Memory management and parallelization of data intensive all-to-all comparison in shared-memory systems

This thesis presents a novel program parallelization technique incorporating with dynamic and static scheduling. It utilizes a problem specific pattern developed from the prior knowledge of the targeted problem abstraction. Suitable for solving complex parallelization problems such as data intensive all-to-all comparison constrained by memory, the technique delivers more robust and faster task scheduling compared to the state-of-the art techniques. Good performance is achieved from the technique in data intensive bioinformatics applications.

Parallelization of an unstructured data based cell centre finite volume code, HIFUN-3D

This work describes the parallelization of High Resolution flow solver on unstructured meshes, HIFUN-3D, an unstructured data based finite volume solver for 3-D Euler equations. For mesh partitioning, we use METIS, a software based on multilevel graph partitioning. The unstructured graph used for partitioning is associated with weights both on its vertices and edges. The data residing on every processor is split into four layers. Such a novel procedure of handling data helps in maintaining the effectiveness of the serial code. The communication of data across the processors is achieved by explicit message passing using the standard blocking mode feature of Message Passing Interface (MPI). The parallel code is tested on PACE++128 available in CFD Center

An automatic parallelization framework for multicomputers

Several researchers have looked into various issues related to automatic parallelization of sequential programs for multicomputers. But there is a need for a coherent framework which encompasses all these issues. In this paper we present a such a framework which takes best advantage of the multicomputer architecture. We resort to tiling transformation for iteration space partitioning and propose a scheme of automatic data partitioning and dynamic data distribution. We have tried a simple implementation of our scheme on a transputer based multicomputer [1] and the results are encouraging.

Online Unsupervised Pattern Discovery in Speech Using Parallelization

Segmental dynamic time warping (DTW) has been demonstrated to be a useful technique for finding acoustic similarity scores between segments of two speech utterances. Due to its high computational requirements, it had to be computed in an offline manner, limiting the applications of the technique. In this paper, we present results of parallelization of this task by distributing the workload in either a static or dynamic way on an 8-processor cluster and discuss the trade-offs among different distribution schemes. We show that online unsupervised pattern discovery using segmental DTW is plausible with as low as 8 processors. This brings the task within reach of today's general purpose multi-core servers. We also show results on a 32-processor system, and discuss factors affecting scalability of our methods.

Integrated parallelization of computations and visualization for large-scale applications

Critical applications like cyclone tracking and earthquake modeling require simultaneous high-performance simulations and online visualization for timely analysis. Faster simulations and simultaneous visualization enable scientists provide real-time guidance to decision makers. In this work, we have developed an integrated user-driven and automated steering framework that simultaneously performs numerical simulations and efficient online remote visualization of critical weather applications in resource-constrained environments. It considers application dynamics like the criticality of the application and resource dynamics like the storage space, network bandwidth and available number of processors to adapt various application and resource parameters like simulation resolution, simulation rate and the frequency of visualization. We formulate the problem of finding an optimal set of simulation parameters as a linear programming problem. This leads to 30% higher simulation rate and 25-50% lesser storage consumption than a naive greedy approach. The framework also provides the user control over various application parameters like region of interest and simulation resolution. We have also devised an adaptive algorithm to reduce the lag between the simulation and visualization times. Using experiments with different network bandwidths, we find that our adaptive algorithm is able to reduce lag as well as visualize the most representative frames.

Decoherence suppression by parallelization

Decoherence as an obstacle in quantum computation is viewed as a struggle between two forces [1]: the computation which uses the exponential dimension of Hilbert space, and decoherence which destroys this entanglement by collapse. In this model of decohered quantum computation, a sequential quantum computer loses the battle, because at each time step, only a local operation is carried out but g*(t) number of gates collapse. With quantum circuits computing in parallel way the situation is different- g(t) number of gates can be applied at each time step and number gates collapse because of decoherence. As g(t) ≈ g*(t) competition here is even [1]. Our paper improves on this model by slowing down g*(t) by encoding the circuit in parallel computing architectures and running it in Single Instruction Multiple Data (SIMD) paradigm. We have proposed a parallel ion trap architecture for single-bit rotation of a qubit.

Hybrid aggregated-vector algorithm for efficient parallelization of fast multipole method

An efficient parallelization algorithm for the Fast Multipole Method which aims to alleviate the parallelization bottleneck arising from lower job-count closer to root levels is presented. An electrostatic problem of 12 million non-uniformly distributed mesh elements is solved with 80-85% parallel efficiency in matrix setup and matrix-vector product using 60GB and 16 threads on shared memory architecture.

Compiler/Runtime Framework for Dynamic Dataflow Parallelization of Tiled Programs

Task-parallel languages are increasingly popular. Many of them provide expressive mechanisms for intertask synchronization. For example, OpenMP 4.0 will integrate data-driven execution semantics derived from the StarSs research language. Compared to the more restrictive data-parallel and fork-join concurrency models, the advanced features being introduced into task-parallelmodels in turn enable improved scalability through load balancing, memory latency hiding, mitigation of the pressure on memory bandwidth, and, as a side effect, reduced power consumption. In this article, we develop a systematic approach to compile loop nests into concurrent, dynamically constructed graphs of dependent tasks. We propose a simple and effective heuristic that selects the most profitable parallelization idiom for every dependence type and communication pattern. This heuristic enables the extraction of interband parallelism (cross-barrier parallelism) in a number of numerical computations that range from linear algebra to structured grids and image processing. The proposed static analysis and code generation alleviates the burden of a full-blown dependence resolver to track the readiness of tasks at runtime. We evaluate our approach and algorithms in the PPCG compiler, targeting OpenStream, a representative dataflow task-parallel language with explicit intertask dependences and a lightweight runtime. Experimental results demonstrate the effectiveness of the approach.

Scaling the iHMM: Parallelization versus Hadoop

This paper compares parallel and distributed implementations of an iterative, Gibbs sampling, machine learning algorithm. Distributed implementations run under Hadoop on facility computing clouds. The probabilistic model under study is the infinite HMM [1], in which parameters are learnt using an instance blocked Gibbs sampling, with a step consisting of a dynamic program. We apply this model to learn part-of-speech tags from newswire text in an unsupervised fashion. However our focus here is on runtime performance, as opposed to NLP-relevant scores, embodied by iteration duration, ease of development, deployment and debugging. © 2010 IEEE.

Structure Sharing and Parallelization in a GB Parser

By utilizing structure sharing among its parse trees, a GB parser can increase its efficiency dramatically. Using a GB parser which has as its phrase structure recovery component an implementation of Tomita's algorithm (as described in [Tom86]), we investigate how a GB parser can preserve the structure sharing output by Tomita's algorithm. In this report, we discuss the implications of using Tomita's algorithm in GB parsing, and we give some details of the structuresharing parser currently under construction. We also discuss a method of parallelizing a GB parser, and relate it to the existing literature on parallel GB parsing. Our approach to preserving sharing within a shared-packed forest is applicable not only to GB parsing, but anytime we want to preserve structure sharing in a parse forest in the presence of features.

A scalable strategy for the parallelization of multiphysics unstructured mesh-iterative codes on distributed-memory systems

Realizing scalable performance on high performance computing systems is not straightforward for single-phenomenon codes (such as computational fluid dynamics [CFD]). This task is magnified considerably when the target software involves the interactions of a range of phenomena that have distinctive solution procedures involving different discretization methods. The problems of addressing the key issues of retaining data integrity and the ordering of the calculation procedures are significant. A strategy for parallelizing this multiphysics family of codes is described for software exploiting finite-volume discretization methods on unstructured meshes using iterative solution procedures. A mesh partitioning-based SPMD approach is used. However, since different variables use distinct discretization schemes, this means that distinct partitions are required; techniques for addressing this issue are described using the mesh-partitioning tool, JOSTLE. In this contribution, the strategy is tested for a variety of test cases under a wide range of conditions (e.g., problem size, number of processors, asynchronous / synchronous communications, etc.) using a variety of strategies for mapping the mesh partition onto the processor topology.

Bulk synchronous parallelization of industrial electromagentic software

The parallelization of existing/industrial electromagnetic software using the bulk synchronous parallel (BSP) computation model is presented. The software employs the finite element method with a preconditioned conjugate gradient-type solution for the resulting linear systems of equations. A geometric mesh-partitioning approach is applied within the BSP framework for the assembly and solution phases of the finite element computation. This is combined with a nongeometric, data-driven parallel quadrature procedure for the evaluation of right-hand-side terms in applications involving coil fields. A similar parallel decomposition is applied to the parallel calculation of electron beam trajectories required for the design of tube devices. The BSP parallelization approach adopted is fully portable, conceptually simple, and cost-effective, and it can be applied to a wide range of finite element applications not necessarily related to electromagnetics.

An environment for OpenMP code parallelization

This chapter discusses the code parallelization environment, where a number of tools that address the main tasks, such as code parallelization, debugging, and optimization are available. The parallelization tools include ParaWise and CAPO, which enable the near automatic parallelization of real world scientific application codes for shared and distributed memory-based parallel systems. The chapter discusses the use of ParaWise and CAPO to transform the original serial code into an equivalent parallel code that contains appropriate OpenMP directives. Additionally, as user involvement can introduce errors, a relative debugging tool (P2d2) is also available and can be used to perform near automatic relative debugging of an OpenMP program that has been parallelized either using the tools or manually. In order for these tools to be effective in parallelizing a range of applications, a high quality fully inter-procedural dependence analysis, as well as user interaction is vital to the generation of efficient parallel code and in the optimization of the backtracking and speculation process used in relative debugging. Results of parallelized NASA codes are discussed and show the benefits of using the environment.

Generating OpenMP code using an interactive parallelization environment

Code parallelization using OpenMP for shared memory systems is relatively easier than using message passing for distributed memory systems. Despite this, it is still a challenge to use OpenMP to parallelize application codes in a way that yields effective scalable performance when executed on a shared memory parallel system. We describe an environment that will assist the programmer in the various tasks of code parallelization and this is achieved in a greatly reduced time frame and level of skill required. The parallelization environment includes a number of tools that address the main tasks of parallelism detection, OpenMP source code generation, debugging and optimization. These tools include a high quality, fully interprocedural dependence analysis with user interaction capabilities to facilitate the generation of efficient parallel code, an automatic relative debugging tool to identify erroneous user decisions in that interaction and also performance profiling to identify bottlenecks. Finally, experiences of parallelizing some NASA application codes are presented to illustrate some of the benefits of using the evolving environment.