Predicting the Performance Measures of a 2-Dimensional Message Passing Multiprocessor Architecture by Using Machine Learning Methods

Abstract

2-dimensional Simultaneous Optical Multiprocessor Exchange Bus (2D SOME-Bus) is a reliable, robust implementation of petaflops-performance computer architecture. In this paper, we develop models to predict the performance measures (i.e. average channel utilization, average channel waiting time, average network latency, average processor utilization and average input waiting time) of a message passing architecture interconnected by the 2D SOME-Bus by using Multilayer Feed-forward Artificial Neural Network (MFANN), Support Vector Regression (SVR) and Multiple Linear Regression (MLR). OPNET Modeler is used to simulate the message passing 2D SOME-Bus multiprocessor architecture and to create the training and testing datasets. Using 10-fold cross validation, the performance of the prediction models have been evaluated using several performance metrics. The results show that the SVR model using the radial basis function kernel (SVR-RBF) yields the lowest prediction error among all models.

An Intelligent Edge-centric Queries Allocation Scheme based on Ensemble Models

“…The interesting is that the statistical error in our measurements is also limited (for the majority of the scenarios) exhibiting the stability of our model. In this set of our experiments, the CS exhibits the best performance compared to the MV S. The next set of experiments deals with the dataset provided by [4]. Our results are presented in Figures 11 and 12.…”

“…For the generation of each value, we consider an maximum a equal to 10. [4]. From this dataset, we adopt the processor utilization dimension that depicts the percent of time that threads are running in a processor.…”

Autonomous Data Density pruning fuzzy neural network for Optical Interconnection Network

“…Researchers in the artificial intelligence field commonly study the techniques to predict performance measures of a multiprocessor architecture. In recent years, artificial intelligence and machine learning techniques have been applied in parallel computing as in Akay et al [3,4], Zayid et al [79] and Aci et al [2]. Other intelligent techniques such as clustering assistance in multiprocessors evaluation [59], a framework proposed by [54], sorting and matrix operations [44], a serial synchronized multistage interconnection [77], Hopfield neural network [27].…”

“…Parallel computing processing was performed using simulations with a software called OPNET Modeler [4]. This software simulates the passage of pensions, the procedure used as the mechanism of communication, in which any processor can send to the network a point-to-point message destined to any other processor.…”

“…Works were performed to predict the 2D SOME-Bus multiprocessor architecture using the message transmission programming model [2], for such a database was created for training and testing of intelligent models using OPNET Modeler. It has as the main feature to simulate the 2D SOME-Bus pass-through multiprocessor architecture, thus allowing the collected values to reflect the communication behavior of real systems that communicate in parallel [4].…”