Title :
Fragmentation aware placement algorithm for a reconfigurable system
Author :
Esmaeildoust, Mohammad ; Fazlali, Mahmood ; Zakerolhosseini, Ali ; Karimi, Mohsen
Author_Institution :
Dept. of Comput. Eng., Shahid Beheshti Univ., Tehran
Abstract :
Partially runtime reconfiguration allows runtime mapping and execution of several tasks on a FPGA. In order to enhance the efficiency of the reconfigurable computing, multitasking on FPGAs is proposed. One of the challenging problems in multitasking systems is online task placement. This paper focuses on task placement in a reconfigurable system. Several techniques for nonrectangular tasks are investigated and an algorithm that decreases the fragmentation of a FPGA is presented. The algorithm is able to locate a feasible location with up to 26% improvement in the fragmentation of a FPGA, with a lower rejection rate compared to other placement algorithms.
Keywords :
field programmable gate arrays; FPGA; fragmentation aware placement algorithm; online task placement; reconfigurable system; runtime mapping; Field programmable gate arrays; Hardware; Heuristic algorithms; Logic devices; Multitasking; Partitioning algorithms; Reconfigurable logic; Resource management; Runtime; Shape;
Conference_Titel :
Electrical Engineering, 2008. ICEE 2008. Second International Conference on
Conference_Location :
Lahore
Print_ISBN :
978-1-4244-2292-0
Electronic_ISBN :
978-1-4244-2293-7
DOI :
10.1109/ICEE.2008.4553918