Title :
HTG prediction algorithm for SoC design
Author :
Chen, Ruei-Xi ; Kao, Jian-Jyh ; Fan, Jeffrey
Author_Institution :
Comput. Sci. & Inf. Eng. Dept., St. John´´s Univ., Taipei, Taiwan
Abstract :
Efficient design methodology is the key toward success for System-on-a-Chip (SoC) design. In this paper, we proposed a design methodology based on the Hot-Task Group (HTG) to improve the design efficiency. The method is using a fitness function to collect the valuable functionality task of applications, and then extracts the best HTG for hardware design consideration. This method not only significantly reduces the design complexity but also avoid the design traps that traditional methods encountered. A study case of H.264 encoder reveals the design efficiency.
Keywords :
integrated circuit design; system-on-chip; H.264 encoder; HTG prediction algorithm; SoC design; design complexity reduction; hardware design consideration; hot-task group prediction algorithm; system-on-a-chip design; Algorithm design and analysis; Complexity theory; Design methodology; Hardware; Partitioning algorithms; Prediction algorithms; System-on-a-chip; H.264 encoder; HW/SW partion; Hot-Task group; fitness function;
Conference_Titel :
System Theory (SSST), 2012 44th Southeastern Symposium on
Conference_Location :
Jacksonville, FL
Print_ISBN :
978-1-4577-1492-4
DOI :
10.1109/SSST.2012.6195158