Heterogeneous Parallel Programming
نویسندگان
چکیده
54 Published By: Blue Eyes Intelligence Engineering & Sciences Publication Pvt. Ltd. Abstract—This paper presents Heterogeneous Parallel Computing (HPC), which is a well-orchestrated and co-ordinated effective use of a suite of diverse high performance machines to provide super-speed processing for computationally demanding tasks with diverse computational needs. GPUs are accoutered with a much more throughput oriented design as compared to that of the CPUs thus making them a powerful alternative to boast overall performance. It is now used all the way from mobile computing to supercomputing, like in Blue Star Super Computers. Upcoming Exascale and Petascale systems have embraced even heterogeneity in order to overcome power limitations. This paper also illustrates programming example using CUDA C to demonstrate the efficiency achieved in problems like matrix multiplication using a more heterogeneous approach as compared to that of sequential approach. It also explains how Heterogeneous Parallel Programming is a plausible, novel technique which allows to exploit inherent capabilities of a wide range of computational machines to solve computationally intensive problems that have several types of embedded parallelism by breaking it into separate modules. This paper also puts light on the challenges and concerns which exist when programming in HPC environment and some techniques to alleviate them.
منابع مشابه
Accelerating high-order WENO schemes using two heterogeneous GPUs
A double-GPU code is developed to accelerate WENO schemes. The test problem is a compressible viscous flow. The convective terms are discretized using third- to ninth-order WENO schemes and the viscous terms are discretized by the standard fourth-order central scheme. The code written in CUDA programming language is developed by modifying a single-GPU code. The OpenMP library is used for parall...
متن کاملParallel computing using MPI and OpenMP on self-configured platform, UMZHPC.
Parallel computing is a topic of interest for a broad scientific community since it facilitates many time-consuming algorithms in different application domains.In this paper, we introduce a novel platform for parallel computing by using MPI and OpenMP programming languages based on set of networked PCs. UMZHPC is a free Linux-based parallel computing infrastructure that has been developed to cr...
متن کاملStochastic Comparisons of Series and Parallel Systems with Heterogeneous Extended Generalized Exponential Components
In this paper, we discuss the usual stochastic‎, ‎likelihood ratio, ‎dispersive and convex transform order between two parallel systems with independent heterogeneous extended generalized exponential components. ‎We also establish the usual stochastic order between series systems from two independent heterogeneous extended generalized exponential samples. ‎Finally, ‎we f...
متن کاملLinguistic Support for Heterogeneous Parallel Processing: A Survey and an Approach
Coding a highly parallel application to run on a heterogeneous suite of processors (both metacomputers and mixed-mode computers) with high efficiency, ease of implementation, and portability is a significant challenge. This paper first surveys recently proposed and existing parallel languages from the perspective of programming complex, heterogeneous systems. We then propose two essential featu...
متن کاملHeterogeneous Parallel Computing with Java: Jabber or Justified?
Is Java a good language for programming heterogeneous parallel computing systems? It is a well-designed modern language that, combined with the Java Virtual Machine (JVM), offers a myriad of modern programming features and excellent portability. Howev er, in speedup-oriented heterogeneous computing, our primary concern is obtaining the best possible execution speed from the heterogeneous system...
متن کاملA fuzzy mixed-integer goal programming model for a parallel machine scheduling problem with sequence-dependent setup times and release dates
This paper presents a new mixed-integer goal programming (MIGP) model for a parallel machine scheduling problem with sequence-dependent setup times and release dates. Two objectives are considered in the model to minimize the total weighted flow time and the total weighted tardiness simultaneously. Due to the com-plexity of the above model and uncertainty involved in real-world scheduling probl...
متن کامل