• DocumentCode
    1436678
  • Title

    An Enhanced Canary-Based System With BIST for SRAM Standby Power Reduction

  • Author

    Wang, Jiajing ; Hoefler, Alexander ; Calhoun, Benton H.

  • Author_Institution
    Dept. of Electr. & Comput. Eng., Univ. of Virginia, Charlottesville, VA, USA
  • Volume
    19
  • Issue
    5
  • fYear
    2011
  • fDate
    5/1/2011 12:00:00 AM
  • Firstpage
    909
  • Lastpage
    914
  • Abstract
    To achieve aggressive standby power reduction for static random access memory (SRAM), we have previously proposed a closed-loop VDD scaling system with canary replicas that can track global variations. In this paper, we propose several techniques to enhance the efficiency of this system for more advanced technologies. Adding dummy cells around the canary cell improves the tracking of systematic variations. A new canary circuit avoids the possibility that a canary cell may never fail because it resets into its more stable data pattern. A built-in self-test (BIST) block incorporates self-calibration of SRAM minimum standby VDD and the initial failure threshold due to intrinsic mismatch. Measurements from a new 45 nm test chip further demonstrate the function of the canary cells in smaller technology and show that adding dummy cells reduces the variation of the canary cell.
  • Keywords
    SRAM chips; built-in self test; calibration; integrated circuit testing; BIST; SRAM standby power reduction; built-in self-test; canary circuit; canary-based system; closed-loop VDD scaling; dummy cells; intrinsic mismatch; self-calibration; size 45 nm; static random access memory; Built-in self-test; CMOS technology; Circuits; Leakage current; Random access memory; SRAM chips; Semiconductor device measurement; Stability; Temperature; Voltage; Built-in self test (BIST); data retention voltage (DRV); standby power; static random access memory (SRAM); variation;
  • fLanguage
    English
  • Journal_Title
    Very Large Scale Integration (VLSI) Systems, IEEE Transactions on
  • Publisher
    ieee
  • ISSN
    1063-8210
  • Type

    jour

  • DOI
    10.1109/TVLSI.2010.2042184
  • Filename
    5428790