Multiprocessor Organization—a Survey
ACM Computing Surveys (CSUR)
The design and performance of a hierarchical multi/microprocessor.
The design and performance of a hierarchical multi/microprocessor.
Survey on special purpose computer architectures for AI
ACM SIGART Bulletin
Optimal granularity of parallel evaluation of AND trees
ACM '86 Proceedings of 1986 ACM Fall joint computer conference
A tree generating algorithm for designing optimal hierarchical distributed systems
SAC '97 Proceedings of the 1997 ACM symposium on Applied computing
A hardware support mechanism for scheduling resources in a parallel machine environment
ISCA '81 Proceedings of the 8th annual symposium on Computer Architecture
The Binary Tree as an Interconnection Network: Applications to Multiprocessor Systems and VLSI
IEEE Transactions on Computers
Divide-and-Conquer for Parallel Processing
IEEE Transactions on Computers
| Hi-index | 0.01 |
The paper describes the results of simulation experiments of a tree organized multicomputer now being constructed in the Department of Computer Science at Stony Brook. First the structure of the multicomputer is introduced. This is based on, (i) separate local memories, (ii) a tree organization mirrored on that of social structures, and (iii) a distributed file system. The simulation studies were designed to illuminate the performance of the multicomputer when cooperating on the solution of selected single computational tasks. The results showed (i) the potential for an order of magnitude increase in the number of cooperating processors before performance saturation effects make themselves felt, (ii) a high frequency of processor resets relative to the completion time of the composite job, and (iii) a lower utilization of processors near the root of the tree relative to those near the leaves. The implications of these results for the design of the hardware and software of this multicomputer are discussed in a conclusion.