Masters Theses

Permanent URI for this collection

https://jewlscholar.mtsu.edu/handle/mtsu/3482

Browse

Browsing Masters Theses by Author "Aderholdt, Ferrol"
  • Item
    A COMPARISON OF PARALLEL IMPLEMENTATIONS OF PATHFINDING ALGORITHMS
    (Middle Tennessee State University, 2021) Johnson, Charles W ; Pettey, Chrisila ; Seo, Suk ; Aderholdt, Ferrol
    Pathfinding algorithms are in constant use today – be it for generating driving directions or maintaining the routing tables that are used for routing traffic around the Internet. As would be expected, a given algorithm’s performance varies depending upon both the type and the scale of the application. Additionally, the use of parallel computing techniques may or may not decrease the runtime for a given problem. This thesis examines three commonly used pathfinding algorithms – Dijkstra, Bellman-Ford, and Floyd-Warshall – and seeks to illustrate their strengths and weaknesses when analyzing graphs that range from small and sparse to large and dense in various parallel computing environments. As well, focus is given to the use of Graphics Processing Units (GPUs) as parallel computing devices themselves and the consequential reduction in the amount of hardware necessary for a given task.
  • Item
    An OpenSHMEM Runtime for UPC
    (Middle Tennessee State University, 2020) Lewis, Daniel Youree ; Aderholdt, Ferrol ; Phillips, Joshua ; Barbosa, Salvador
    Partitioned Global Address Space (PGAS) languages are becoming more important aswe move into the Exascale era. The complexities brought on by the architectures of thesemachines make traditional message passing approaches challenging. Currently, we lack aportable and flexible runtime for PGAS languages. The suggestion is that this is possiblewith an OpenSHMEM-based runtime. OpenSHMEM is a PGAS library for the C language aiming to provide a standardApplication Programming Interface (API) for Symmetric Hierarchical MEMory (SHMEM).OpenSHMEM is a portable, flexible, and performant API for PGAS runtimes throughits use of thread safety semantics and grouping of processes through teams. This allowsOpenSHMEM to make performance-oriented decisions on behalf of the language or librarybeing implemented. This thesis demonstrates a mapping and implementation of a Unified Parallel C (UPC)runtime based on OpenSHMEM, and an evaluation our runtime’s performance in comparisonto the popular Berkeley UPC implementation. Berkeley UPC uses a runtime based onGlobal-Address Space Networking (GASNet), a language independent middleware forPGAS programming languages.
  • Item
    Parallelizing A Content-Aware Image Resizing with OpenSHMEM
    (Middle Tennessee State University, 2020) Omotoso, Bukola Grace ; Aderholdt, Ferrol ; Seo, Suk ; Pettey, Chrisila
    ABSTRACT Cropping and Scaling are the most common image resizing techniques but neither of these considers the content of the image. Seam carving is a content-aware image resizing technique that removes or duplicates the least visible pixels from an image, thereby making it more effective than Cropping and Scaling. For large batches of images, it may be unrealistic to do Seam carving on one processing element due to memory constraints. To solve this problem, a parallel approach to image resizing that helps to ease data transfer between processing elements needs to be considered. Partitioned Global Address Space (PGAS) programming models have been attracting attention as a parallel computing model and it is often used to implement one-sided Remote Memory Access (RMA) from multi-host systems, such as computer clusters. OpenSHMEM is a distributed, PGAS programming model that has light-weight semantics and high performance RMA and atomic memory operations. In this thesis, we parallelize Seam carving using Pthreads, OpenSHMEM and MPI. We evaluate the relative performance gained with multiple threads and processing elements (PEs).