Solutions by Optimization of the 2-Dimensional Heat Conductivity Problem on Grid Machines using C++ and OpenMP
Aliyu Sani, Mu'awuya Dalhatu, Emmanuel Masa-Ibi, Abdulrashid Aliyu · International Journal of Computer Applications · 2020
Optimization is a method of transforming a piece of code into another functionally equivalent piece of code to improve code quality and efficiency.The two most important characteristics are the speed and size of the code.Other characteristics include: energy required and the time it takes to compile the code.This paper uses source code (jacobi2d.c)to provide solutions by optimization of the 2-Dimentional heat conductivity problem on grid machine using C++ compiler and OpenMP (Open Multi-Processing).The paper discuss the benefits of parallel grid computing.The grid server on which this paper used has 10 computers, grid-01 to grid-08 have 2 cores and 2 Gigabytes memory while 09 and 10 have 8 cores and 8 Gigabytes memory.The process of solving rectangular 2-dimensional heat conductivity problem to optimized performance is done in number of steps from step 1 to step 4, optimization starts in step 3 and ends in step 4. In step 4, performance speedup of step 3 is compared with performance speedup of step 4 and optimal (result) performance is achieved in step 4. It is concluded that performance is optimized with parallel processing using grid machine with performance level -3 and large problem size.