期刊文献+

一种基于In-Situ AVS技术的低功耗处理器实现方法

A Low Power Implememtation Method on Processor Based on In-Situ AVS Technique
下载PDF
导出
摘要 为了进一步降低ARM Cortex M0处理器的功耗,提出了一种基于In-Situ AVS(Adaptive Voltage Scaling)技术的低功耗电路实现方法.该方法通过电路的路径延时估算出监测窗口大小(ΔT)和电压调整的错误(pre-error)数量阈值(nlimit),将部分关键路径的触发器替换成实时延时检测电路,对重要的几条路径的延时和错误进行实时监测,经AVS控制单元和电压调整模块随着PVTA的变化自适应地调整电压,有效地降低电路功耗.在SMIC 180nm工艺下设计了一款ARM Cortex M0处理器,将此方法应用于处理器的一个关键模块,即AHB到APB的桥接电路(AHB_to_APB).测试结果表明,在一个观测区间(N=1 000)内错误率为8.9E-4时,电路功耗降低了28%. In order to excessively reduce power dissipation of the ARM Cortex M0 processor, we propose a Low power implememtation circiut based on In-Situ AVS Technique. The main process is to estimate the monitoring window(△T) and pre-error limit number(nlimit ) which detelmings the voltage adjusting through the path delay of the circuit. Besides, the In-Situ Delay Monitor replaces the flip-flops at the end of critical paths. After detecting the path delay and pre-error using the In-Situ Delay Monitor, the circuit voltage could be adaptively adjusted through the AVS control unit and voltage regulator with the varying of PVTA, which is an efficient scheme to tune the supply voltage of digital circuits according to variations. We have designed a Arm Cortex M0 processor in the SMIC 180 ns process, and applied this method to a critical module(AHB to APB bridge). As a result, We simulated the power dissipation, and find that the circuit applied AVS technique has reduced 28~ with the pre-error rate of 8. 9E-4 during a observation interval compared to the design former.
出处 《微电子学与计算机》 CSCD 北大核心 2015年第10期12-16,共5页 Microelectronics & Computer
基金 中科院国际合作重点项目(GJHZ201304)
关键词 低功耗 IN-SITU AVS 监测窗口 实时延时检测电路 low power dissipatiom In-Situ AVS monitoring window In-Situ Delay Monitor
  • 相关文献

参考文献10

  • 1Martin Wirnshofer, Leonhard Heil3,et al. A variation- aware adaptive voltage scaling technique based on in-si- tu delay monitoring[C]//Design and Diagnostics of E- lectronic Circuits b- Systems (DDECS). Cottbus, 2011 : 261-266.
  • 2Minghao Jin. On-chip PVT compensation technology for adaptive voltage sealing control in computer power management application[C]// IEEE Automation and Computing (ICAC)Laughborough Loughborough, 2012 : 1-5.
  • 3Das S. RazorII: In situ error detection and correction for PVT and SER tolerance[J]. IEEE Journal of Solid- State Circuits, 2009,44(1) : 32-48.
  • 4David Bull, Shidhartha Das. A power-efficient 32 bit ARM processor using timing-error detection and cor- rection for transient-error tolerance and adaptation to PVT variation[J]. IEEE Journal of solid-state cir- cuits, 2011,46(1): 18-31.
  • 5Bowman K. Energy-efficient and metastability-immune timingerror detection and instruction replay-based re- covery circuits for dynamic variation tolerance[C]// IEEE J. Solid-State Circuits. San Francisco, 2009: 49- 63.
  • 6Vincent L. Dynamic variability monitoring using statis- tical tests for energy efficient adaptive architectures [J]. IEEE Circuits and Systems Society, 2014,61 (6) :1741-1754.
  • 7Tuck-Boon Chan. On aging-aware signoff for circuits with adaptive voltage scalingrJT. IEEE Circuits and Systems Society, 2014,61(10) : 2920-2930.
  • 8Peng Shenyu. Instruction-cycle-based dynamic voltage scaling power management for low-power digital signal processor with 53% power savings[J]. IEEE Solid- State Circuits Society, 2013, 48(11) : 2649-2661.
  • 9Sridhara S R, DiRenzo 1VL Microwatt embedded pro- cessor platform for medical system-on-chip applications [J]. IEEE J. Solid-State Circuits, 2011, 46(4):721 -730.
  • 10Lee Y H, Chiu C C. A near-optimum dynamic voltage scaling (DVS) in 65 nm energy-efficient power man- agement with frequency-based control (FBC) for SoC system[J]. IEEE J. Solid-State Circuits, 2012,47 (1) ; 2563-2575.

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部