Abstract
This paper deals with the problem of thread synchronisation in manycore systems. In particular, it considers the open-source GPU-like architecture developed within the MANGO H2020 project. The thread synchronisation hardware relies on a distributed master and on a lightweight control unit to be deployed within the core. It does not rely on memory access for exchanging synchronisation information since it uses hardware-level messages. The solution supports multiple barriers for different application kernels possibly being executed simultaneously. The results for different NoC sizes provide indications about the reduced synchronisation times and area overheads incurred by our solution.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: International Journal of Grid and Utility Computing
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
