Abstract:Parallel principles are the most effective way how to increase performance in parallel computing (parallel computers and algorithms too). In this sense the paper is devoted to a complex performance evaluation of matrix parallel algorithms (MPA). At first the paper describes the typical matrix parallel algorithms and then it summarizes common properties of them to complex performance modeling of MPA. To complex performance analysis we are able to take into account all overheads influence performance of parallel algorithms (parallel computer architecture, parallel computation, communication etc.). To be le to analyze MPA in their abstract form we have defined needed decomposition models of MPA. For these decomposition strategies we derived analytical relation for defined complex performance criterions including isoefficiency functions, which allow us to predict performance although for hypothetical parallel computer. In its experimental part the paper considers the achieved results using defined complex performance criterions including issoefficiency function for performance prediction also for hypothetical future parallel computers. Such idea of common abstract analysis could be very useful in deriving complex performance criterions for groups of other similar parallel algorithms (PA) as for example numerical integration PA, optimization PA etc.
Keywords:
Trends in Parallel ComputingBasic common properties in parallel computing (parallel computers, parallel algorithms) computing, which are reaching continuous demands to performance acceleration, are as follows embedded parallel principles on various levels of technical (hardware) and program support means (software) [8] using of homogenous shared resources so in computing nodes of parallel computers (processors, cores, computers) as in parallel algorithms too [24] using of high speed communication networks reducing communication latency [39] increased client/server computing on symmetrical multiple processors or cores (SMP) trends to unified modeling of parallel computers (shared memory, distributed memory) and in parallel algorithms (shared memory, distributed memory, hybrid) continuous demands to increase mobility and data migration [23] the development of hardware neutral parallel programming language, such as Java, provides a virtual computational environment in which computing nodes of parallel computer appear to be homogenous continuous improvements in network technology and communication middleware in order to use shared parallel resources in unified manner (cloud computing, Internet computing). Current trends also in high performance computing (HPC) are to use networks of workstations (NOW, SMP) as a cheaper alternative to traditionally used massively parallel multiprocessors or supercomputers and to profit from unifying of both mentioned disciplines [19]. The individual powerful computing nodes (workstations) could be so single personal computer (PC) as parallel computers based on modern SMP parallel computers implemented within computing no...