University of Rhode Island DigitalCommons@URI Open Access Master's Theses 2013 On Performance of GPU and DSP Architectures for Computationally Intensive Applications John Faella University of Rhode Island,
[email protected] Follow this and additional works at: https://digitalcommons.uri.edu/theses Recommended Citation Faella, John, "On Performance of GPU and DSP Architectures for Computationally Intensive Applications" (2013). Open Access Master's Theses. Paper 2. https://digitalcommons.uri.edu/theses/2 This Thesis is brought to you for free and open access by DigitalCommons@URI. It has been accepted for inclusion in Open Access Master's Theses by an authorized administrator of DigitalCommons@URI. For more information, please contact
[email protected]. ON PERFORMANCE OF GPU AND DSP ARCHITECTURES FOR COMPUTATIONALLY INTENSIVE APPLICATIONS BY JOHN FAELLA A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENE IN ELECTRICAL ENGINEERING UNIVERSITY OF RHODE ISLAND 2013 MASTER OF SCIENCE THESIS OF JOHN FAELLA APPROVED: Thesis Committee: Major Professor Dr. Jien-Chung Lo Dr. Resit Sendag Dr. Lutz Hamel Nasser H. Zawia DEAN OF THE GRADUATE SCHOOL UNIVERSITY OF RHODE ISLAND 2013 ABSTRACT This thesis focuses on the implementations of a support vector machine (SVM) algorithm on digital signal processor (DSP), graphics processor unit (GPU), and a common Intel i7 core architecture. The purpose of this work is to identify which of the three is most suitable for SVM implementation. The performance is measured by looking at the time required by each of the architectures per prediction. This work also provides an analysis of possible alternatives to existing implementations of computationally intensive algorithms, such as SVM.