Type of Document	Thesis
Author	Ren, Jingyuan
URN	etd-11082006-195809
Title	Design and Implementation of Scalable Parallel Random Number Generators Library 4.0
Degree	Master of Science
Department	Computer Science, Department of
Advisory Committee	Advisor Name Title Michael Mascagni Committee Chair Ashok Srinivasan Committee Member Piyush Kumar Committee Member
Keywords	Random Number Generators Pseudorandom Numbers Random Numbers SPRNG
Date of Defense	2006-11-06
Availability	unrestricted
Abstract This thesis presents the Scalable Parallel Random Number Generators Library (SPRNG) version 4.0. SPRNG is a widely used software package for both serial and parallel pseudorandom number generation. Its generators satisfy every known property of good pseudorandom numbers, and have passed some of the most stringent tests for pseudo-randomness available today. The previous versions of SPRNG used C as the implementation language and provided interfaces for FORTRAN and C++. Although SPRNG functions were callable from C++, the concepts of object-oriented programming were not applied in the implementation of earlier versions of SPRNG. In the new version described here, we replace the implementation language with object oriented C++, while still providing an interface for FORTRAN. Furthermore, we remove SPRNG’s dependence on the GNU Multi Precision (GMP) Library and supplant GMP with our own code for arithmetic on arbitrarily large integers and rational numbers. This additional code is tested for both accuracy and efficiency, and the results appear here. Thus, a new and more versatile version of SPRNG is the result of this work. SPRNG and its documentation can be downloaded from http://sprng.fsu.edu.
Files	Filename       Size       Approximate Download Time (Hours:Minutes:Seconds)     28.8 Modem   56K Modem   ISDN (64 Kb)   ISDN (128 Kb)   Higher-speed Access    final_ms_thesis.pdf 639.13 Kb 00:02:57 00:01:31 00:01:19 00:00:39 00:00:03