Paul Werstein

Load Balancing in a Cluster Computer

This paper proposes a load balancing algorithm for distributed use of a cluster computer. It uses load information including CPU queue length, CPU utilisation, memory utilisation and network traffic to decide the load of each node. This algorithm is compared to an algorithm using only the CPU queue length. The performance evaluation results show that the proposed algorithm performs well

Performance evaluation of view-oriented parallel programming

This paper evaluates the performance of a novel view-oriented parallel programming style for parallel programming on cluster computers. View-oriented parallel programming is based on distributed shared memory which is friendly and easy for programmers to use. It requires the programmer to divide shared data into views according to the memory access pattern of the parallel algorithm. One of the advantages of this programming style is that it offers the performance potential for the underlying distributed shared memory system to optimize consistency maintenance. Also it allows the programmer to participate in performance optimization of a program through wise partitioning of the shared data into views. Experimental results demonstrate a significant performance gain of the programs based on the view-oriented parallel programming style.

GPU as a General Purpose Computing Resource

In the last few years, GPUs(Graphics Processing Units) have made rapid development. Their ever-increasing computing power and decreasing cost have attracted attention from both industry and academia. In addition to graphics applications, researchers are interested in using them for general purpose computing. Recently, NVIDIA released a new computing architecture, CUDA (compute united device architecture), for its GeForce 8 series, Quadro FX, and Tesla GPU products. This new architecture can change fundamentally the way in which GPUs are used. In this paper, we study the programmability of CUDA and its GeForce 8 GPU and compare its performance with general purpose processors, in order to investigate its suitability for general purpose computation.

A performance comparison of DSM, PVM, and MPI

We compare the performance of the Treadmarks DSM system with two popular message passing systems (PVM and MPI). The comparison is done on 1, 2, 4, 8, 16, 24, and 32 nodes. Applications are chosen to represent three classes of problems: loosely synchronous, embarrassingly parallel, and synchronous. The results show DSM has similar performance to message passing for the embarrassingly parallel class. However the performance of DSM is lower than PVM and MPI for the synchronous and loosely synchronous classes of problems. An analysis of the reasons is presented.

A Remote Memory Swapping System for Cluster Computers

This paper describes the use of remote memory for virtual memory swapping in a cluster computer. Our design uses a lightweight kernel-to-kernel communications channel for fast, efficient data transfer. Performance tests are made to compare our system to normal hard disk swapping. The tests show significantly improved performance when data access is random.

Rule-based consistency enforcement for knowledge-based systems

A rule-based approach for the automatic enforcement of consistency constraints is presented. In contrast to existing approaches that compile consistency checks into application programs, the approach centralizes consistency enforcement in a separate module called a knowledge-base management system. Exception handlers for constraint violations are represented as rule entities in the knowledge base. For this purpose, a new form of production rule called the activation pattern controlled rule is introduced: in contrast to classical forward chaining schemes, activation pattern controlled rules are triggered by the intent to apply a specific operation but not necessarily by the result of applying this operation. Techniques for implementing this approach are discussed, and experiments in speeding up the system performance are described. Furthermore, an argument is made for more tolerant consistency enforcement strategies, and how they can be integrated into the rule-based approach to consistency enforcement is discussed. >

View-Oriented parallel programming and view-based consistency

This paper proposes a novel View-Oriented Parallel Programming style for parallel programming on cluster computers. View-Oriented Parallel Programming is based on Distributed Shared Memory. It requires the programmer to divide the shared memory into views according to the nature of the parallel algorithm and its memory access pattern. The advantage of this programming style is that it can help the Distributed Shared Memory system optimise consistency maintenance. Also it allows the programmer to participate in performance optimization of a program through wise partitioning of the shared memory into views. The View-based Consistency model and its implementation, which supports View-Oriented Parallel Programming, is discussed as well in this paper. Finally some preliminary experimental results are shown to demonstrate the performance gain of View-Oriented Parallel Programming.

Maotai 2.0: Data Race Prevention in View-Oriented Parallel Programming

This paper proposes a data race prevention scheme, which can prevent data races in the View-Oriented Parallel Programming (VOPP) model. VOPP is a novel shared-memory data-centric parallel programming model, which uses views to bundle mutual exclusion with data access. We have implemented the data race prevention scheme with a memory protection mechanism. Experimental results show that the extra overhead of memory protection is trivial in our applications. We also present a new VOPP implementation--Maotai 2.0, which has advanced features such as deadlock avoidance, producer/consumer view and system queues, in addition to the data race prevention scheme. The performance of Maotai 2.0 is evaluated and compared with modern programming models such as OpenMP and Cilk.

Integration of rules into a knowledge base management system

The centralization and integration of the data management functions of large computerized knowledge bases into a module called the knowledge base management system (KBMS), a first step on the way to large, integrated knowledge bases, are discussed. One prerequisite for the success of this approach is the feasibility of integrating large sets of rules that share the same knowledge base. Rule integration has to cope with objects that have inferential/operational capabilities. The conceptual and implementation problems of such an integration are discussed. Experiences in integrating rules into a prototype KBMS called DALI are surveyed. For this integration, a new form of production rule called the activation pattern controlled rule which plays a key role in the system considered, is introduced. Such rules are triggered by a call to the KBMS; that is, by the intention to apply a certain operation but not necessarily by the result of applying the operation. It is demonstrated that exception handling mechanisms can easily be represented when using this form of rule.<<ETX>>

Data race: tame the beast

Data races hamper parallel programming and threaten the reliability of future software. This paper proposes the data race prevention scheme View-Oriented Data race Prevention (VODAP), which can prevent data races in the View-Oriented Parallel Programming (VOPP) model. VOPP is a novel shared-memory data-centric parallel programming model, which uses views to bundle mutual exclusion with data access. We have implemented the data race prevention scheme with a memory protection mechanism. Experimental results show that the extra overhead of memory protection is trivial in our applications. The performance is evaluated and compared with modern programming models such as OpenMP and Cilk.