DocumentCode
2922581
Title
ILP formulation for hybrid FPGA MPSoCs optimizing performance, area and memory usage
Author
Sotiropoulou, Calliope-Louisa ; Nikolaidis, Spyridon
Author_Institution
Dept. of Phys., Aristotle Univ. of Thessaloniki, Thessaloniki, Greece
fYear
2011
fDate
11-14 Dec. 2011
Firstpage
748
Lastpage
751
Abstract
In this paper an integer linear programming (ILP) model is proposed for identifying the optimal resources allocation and application partitioning on FPGA-based hybrid MPSoC, including the memory allocation for each processing unit and total BRAM usage on the device. The motivation behind this work is the reduction of the necessary exploration time for the identification of the optimal hybrid MPSoC architecture implementation for each particular application. Whilst many ILP model approaches have been proposed, none of the existing models presents an accurate calculation model for the BRAM usage on the FPGA device and can impose the memory usage and/or BRAM usage as a design constraint. Our work comes to fill this void, especially now that modern multimedia applications have an ever increasing need for memory space and data throughput.
Keywords
field programmable gate arrays; integer programming; linear programming; multiprocessing systems; performance evaluation; system-on-chip; BRAM; ILP; ILP formulation; hybrid FPGA MPSoCs optimizing performance; hybrid MPSoC architecture; integer linear programming; memory space; memory usage; multimedia applications; Data models; Field programmable gate arrays; Hardware; Memory management; Microprocessors; Multiprocessing systems; Program processors;
fLanguage
English
Publisher
ieee
Conference_Titel
Electronics, Circuits and Systems (ICECS), 2011 18th IEEE International Conference on
Conference_Location
Beirut
Print_ISBN
978-1-4577-1845-8
Electronic_ISBN
978-1-4577-1844-1
Type
conf
DOI
10.1109/ICECS.2011.6122382
Filename
6122382
Link To Document