PROJECT TITLE :

Atomic Streaming: A Framework of On-Chip Data Supply System for Task-Parallel MPSoCs

ABSTRACT:

State of the art fabrication technology for integrating varied hardware resources like Processors/DSPs and memory arrays into a single chip allows the emergence of Multiprocessor System-on-Chip (MPSoC). Stream programming paradigm based mostly on MPSoC is highly efficient for single functionality state of affairs due to its dedicated and predictable information offer system. But, when memory traffic is heavily shared among parallel tasks in applications with multiple interrelated functionalities, performance suffers through task interferences and shared memory congestions that cause poor parallel speedups and memory bandwidth utilizations. This paper proposes a framework of Stream Processing based on-chip data provide system for task-parallel MPSoCs. During this framework, stream address generations and data computations are decoupled and parallelized to permit full utilization of on-chip resources. Task granularities are dynamically tuned to jointly optimize the overall application performance. Experiments show that proposed framework furthermore because the tuning scheme are effective for joint optimization in task-parallel MPSoCs.