IMPLEMENTING NEURAL NETWORK MODELS ON PARALLEL COMPUTERS

被引：31

作者：

FORREST, BM ^{[1
]}

ROWETH, D ^{[1
]}

STROUD, N ^{[1
]}

WALLACE, DJ ^{[1
]}

WILSON, GV ^{[1
]}

机构：

[1] UNIV EDINBURGH,DEPT PHYS,JAMES CLERK MAXWELL BLDG,KINGS BLD,MAYFIELD RD,EDINBURGH EH9 3JZ,MIDLOTHIAN,SCOTLAND

来源：

COMPUTER JOURNAL | 1987年 / 30卷 / 05期

关键词：

COMPUTER ARCHITECTURE - Medical Applications - COMPUTER PROGRAMMING - Algorithms - COMPUTER SYSTEMS; DIGITAL - Parallel Processing;

D O I：

10.1093/comjnl/30.5.413

中图分类号：

TP3 [计算技术、计算机技术];

学科分类号：

0812 ;

摘要：

The remarkable processing capabilities of the nervous system must derive from the large numbers of neurons participating (roughly 10**1**0), since the time-scales involved are of the order of a millisecond, rather than the nanoseconds of modern computers. The neural network models which attempt to capture this behavior are inherently parallel. We review the implementation of a range of neural network models on SIMD and MIMD computers. On the ICL Distributed Array Process (DAP), a 4096-processor SIMD machine, we have studied training algorithms in the context of the Hopfield net, with specific applications including the storage of words and continuous text in content-addressable memory.

引用

页码：413 / 419

页数：7

共 31 条

[1] STORING INFINITE NUMBERS OF PATTERNS IN A SPIN-GLASS MODEL OF NEURAL NETWORKS [J].