Enhanced Throughput and Energy-Efficient MRAM-Based NB-LDPC Decoder

In this paper, an adaptive and efficient memorization system based on Spin-Transfer-Torque Magneto-resistive Random Access Memory (STT-MRAM) is proposed to be implemented within Non Binary-Low Density Parity Check (NB-LDPC) Decoder. Based on the available MRAM types and array dimensions, new data mapping and efficient structure of the memorization system are proposed. The power consumption of the NB-LDPC decoder with STT-MRAM is evaluated and compared to the Static Random Access Memory (SRAM)-based one. Simulation results show that the proposed memorization system offers an energy reduction that reaches 73.35% for MRAM type I and 65.15% for MRAM type II. Moreover MRAM type I increases the throughput by 21.05% and MRAM type II by 43.75% for the NB-LDPC Decoder.

[1]  David Declercq,et al.  Performance Evaluation of Non-Binary LDPC Codes on Wireless Channels , 2009 .

[2]  Emmanuel Boutillon,et al.  Design of a GF(64)-LDPC Decoder Based on the EMS Algorithm , 2013, IEEE Transactions on Circuits and Systems I: Regular Papers.

[3]  H. Ohno,et al.  A 1 Mb Nonvolatile Embedded Memory Using 4T2MTJ Cell With 32 b Fine-Grained Power Gating Scheme , 2013, IEEE Journal of Solid-State Circuits.

[4]  Vincent Gripon,et al.  NoC-MRAM architecture for memory-based computing: Database-search case study , 2017, 2017 15th IEEE International New Circuits and Systems Conference (NEWCAS).

[5]  Ronald F. DeMara,et al.  Survey of STT-MRAM Cell Design Strategies , 2017, ACM J. Emerg. Technol. Comput. Syst..

[6]  D. Mackay,et al.  Low density parity check codes over GF(q) , 1998, 1998 Information Theory Workshop (Cat. No.98EX131).

[7]  Zouhair El-Bazzal,et al.  MRAM-based memorization system for NB-LDPC decoder , 2017, 2017 29th International Conference on Microelectronics (ICM).