An efficient partitioning and scheduling algorithm for streaming applications on FPGA with resource constraint

Authors:
Radha Guha;Nader Bagherzadeh
Affiliations:
University of California, Irvine, California;University of California, Irvine, California
Venue:
PDCS '07 Proceedings of the 19th IASTED International Conference on Parallel and Distributed Computing and Systems
Year:
2007

Citing 5
Cited 0

A HW/SW partitioning algorithm for dynamically reconfigurable architectures

Proceedings of the conference on Design, automation and test in Europe
Hardware-software cosynthesis for run-time incrementally reconfigurable FPGAs

ASP-DAC '00 Proceedings of the 2000 Asia and South Pacific Design Automation Conference
Design Methodology for a Tightly Coupled VLIW/Reconfigurable Matrix Architecture: A Case Study

Proceedings of the conference on Design, automation and test in Europe - Volume 2
Efficient search space exploration for HW-SW partitioning

Proceedings of the 2nd IEEE/ACM/IFIP international conference on Hardware/software codesign and system synthesis
Physically-aware HW-SW partitioning for reconfigurable architectures with partial dynamic reconfiguration

Proceedings of the 42nd annual Design Automation Conference

Quantified Score

Hi-index	0.00

Visualization

Abstract

Hardware and software partitioning and reconfiguration cost minimization are two major design challenges for the performance improvement of data-intensive streaming applications by hardware and software co-design approach. In this paper a novel partitioning and scheduling algorithm is proposed for a hybrid system of a microprocessor and a reconfigurable hardware with resource constraint to optimize between hardware resource usage and execution time of an application. We balance the exploitation of spatial parallelism and temporal parallelism in streaming applications by considering the reconfiguration cost vs. the data transfer cost. As the reconfiguration cost is configuration bits size plus the configuration time we try to keep it minimum by exploiting the temporal parallelism maximally.