Project 1: Cache Architecture and Management Strategies for Smart Homes
Project Brief: This is a system modeling simulation project to develop:
-
Reconfigurable cache architecture to better suit various features of smart homes.
-
Cache architecture to support Real-time Artificial Intelligence.
-
Benchmark applications to test cache management strategies.
​
Status: Using unconventional system modeling simulation techniques, we have developed a strategy to improve prefetching accuracy in Ubuntu 16.04 Operating System. When the strategy is applied during real execution of memory-bound applications on Ubuntu 16.04 Operating System running on Processor intel i5-4690K, execution time can be reduced up to 50%. To test the superiority of the invented strategy, we have developed benchmark applications that can be used to measure prefetching efficiency of cache memory architectures for different memory subsystem setup and multiprocessing environment. We are currently writing two conference articles on these two topics.
​
Funding Source: Private Funding
​
Project Duration: 2 Years.
Commencement Date: Aug 01, 2017
​
Project Location: Singapore.
​
Related Articles:
-
Title: A Self-Reconfiguring Cache Architecture to Improve Control Quality in Cyber-Physical Systems. Authors: Mohammad Shihabul Haque, Sriram Vasudevan, Alamuri Sriram Nihar, Arvind Easwaran, Akash Kumar, Y.C. Tay. Conference: ISORC 2018.
