x n extended multi-mesh network is presented. The A parallel algorithm for prefix computation of N = n(4) network is a modified version of an ea...
详细信息
A parallel algorithm for prefix computation of N = n(4) elements on an n x n extended multi-mesh network is presented. The A parallel algorithm for prefix computation of N = n(4) network is a modified version of an earlier multi-mesh network with a 4-regular structure. The algorithm takes O(N-1/4) time on N processors (13N(1/4)-5 communication steps and log N + 4 arithmetic/logic steps). (C) 2002 Elsevier Science B.V. All rights reserved.
暂无评论