Circuit Design for Non-volatile Magnetic
Memory
Tony Tae-Hyoung Kim
Abstract High performance low power memory is a topmost requirement in
advanced computing systems. Magnetic memory has been considered as a promising
solution because of its performance and non-volatility. However, it has various design
challenges such as small tunneling magnetoresistance (TMR) ratios and large variability that need to be tackled for reliable operation. This chapter will discuss those
challenges and introduce state-of-the-art write and read techniques.
1 Introduction
The performance of most computing systems is generally limited by the characteristics of memory. Over the last few decades, charge-storage-based memories such
as static random access memory (SRAM) and dynamic random access memory
(DRAM) have been the mainstream memory solutions. However, these memory
technologies are facing challenges in scaling, which also limits the implementable
memory density and the system performance. In addition, many emerging systems
with frequency idle states require to store data with extremely low power consumption. No conventional memory technology can satisfy the above requirements. To
tackle these issues, various non-volatile memory devices such as magnetic memory
(MRAM), phase-change memory (PCRAM), and resistive memory (RRAM) have
been explored. Since these memory devices have two terminals, they demonstrate
high scalability (<10 nm). In addition, the non-volatile characteristics allows these
memory technologies applicable to many ultra-low power systems requiring almostzero power during idle or standby modes. However, the above non-volatile memory
candidates have various design issues to be overcome, which is the main focus of
this chapter.
This chapter introduces the overview of magnetic memory cells in circuit
designer’s point of view. After that, basic memory architecture for magnetic memory
T. T.-H. Kim (B)
School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore,
Singapore
e-mail: THKIM@ntu.edu.sg
© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021
W. S. Lew et al. (eds.), Emerging Non-volatile Memory Technologies,
https://doi.org/10.1007/978-981-15-6912-8_6
203
Memory
Tony Tae-Hyoung Kim
Abstract High performance low power memory is a topmost requirement in
advanced computing systems. Magnetic memory has been considered as a promising
solution because of its performance and non-volatility. However, it has various design
challenges such as small tunneling magnetoresistance (TMR) ratios and large variability that need to be tackled for reliable operation. This chapter will discuss those
challenges and introduce state-of-the-art write and read techniques.
1 Introduction
The performance of most computing systems is generally limited by the characteristics of memory. Over the last few decades, charge-storage-based memories such
as static random access memory (SRAM) and dynamic random access memory
(DRAM) have been the mainstream memory solutions. However, these memory
technologies are facing challenges in scaling, which also limits the implementable
memory density and the system performance. In addition, many emerging systems
with frequency idle states require to store data with extremely low power consumption. No conventional memory technology can satisfy the above requirements. To
tackle these issues, various non-volatile memory devices such as magnetic memory
(MRAM), phase-change memory (PCRAM), and resistive memory (RRAM) have
been explored. Since these memory devices have two terminals, they demonstrate
high scalability (<10 nm). In addition, the non-volatile characteristics allows these
memory technologies applicable to many ultra-low power systems requiring almostzero power during idle or standby modes. However, the above non-volatile memory
candidates have various design issues to be overcome, which is the main focus of
this chapter.
This chapter introduces the overview of magnetic memory cells in circuit
designer’s point of view. After that, basic memory architecture for magnetic memory
T. T.-H. Kim (B)
School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore,
Singapore
e-mail: THKIM@ntu.edu.sg
© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021
W. S. Lew et al. (eds.), Emerging Non-volatile Memory Technologies,
https://doi.org/10.1007/978-981-15-6912-8_6
203