Survey on Parallel Programming Model

Abstract

Abstract. The development of microprocessors design has been shifting to multi-core architectures. Therefore, it is expected that parallelism will play a significant role in future generations of applications. Throughout the years, there has been a myriad number of parallel programming models proposed. In choosing a parallel programming model, not only the performance aspect is important, but also qualitative the aspect of how well parallelism is abstracted to developers. A model with a well abstraction of parallelism leads to a higher application-development productivity. In this paper, we propose seven criteria to qualitatively evaluate parallel programming models. Our focus is on how parallelism is abstracted and presented to application developers. As a case study, we use these criteria to investigate six well-known parallel programming models in the HPC community.

A taxonomy of task-based parallel programming technologies for high-performance computing

“…1 presents our API taxonomy. A subset of these features was adapted from previous work by Kasim et al [19]. To these existing characteristics, we added additional information of general importance-such as technological readiness levels-as well as features which relate to new capabilities particularly relevant for modern HPC like support for heterogeneity and resilience management.…”

A package for OpenCL based heterogeneous computing on clusters with many GPU devices

“…A survey of parallel programming models, including OpenMP, CUDA and MPI, was presented by Kasim et al [14]. MGP is an extension of the OpenMP programming model, which allows running of GPU code in a simple, OpenMP-like manner.…”

PREPARING FOR EXASCALE: ORNL Leadership Computing Application Requirements and Strategy

“…Factors such as the rise of heterogeneous computing, the quantum jump in the amount of parallelism required in applications, and the increasing need for locality are creating a disruption in the methods of [Kasim et al 2008].…”