Retention-Aware Refresh Techniques for Reducing Power and Mitigation of Data Retention Faults in DRAM

  • Shyue-Kung LuEmail author
  • Hung-Kai Huang
  • Chun-Lung Hsu
  • Chi-Tien Sun
  • Kohei Miyase


Due the leakage mechanisms exist in DRAM cells, DRAM cells lose stored information over time. Periodic refresh operations are inevitable for retaining the stored information. However, refresh operations are very power hungry and impact the bandwidth substantially since the refresh period is usually determined by the leakiest DRAM cells. A straightforward way for reducing refresh power is by merely extending the single standard refresh period. The incurred side effect is that more data retention faults (DRFs) will be generated and the fabrication yield will be sacrificed. To cure these dilemmas, a novel address remapping technique—the sub-bank address remapping (SBAR) technique is proposed in this paper. SBAR manipulates the logical-to-physical address remapping for each sub-bank such that the leakiest cells can be clustered and refreshed with their most suitable refresh periods. For the majority of DRAM cells, they can be refreshed with a longer refresh period such that the refresh power can be effectively reduced. The corresponding hardware architectures are also proposed. Experimental results show that we can respectively save 74.97% refresh power with less than 0.1% hardware overhead for a 1-Gb DRAM. Moreover, the fabrication yield can also be improved significantly.


DRAM Data retention fault Refresh period Low power Yield 


Supplementary material

10836_2019_5817_MOESM1_ESM.pdf (297 kb)
ESM 1 (PDF 296 kb)


  1. 1.
    Agrawal A, Ansari A, Torrellas J (Feb. 2014) Mosaic: exploiting the spatial locality of process variation to reduce refresh energy in on-chip eDRAM modules. In: Proc. of the 20th Int. Symp. On high-performance computer architecture (HPCA’14), pp 84–95Google Scholar
  2. 2.
    Ahn JH, Jeong BH, Kim SH, Chu SH, Cho SW, Lee HJ, Kim MH, Park SI, Shin SW, Lee JH, Han BS, Hong JK, Moran PB, Kim YT (2006) Adaptive self refresh scheme for battery operated high-density mobile DRAM applications. IEEE Asian Solid-State Circuits Conf, Nov:319–322Google Scholar
  3. 3.
    Baek S, Cho S, Melhem R (Dec. 2014) Refresh now and then. IEEE Trans on Computers 63(12):3114–3126MathSciNetCrossRefzbMATHGoogle Scholar
  4. 4.
    Bose RC, Ray-Chaudhuri DK (Mar. 1960) On a class of error-correcting binary group codes. Inf Control 3(1):68–79MathSciNetCrossRefzbMATHGoogle Scholar
  5. 5.
    Ghosh M, Lee HHS (2007) Smart refresh: an enhanced memory controller design for reducing energy in conventional and 3D die-stacked DRAMs. Proc IEEE/ACM International Symp on Microarchitecture, Dec:134–145Google Scholar
  6. 6.
    Hamamoto T, Sugiura S, Sawada S (June 1998) On the retention time distribution of dynamic random access memory (DRAM). IEEE Trans on Electron Devices 45(6):1300–1309CrossRefGoogle Scholar
  7. 7.
    Hamburgen WR, Wallach DA, Viredaz MA, Brakmo LS, Waldspurger CA, Barlett JF, Mann T, Farkas KI (April 2001) Itsy: stretching the bound of mobile computing. IEEE Computer 34(4):28–36CrossRefGoogle Scholar
  8. 8.
    Y. Han, Y. Wang, H. W. Li, and X. W. Li, "Data-aware DRAM refresh control to squeeze the margin of retention time in 3D memory," in Proc. IEEE/ACM Int’l Conf on Computer-aided Design, pp. 295–300, Nov. 2014Google Scholar
  9. 9.
    Huang C-T, Huang J-R, Wu C-E, Wu C-W, Chang T-Y (Jan.–Mar. 1999) A programmable BIST core for embedded DRAM. IEEE Design & Test of Computers 16(1):59–70CrossRefGoogle Scholar
  10. 10.
    Idei Y, Shimohigashi K, Aoki M, Noda H, Iwai H, Sato K, Tachibana T (Feb. 1998) Dual-period self-refresh scheme for low-power DRAM’s with on-chip PROM mode register. IEEE Jour of Solid-State Circuits 33(2):253–259CrossRefGoogle Scholar
  11. 11.
    Kim J, Papaefthymiou MC (Dec. 2003) Block-based multi-period dynamic memory design for low data-retention power. IEEE Trans on VLSI Systems 11(6):1006–1018CrossRefGoogle Scholar
  12. 12.
    Liu S, Pattabiraman K, Moscibroda T, Zorn BG (2011) Flikker: saving DRAM refresh-power through critical data partitioning. Proc Int’l Conf on Architectural Support for Programming Languages and Operating Systems, March:213–224Google Scholar
  13. 13.
    Liu J, Jaiyen B, Veras R, Mutlu O (2012) RAIDR: retention-aware intelligent DRAM refresh. Proc. of Int’l Symp. on Computer Architecture, June:1–12Google Scholar
  14. 14.
  15. 15.
  16. 16.
  17. 17.
    Mittal S, Vetter JS (June 2016) A survey of techniques for architecting DRAM caches. IEEE Trans Parallel and Distributed Systems 27(6):1852–1863CrossRefGoogle Scholar
  18. 18.
    J. Mukundan, H. Hunter, K.-h. Kim, J. Stuecheli, and J. F. Mart’inez, “understanding and mitigating refresh overheads in high-density DDR4 DRAM systems,” in Proc. IEEE Int’l Symp. On computer architecture (ISCA), 2013, pp. 48–59Google Scholar
  19. 19.
    Nagai T, Wada M, Iwai H, Kaku M, Suzuki A, Takai T, Itoga N, Miyazaki T, Takenaka H, Hojo T, Miyano S (2006) A 65nm low-power embedded DRAM with extended data-retention sleep mode. IEEE Int’l Solid-State Circuits Conf, Feb:567–576Google Scholar
  20. 20.
    Nakagome Y, Aoki M, Ikenaga S, Horiguchi M, Kimura S, Kawamoto Y, Itoh K (Oct. 1988) The impact of data-line interference noise on DRAM scaling. IEEE J Solid State Circuits 23(5):1120–1127CrossRefGoogle Scholar
  21. 21.
    Ohsawa T, Kai K, Murakami K (1998) Optimizing the DRAM refresh count for merged DRAM/logic LSIs. Proc Int’l Symp on Low-Power Electronics and Design, Aug:82–87Google Scholar
  22. 22.
    Pyo SS, Lee CH, Kim GH, Choi KM, Jun YH, Kong BS (2009) 45nm low-power embedded pseudo-SRAM with ECC-based auto-adjusted self-refresh scheme. IEEE Int’l Symp on Circuits and Systems (ISCAS), May:2517–2520Google Scholar
  23. 23.
    Reviriego P, Sanchez-Macian A, Maestro JA (2012) Low power embedded DRAM caches using BCH code partitioning. IEEE Int’l On-Line Testing Symposium (IOLTS), June:79–83Google Scholar
  24. 24.
    M. A. Viredaz and D. A. Wallach, “Power evaluation of a handheld computer,” IEEE Micro, vol. 23, no. 1, pp. 66–74, Jan/Feb 2003Google Scholar
  25. 25.
    Wang Y, Han Y, Wang C, Li HW, Li XW (2015) RADAR: a case for retention-aware DRAM assembly and repair in future FGR DRAM memory. Proc IEEE/ACM Design, Automation Conference, June:1–6Google Scholar
  26. 26.
    Wang Y, Han Y, Wang C, Li HW, Li XW (May 2017) Retention-aware DRAM assembly and repair for future FGR memories. IEEE Trans Computer-Aided Design of Integrated Circuits and Systems 36(3):705–718Google Scholar
  27. 27.
    Wilkerson C, Alameldeen AR, Chishti Z, Wu W, Somasekhar D, Lu SL (2010) Reducing cache power with low-cost, multi-bit error-correcting codes. Proc of Int’l Symp on Computer Architecture, June 38:83–93Google Scholar
  28. 28.
    Yu YC, Hou CS, Chang LJ, Li JF, Lo CY, Kwai D, Wu CW (Apr. 2013) A hybrid ECC and redundancy technique for reducing refresh power of DRAMS. in Proc VLSI Test Symp (VTS):1–6Google Scholar

Copyright information

© Springer Science+Business Media, LLC, part of Springer Nature 2019

Authors and Affiliations

  1. 1.Department of Electrical EngineeringNational Taiwan University of Science and TechnologyTaipeiTaiwan
  2. 2.Information and Communications Research LaboratoriesIndustrial Technology Research InstituteZhudongTaiwan
  3. 3.Department of Creative InformaticsKyushu Institute of TechnologyFukuokaJapan

Personalised recommendations