Towards A Mathematical Specification of the APE100 Architecture:the APESE Model
Egon Börger, Giuseppe Del Castillo, Paola Glavan, Dean Rosenzweig · 1994
This paper provides the first step of a full mathematical description of the APE100 parallel architecture. The description consists of several models, at different levels of abstraction, corresponding to views of the architecture provided by different languages used within the APE100 compilation chain (a crucial part of the software environment of APE100). Here we present the primary model , based on the relevant subset of APESE, a high level language specially designed for APE100 and constituting the source language in the compilation chain: APESE reflects closely the APE100 model of parallel execution and is therefore adequate for an abstract description of the main features of the latter. Stepwise refinement will lead us from the APESE model of this paper to the hardware level of APE100, using evolving algebras as specification method (see [1,2]). APE100 provides a beautiful example of an evolving algebra which models lock-step parallelism: at each step all rules which can be fired are fired simultaneously. Keyword Codes: C.1; C.3; J.2 Keywords: Processor Architectures; Special--Purpose and Application--Based Systems; Physical Sciences and Engineering