DocumentCode
2430865
Title
High-Level Power Estimation Model for SOC with FPGA Prototyping
Author
Wang, Lijia ; Wang, Xin´an ; Wang, Teng ; Yang, Qihua
Author_Institution
Key Lab. of Integrated Microsyst. Sci. Eng. & Applic., Peking Univ., Shenzhen, China
fYear
2012
fDate
3-5 Nov. 2012
Firstpage
491
Lastpage
495
Abstract
Power consumption reduction has become a first-order requirement for modern IC design, and techniques of high level power estimation are urgently in need for the sake of achieving maximal power reduction with minimal cost by optimizing the design early in the design process. This paper presents a high level power estimation model for system-on-chip (SOC) architectures on FPGA. Working with a system-level SOC synthesis framework, the proposed model takes primary power data as foundation and power estimation is conducted with regard to major event signatures such as bus and memory read/write. Besides, the anomalistic change of LUT numbers in the hardware implementation process is then taken into account, which allows more accurate and efficient power estimation providing invaluable directions for further optimization process. With the proposed model, SOC architectures are designed, and power estimation results of the model are compared with the results from real measurements on a Xilinx Virtex-5 FPGA board.
Keywords
field programmable gate arrays; logic design; network synthesis; power consumption; system-on-chip; IC design; SOC; Xilinx Virtex-5 FPGA board; anomalistic change; high-level power estimation model; power consumption reduction; system-on-chip; Computational modeling; Estimation; Field programmable gate arrays; Integrated circuit modeling; Power demand; System-on-a-chip; Table lookup; FPGA; High level Power Estimation; SoC; event signature;
fLanguage
English
Publisher
ieee
Conference_Titel
Computational Intelligence and Communication Networks (CICN), 2012 Fourth International Conference on
Conference_Location
Mathura
Print_ISBN
978-1-4673-2981-1
Type
conf
DOI
10.1109/CICN.2012.124
Filename
6375163
Link To Document