Abstract—Complex numerical computations often use repeated arithmetic operations. Pipelining is an architectural approach for speeding up the floating-point arithmetic operations. In pipelining, a task is partitioned into simpler computational steps that can be executed concurrently. In high performance computing, the multi-core technology has emerged as one of the major technologies that support such pipeline architecture. In this paper, the performance of Intel Dual Core processor has been evaluated for floating-point arithmetic operations. A common software program computing floating-point arithmetic has been developed in Visual C++ and Visual C# languages and the performance of these programs is measured on Intel Dual Core processor. To demonstrate the performance, a well known Unified Modeling Language (UML) is used for modeling of floating-point nonlinear pipeline. The UML class, activity and sequence diagrams are designed and results are presented in the form of tables and graphs.

Index Terms—Nonlinear Pipeline, Dual Core Processor,Floating-point, UML class diagram, UML sequence diagram

Cite: Vipin Saxena and Manish Shrivastava, "Performance Evaluation of Nonlinear Pipeline through UML," International Journal of Computer and Electrical Engineering vol. 2, no. 5, pp. 860-866, 2010.