Publication Type : Conference Paper
Publisher : IEEE
Source : 2025 International Conference on Robotics and Mechatronics (ICRM)
Url : https://doi.org/10.1109/icrm66809.2025.11349018
Campus : Amritapuri
School : School of Engineering
Center : Humanitarian Technology (HuT) Labs
Department : Electronics and Communication
Year : 2025
Abstract : In modern CPUs, predicting which way a branch will go helps improve speed and reduce power usage. This paper shows a smart system that combines a basic 2-bit counter and an advanced AI-based method using a Bayesian Neural Network with Deep Q-Learning (BNN-DQN). If the branch is easy to guess, the system uses the simple method. If it’s hard, it switches to the smart method. This helps save energy while keeping accuracy high. We tested this using ChampSim and found that our system works much better than just using the simple method. It improves accuracy by about 2% and uses less energy overall.
Cite this Research Publication : Abhinav Manoj Nair, Arjun Pananjikkal Sujeendran, Rajesh Kannan Megalingam, Branch Prediction Using Reinforcement Learning, 2025 International Conference on Robotics and Mechatronics (ICRM), IEEE, 2025, https://doi.org/10.1109/icrm66809.2025.11349018