Performance Analysis Of Parallel Programs Research Articles

Empirical Performance Analysis of Parallel Programs on Wired and Wireless Local Area Networks Using Beowulf Clusters

DOI: 10.9734/BJMCS/2016/21744 Editor(s): (1) Dariusz Jacek Jakobczak, Chair of Computer Science and Management in this department, Technical University of Koszalin, Poland. Reviewers: (1) Ahmet Sayar, Kocaeli University, Turkey. (2) Anand Nayyar, KCL Institute of Management and Technology, India. (3) Xinwei Niu, SASTRA University, Thanjavur, India. Complete Peer review History: http://sciencedomain.org/review-history/12383

RPPA: A Remote Parallel Program Performance Analysis Tool

Parallel program performance analysis plays an important role in exploring parallelism and improving efficiency of parallel programs. To ease the performance analysis for remote programs, this paper presents a remote parallel program performance analysis tool, RPPA (Remote Parallel Performance Analyzer), which is based on dynamic code instrumentation. A hierarchical structure is adopted by RPPA which consists of 3 parts: client, server and computing nodes. Performance analysis tasks are submitted to the server via the graphical user interface of the client, and then actual analysis processes are started on the computing nodes by the server to collect performance data for visualization in the client. The performance information gained by RPPA is comprehensive and intuitive, hence it is quite helpful for users to analyze performance, locate bottlenecks of programs, and optimize programs.

HieraAnalyses – a tool for hierarchical analysis of parallel programs

Detailed information for performance analysis of parallel programs can be collected through trace files. Generally, trace files contain a register of individual events that occurred during program execution. Considering that the events traced are commonly of low level, like communication operations in a parallel system, and that it is increasingly common for the application programmer to use higher level abstractions (e.g., a parallel eigenvalues routine), a semantic gap exists between the collected information and the concepts used for the development of the application, hindering an effective use of that information. In this work, a new approach to trace files is proposed, where the files retain information about the different hierarchical levels in the application. The files follow an XML format, where routines are XML tags, with auxiliary routines called during its execution as child tags. The approach is demonstrated by its implementation for the MPI library level and the OOPS level, this last one being an object-oriented framework with higher level abstractions for the development of parallel programs that uses MPI for its implementation. To complement the work, some analysis tools using the file format are presented.

Visualization of do-loop performance using AVS

Performance visualization is the use of graphical display techniques for the analysis of performance data in order to improve understanding of complex performance phenomena. Performance visualization systems for parallel programs have been helpful in the past and they are commonly used in order to improve parallel program performance. However, despite the advances that have been made in visualizing scientific data, techniques for visualizing performance of parallel programs remain ad hoc and performace visualization becomes more difficult as the parallel system becomes more complex. Massively parallel processors can produce a huge amount of performance data and sophisticated methods for representing and displaying this data are required. The use of scientific visualization tools (e.g. AVS, Application Visualization System) to display performance data is becoming a very poewrful alternative to support performance analysis of parallel programs. One advantage of this approach is that no tool development is required and that every feature of the data visualization tool can be used for further data analysis. In this paper the Do-Loop-Surface (DLS) display, an abstract view of the performance of a particular do-loop in a program implemented using AVS, is presented as an example on how a data visualization tool can be used to define new abstract representations of performance, helping the user to analyze complex data potentially generated by a large number of processors.

Performance analysis of parallel programs based on directed acyclic graphs

Task precedence graphs are widely used for modeling and evaluation of parallel applications. Their nodes represent the subtasks of the parallel program and the edges represent the precedence relations between the subtasks. The execution times of the subtasks are described by random variables and their distributions. In our paper we introduce a new class of distributions, particularly suited for the modeling and evaluation of parallel programs. Exponential polynomials introduced by Sahner and Trivedi have the disadvantage that a large number of parameters is needed for the representation of realistic task execution times, which usually have a small value of variation. We extend this class to derive the class of truncated \(\theta\)-exponential polynomials which allow the representation of realistic task execution times with fewer parameters. Additionally this class of distributions has the advantage that minimum as well as maximum execution times can be guaranteed. Models with a large number of subtasks \(n\) can not be evaluated on a computer using exact analytical methods because of memory requirements and numerical inaccuracies, which accumulate, when the operations of analysis are applied. Using extreme value theory we derive approximate formulas for the parallel independent execution of \(n\) subtasks, a structure, which can be found in every parallel program. The obtained results for truncated and not truncated distributions show, that distributions with an infinite domain are not suitable, particularly for massively parallel structures.

Performance analysis of parallel programs based on model calculations

Modeling parallel programs can help the developer during the design phase of a particular implementation on the one hand and on the other hand provide principle insights which are needed to establish design principles for the development of parallel programs. It is important that the model used provides a sufficient but not too detailed representation of the parallel program. Precedence graphs provide an easy to understand representation of the structure of parallel programs. Information about the behavior of parallel programs during execution can be obtained by assigning distributions of the execution times of the subtasks to the nodes of these graphs. In this paper we introduce truncated θ- exponential polynomials as a class of distributions suitable for the modeling of task run-times of massively parallel programs. As opposed to classical exponential polynomials known from literature, truncated θ-exponential polynomials allow the representation of distributions with finite domain and are therefore appropriate for the analysis of massively parallel systems. We present approximation formulas which allow the approximate calculation of the total execution time for large models in particular. We provide formulas for the parallel execution of a large number of subtasks. We also give a method to estimate the total execution time of regular graphs, which are also very important in the field of modeling parallel programs.

Teaching empirical performance analysis of parallel programs

Performance is a central issue in parallel computing. In this paper, we describe our approach to teaching advanced undergraduates and graduate students about the fundamentals of measuring and analyzing the performance of programs running on a variety of parallel machines. This approach can be applied to virtually any type of parallel machine, as well as to parallel program simulators. Although performance analysis can serve many purposes, we focus on the needs of the parallel programmer: understanding the behavior of algorithms and programs, and making informed choices among them.