On the Power of the Multiple Associative Computing (MASC) Model Related to That of Reconfigurable Bus-Based Models

Abstract

IntroductionThe MASC (for Multiple Associative Computing) model is a multi-SIMD model that uses control parallelism to coordinate the interaction of data parallel threads. The MASC model extends the concept of associative computing and provides a complete paradigm to support general parallel computation. Motivated from the STARAN computer built in 1970s by Goodyear Aerospace, the MASC model was proposed by Potter and Baker in 1994 [14] and has been studied at Kent State University as a practical model for years. Since the MASC model is a strictly synchronous model that supports SIMD computation, it is sometimes called a multi-SIMD or MSIMD model (i.e., a SIMD enhanced with multiple instruction streams).In contrast to a number of other parallel models, the MASC model possesses certain constant time global properties such as constant time broadcasting, constant time global reduction operations, and constant time associative search, which have all been justified in [7]. These properties have made MASC able not only to solve general parallel processing problems effectively [2,7,13,17] but also to solve some problems in special areas such as real-time air traffic control in an extremely efficient manner [9,10]. Possible techniques for implementing the MASC model have been explored in [5,6,19,21].The power of a computational model is indicated both by the efficiency of algorithms it supports and by the efficiency with which it can simulate other computational models. In order to evaluate the power of the MASC model, simulation and comparison with other popular parallel models such as PRAM (Parallel Random Access Machine) and MMB (Mesh with Multiple Buses) have been studied previously. Constant time simulation of PRAM with MASC and constant time simulations between MMB with MASC have both been established [16,3]. In this paper, we present our work on simulating MASC and reconfigurable bus-based models. We aim to relate the power of the MASC model with that of the RM (Reconfigurable Mesh) model, as RM has been widely accepted as an extremely powerful computation model. In order to establish a relationship between MASC and RM, we use a bridge model -RMBM (Reconfigurable Multiple Bus Machine) and discuss simulations between RMBM and MASC first. Then we

Comparisons of air traffic control implementations on an associative processor with a MIMD and consequences for parallel computing

“…Many algorithms have been designed for the ASC model and software projects for AP platform (often using the ASC language) in order to demonstrate the versatility of the ASC model and the AP system. These include graph algorithms [67,44,42], convex hull [68,69,70,71], string matching [72,73], sequence alignment [74,75], NPComplete [76,77], databases [78,79], parallel compilers [80,81,82], supporting functional and logic-based languages [83,84,85], and supporting expert system languages [86,87,88]. In [39], Batcher lists fast Fourier transforming, sonar post-processing, string search, file processing, air traffic control, image processing, data management, position locating and reporting, bulk filtering, and radar processing as STARAN applications, and for the first five of these, provides a comparison of the STARAN's performance with several other systems.…”