To ensure increased performance of multiprocessor system on chip (MPSoC), several research has been done in the academic or industrial setting, the majority of these searches are based on simulation platforms with simple architectures. Seen that architecture memory has a direct impact on system performance ,the main objective of this article is to make a comparison between different approaches for memory architectures, architectures with shared memory, distributed memory and distributed shared memory, in order to determine the most suitable for which we will base the new architecture of open-source platform , Library of Power and Temperature at the Transaction-Level- Modeling (LIBTLMPWT), this platform has the advantage of calculating the energy consumption of each component using SystemC /TLM, which ensures the speed of the system with a high level of abstraction.
Volume 11 | 11-Special Issue
Pages: 1067-1071
DOI: 10.5373/JARDCS/V11SP11/20193138