Efficient two-layered cycle-accurate modeling technique for processor family with same instruction set architecture

Chien De Chiang*, Juinn-Dar Huang

*此作品的通信作者

    研究成果: Conference contribution同行評審

    1 引文 斯高帕斯(Scopus)

    摘要

    In this paper, we propose a new processor modeling technique that partitions a cycle-accurate model into two layers, an inner functional kernel and an outer timing shell. The kernel is an untimed but high-speed instruction set simulator (ISS) and is suitable for software development; while the timing shell provides additional timing details for cycle-accurate hardware behavior. When a new processor member is added to the family, it demands only a new timing shell because the kernel is identical to that of its ancestors sharing the same instruction set architecture (ISA). It not only helps ensure functional consistency but significantly reduces the model development time. We take two processors with a same ISA, an ARM7-like one and an ARM9-like one, as our modeling examples to demonstrate the feasibility of the proposed technique. Finally, the experimental results show that, on average our two-layered cycle-accurate model is about 30 times faster than the RTL model in simulation.

    原文English
    主出版物標題2009 International Symposium on VLSI Design, Automation and Test, VLSI-DAT '09
    頁面235-238
    頁數4
    DOIs
    出版狀態Published - 2009
    事件2009 International Symposium on VLSI Design, Automation and Test, VLSI-DAT '09 - Hsinchu, 台灣
    持續時間: 28 4月 200930 4月 2009

    出版系列

    名字2009 International Symposium on VLSI Design, Automation and Test, VLSI-DAT '09

    Conference

    Conference2009 International Symposium on VLSI Design, Automation and Test, VLSI-DAT '09
    國家/地區台灣
    城市Hsinchu
    期間28/04/0930/04/09

    指紋

    深入研究「Efficient two-layered cycle-accurate modeling technique for processor family with same instruction set architecture」主題。共同形成了獨特的指紋。

    引用此