Heat Analysis Using Gradient Descent with Multicore Numerical Integration Computation

The high-order numerical integration method known as Weddle's rule can be used to compute integrals precisely. To maximize the performance of Weddle's rule on contemporary processors, the research will investigate the usage of multiprocessing. The integration of function is only one symbolic representation to demonstrate any computational task that will involve large datasets and complex computations. The efforts in computation can be any function instead of this one that is involving load on the server facilitating MOOCs services. Utilizing multiprocessing and parallel computing techniques can significantly enhance the efficiency and scalability of these computations, leading to several real-life server management issues. This study analyzes the heat generated and time taken during the execution of the Integration Process. A different set of processors (1–12) will be used to solve the considered integral problem. Time consumed and heat generated will be noted for each process in regard to number of processors used. The greater number of processors used the more time taken for lesser number of inputs. After that Machine Learning technique Gradient Descent is used to predict heat generated using features like time taken, input size, and Number of processors. In other words this work aims to demonstrate a heat analysis process that is optimized using gradient descent and involves numerical integration computations performed on a multicore system to handle complex calculations efficiently.