RADIC scalability analysis: functional Mmodel

Abstract

In parallel systems, a number of measures of performance are not accurate or representative of their functioning. These measures allow to quantify the benefit of parallelism. Very often, programs are designed and tested for smaller problems on fewer processing elements. However, the real problems these programs are intended to solve are much larger, and the machines contain a great number of processing elements. Hence, it is necessary to create a model that allows to extrapolate the application execution over a few processing elements to larger machine configurations. These measures are more complex if we consider faults. When we take measures we must understand the interaction among system architecture, application architecture and fault tolerant system. In this paper we present a model which analyzes the combination parallel computer, parallel application and RADIC fault tolerance architecture.