Extending goal-oriented parallel computer job scheduling policies to heterogeneous systems

Load balancing is a key problem in parallel computing. Under the environment of increasing heterogeneous parallel systems, load balancing strategies facing new challenges. This paper is a survey of the developments of load balancing strategies on heterogeneous parallel system. First, we compare the differences between homogeneous and heterogeneous systems. Second, we review the status of research on related load balancing strategies and analyze existing load balancing algorithms. Finally, we discuss future challenges for load balancing in heterogeneous parallel computing.

An efficient numerical method to simulate the conjugate heat transfer between a fluid and moving bodies is presented. The numerical approach is based on hierarchical Cartesian meshes and formulated for immersed boundaries such that simulations can be performed for solid bodies of arbitrary shape and without interpolation assumption on parallel high-performance computing systems. To ensure a consistent temperature and heat flux across the fluid–solid interface, a direct iterative procedure incorporating the solid and the fluid domains is used such that a fully energy conserving method is obtained. The method is validated by comparing simulation results for generic three-dimensional problems of stationary and moving geometries with reference data, for which also the conservation properties and the accuracy of the method are demonstrated. Results for the heat transfer of a sphere at a fixed position in an adiabatic domain, the flow over a backward-facing step, and a shape-transforming body are shown. Additionally, results for moving bodies are presented, which demonstrate the accuracy and versatility of the developed method. Finally, the efficient implementation of the algorithm for usage on parallel computing systems is shown by scaling and benchmark plots.

The advancements in technology in this contemporary time require special computing devices to able to manage the increase of the computational power. However, the limitation of the sequential computing systems stands as an obstacle in fulfilling these requirements. Thus, the research community looking for a viable solution able to solve the grand computing problems. A parallel computing system is a collection of processing elements that cooperate and communicate to provide a fast solution for a problem. A new symmetric network named Shifted Peterson Network (SPN) has been proposed in this paper. The basic module (BM) of the suggested network is composed of ten nodes, and the BMs of SPN are connected through external nodes to construct an advanced level of this network. Furthermore, the connectivity of SPN is based on a shifting mechanism of the binary digits of each node. The architecture of SPN discussed in this paper, in addition, the static network performance parameters of this network has been evaluated and compared to popular conventional interconnection networks. The networks evaluated in terms of node degree, diameter, average distance, arc connectivity, bisection width, cost, wiring complexity, packing density, and message traffic density. SPN showed good results in almost all aspects compared to these networks. Therefore, SPN is a suitable choice for the massively parallel computer (MPC) systems.

A Schwarz domain decomposition method for solution of transient unsaturated water flow on parallel computers

This study considers a heterogeneous computing system and corresponding workload being investigated by the Extreme Scale Systems Center (ESSC) at Oak Ridge National Laboratory (ORNL). The ESSC is part of a collaborative effort between the Department of Energy (DOE) and the Department of Defense (DoD) to deliver research, tools, software, and technologies that can be integrated, deployed, and used in both DOE and DoD environments. The heterogeneous system and workload described here are representative of a prototypical computing environment being studied as part of this collaboration. Each task can exhibit a time-varying emph{importance} or emph{utility} to the overall enterprise. In this system, an arriving task has an associated priority and precedence. The priority is used to describe the importance of a task, and precedence is used to describe how soon the task must be executed. These two metrics are combined to create a utility function curve that indicates how valuable it is for the system to complete a task at any given moment. This research focuses on using time-utility functions to generate a metric that can be used to compare the performance of different resource schedulers in a heterogeneous computing system. The contributions of this paper are: (a) a mathematical model of a heterogeneous computing system where tasks arrive dynamically and need to be assigned based on their priority, precedence, utility characteristic class, and task execution type, (b) the use of priority and precedence to generate time-utility functions that describe the value a task has at any given time, (c) the derivation of a metric based on the total utility gained from completing tasks to measure the performance of the computing environment, and (d) a comparison of the performance of resource allocation heuristics in this environment

Over the past several years, a heterogeneous computing (HC) system has become more competative as a commercial computing platform than a homogeneous system. With the growing scale of HC systems, network failures become inevitable. To achieve high performance, communication reliability should be considered while designing reliability-aware task scheduling algorithms. In this paper, we propose a new algorithm called RMSR (Replication-based scheduling for Maximizing System Reliability), which incorporates task communication into system reliability. To maximize communication reliability, an improved algorithm which searches all optimal reliability communication paths for current tasks is proposed. During the task replication phase, the task reliability threshold is determined by users and each task has dynamic replicas. Our comparative studies based on randomly generated graphs show that our RMSR algorithm outperforms existing scheduling algorithms in terms of system reliability. Several factors affecting the performance are analyzed in the paper.

A multiprocessor system is a computer with two or more central processing units (CPUs) with each one sharing the common main memory as well as the peripherals. Multiprocessor system is either homogeneous or heterogeneous system. A homogeneous system is a cluster of processors joined to a high speed network for accomplishing the required task; also it is defined as parallel computing system. Homogeneous is a technique of parallel computing system. A heterogeneous system can be defined as the interconnection of a number of processors, having dissimilar computational speed. Load balance is a method of distributing work between the processors fairly in order to get optimal response time, resource utilization, and throughput. Load balancing is either static or dynamic. In static load balancing, work is distributed among all processors before the execution of the algorithm. In dynamic load balancing, work is distributed among all processors during execution of the algorithm. So problems arise when it cannot statistically divide the tasks among the processors. To use multiprocessor systems efficiently, several load balancing algorithms have been adopted widely. This paper proposes an efficient load balance algorithm which addresses common overheads that may decrease the efficiency of a multiprocessor system. Such overheads are synchronization, data communication, response time, and throughput.

With the development of multicore-processor, the parallel computer system are gaining a rapid progress from supercomputer to personal computer. But the knowledge on parallel computing is seldom involved in the traditional undergraduate stage of computer specialty. Under the current situation, it is very necessary to set the parallel computing as an independent course in the teaching plan of computer specialty. In this paper, the necessity of setting parallel computing course and how to construct the curriculum are discussed. Materials and the arrangement of teaching content, curriculum experiments, and the relationship with the other subjects are also discussed in detail.

ABSTRACTAdding the number of computing nodes is a common approach to achieving higher performance in a parallel computing system. However, with constraint of fixed system architecture and fixed algorithm structure, it is difficult to improve the performance of parallel computing only by extending its scale absolutely. To realize such extension with fixed structure, we analyze key factors from architecture and parallel task, which affect the scalability, and then use the weighted graph to model architecture as well as parallel task. Especially, focusing on the case that architecture graph and parallel task graph are homogeneous, we propose the extension method of graph similarity; for the case that architecture graph and parallel task graph are heterogeneous, a critical-path-unchanged scaling method is proposed. Actually, the above two extending methods do not change the graph's structure. They only adjust the node weight and edge-weight in the relevant graph. Furthermore, through mathematical derivation, some conclusions about the new scaling methods are drawn. Finally, in order to verify the effectiveness, some simulative experiments are conducted on the platform SimGrid. The experimental results show that the proposed methods can realize iso-speed-efficiency extension, and can guide practical extensions for parallel computing.

Multi-objective energy-efficient workflow scheduling using list-based heuristics

We construct and analyse integration methods for solving initial value problems for implicit differential equations (IDEs) that can be efficiently used on parallel computer systems. We construct an IDE method for general IDEs of arbitrarily high index, and two methods that can be applied to partitioned IDEs. The partitioned IDE methods both exploit the special form of the problem and converge faster than the general IDE method. The first partitioned IDE method is suitable for higher-index problems, the second partitioned IDE method only applies to index 1 problems, but is considerably less expensive on parallel computers. This paper presents the results presented in June 1995 at the Seminario Matematico e Fisico organized by the Mathematics Department of the Polytechnics University of Milano.

The time‐dependent Maxwell equations are one of the most important approaches to describing dynamic or wide‐band frequency electromagnetic phenomena. A sequential finite‐volume, characteristic‐based procedure for solving the time‐dependent, three‐dimensional Maxwell equations has been successfully implemented in Fortran before. Due to its need for a large memory space and high demand on CPU time, it is impossible to test the code for a large array. Hence, it is essential to implement the code on a parallel computing system. In this paper, we discuss an efficient and scalable parallelization of the sequential Fortran time‐dependent Maxwell equations solver using High Performance Fortran (HPF). The background to the project, the theory behind the efficiency being achieved, the parallelization methodologies employed and the experimental results obtained on the Cray T3E massively parallel computing system will be described in detail. Experimental runs show that the execution time is reduced drastically through parallel computing. The code is scalable up to 98 processors on the Cray T3E and has a performance similar to that of an MPI implementation. Based on the experimentation carried out in this research, we believe that a high‐level parallel programming language such as HPF is a fast, viable and economical approach to parallelizing many existing sequential codes which exhibit a lot of parallelism. Copyright © 2003 John Wiley & Sons, Ltd.

Introduces the design and implementation of a Remote System for Parallel Computing.Based on Multi-platform.This system support remote operating and running linux system parallel programes on Windows system and provide a kind of uniform mode for the applied customer to operation, The customer does not need to know the concrete structure of the parallel computing system,can operate the parallel system and complete the computation or simulation what they need.This system reduces the complication on developing parallel programs and improves the efficiency on using parallel computing system,and solves the problem of real-time data transmission between server and client machine.It has highly practical value.

Parallel computer systems are among the most complex of man's creations, making satisfactory performance characterization difficult. Despite this complexity, there are strong, indeed, almost irresistible, incentives to quantify parallel system performance using a single metric. The fallacy lies in succumbing to such temptations. A complete performance characterization requires not only an analysis of the system's constituent levels, it also requires both static and dynamic characterizations. Static or average behavior analysis may mask transients that dramatically alter system performance. Although the human visual system is remarkedly adept at interpreting and identifying anomalies in false color data, the importance of dynamic, visual scientific data presentation has only recently been recognized Large, complex parallel system pose equally vexing performance interpretation problems. Data from hardware and software performance monitors must be presented in ways that emphasize important events while eluding irrelevant details. Design approaches and tools for performance visualization are the subject of this paper.

Stochastic computing (SC) is an attractive computing paradigm because of its fault tolerance and hardware efficiency. Unfortunately, existing SC systems suffer from large latency and inaccuracy problems. In this paper, we propose a new parallel stochastic computing (PSC) system which has higher computing accuracy and faster processing speed compared to existing SC systems. It uses the nibble serial data organization to reduce the latency and a combination of memory based weighted binary generator and data shuffling to improve the accuracy. Simulations on finite impulse response (FIR) filters show that the proposed system with 4 nibbles achieves 35% improvement in accuracy with about 3.5 times increase in processing speed compared to traditional SC. Like SC systems, PSC is also more tolerant of soft errors compared to conventional 2's complement implementations. For a 4-tap FIR filter a 10% injected error causes the root-mean-square-error (RMSE) of PSC to be 0.053 compared to RMSE of 0.188 for 2's complement implementation.