On the Optimal Refresh Power Allocation for Energy-Efficient Memories

Kim, Yongjune; Choi, Won Ho; Guyot, Cyril; Cassuto, Yuval

doi:10.1109/globecom38437.2019.9013465

Cited by 6 publications

(7 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In these applications, the impact of bit errors depends on the bit position, i.e., most significant bits (MSBs) are more important than least significant bits (LSBs) [12], [13]. This differential importance has been leveraged to effectively optimize energy in major memory technologies such as SRAMs [14]- [16] and DRAMs [17], [18]. These works attempt to minimize the mean squared error (MSE) since the MSE is a more meaningful fidelity metric than the write-failure probability (or bit error rate) in error tolerant applications.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Optimizing Write Fidelity of MRAMs via Iterative Water-filling Algorithm

Kim¹,

Jeon²,

Choi³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Magnetic random-access memory (MRAM) is a promising memory technology due to its high density, non-volatility, and high endurance. However, achieving high memory fidelity incurs significant write-energy costs, which should be reduced for large-scale deployment of MRAMs. In this paper, we formulate a biconvex optimization problem to optimize write fidelity given energy and latency constraints. The basic idea is to allocate non-uniform write pulses depending on the importance of each bit position. The fidelity measure we consider is mean squared error (MSE), for which we optimize write pulses via alternating convex search (ACS). By using Karush-Kuhn-Tucker (KKT) conditions, we derive analytic solutions and propose an iterative water-filling-type algorithm by leveraging the analytic solutions. Hence, the proposed iterative water-filling algorithm is computationally more efficient than the original ACS while their solutions are identical. Although the original ACS and the proposed iterative water-filling algorithm do not guarantee global optimality, the MSEs obtained by the proposed algorithm are comparable to the MSEs by complicated global nonlinear programming solvers. Furthermore, we prove that the proposed algorithm can reduce the MSE exponentially with the number of bits per word.For an 8-bit accessed word, the proposed algorithm reduces the MSE by a factor of 21. We also evaluate the proposed algorithm for MNIST dataset classification supposing that the model parameters of deep neural networks are stored in MRAMs. The numerical results show that the optimized write pulses can achieve 40 % write energy reduction for a given classification accuracy.

show abstract

Section: Introductionmentioning

confidence: 99%

“…Prior optimization studies on voltage swings of SRAMs [16] and refresh operations of DRAMs [18] are similar in spirit, viz. minimizing the MSE for given resource constraints.…”

Section: Introductionmentioning

confidence: 99%

Optimizing Write Fidelity of MRAMs via Iterative Water-filling Algorithm

Kim¹,

Jeon²,

Choi³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…This differential importance has been leveraged to effectively optimize energy Y. Kim, Y. Jeon, and C. Guyot are with Western Digital Research, Milpitas, CA 95035 USA (e-mail: {yongjune.kim, yoocharn.jeon, cyril.guyot}@wdc.com). Y. Cassuto is with the Viterbi Department of Electrical Engineering, Technion -Israel Institute of Technology, Haifa, Israel, (e-mail: ycassuto@ee.technion.ac.il) in major memory technologies such as SRAMs [13]- [16] and DRAMs [17], [18].…”

Section: Introductionmentioning

confidence: 99%

“…Prior optimization studies on voltage swing of SRAMs [15], [16] and refresh operations of DRAMs [18] are similar in spirit, viz. minimizing the MSE for given resource constraints.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Optimizing the Write Fidelity of MRAMs

Kim,

Jeon,

Guyot

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

Magnetic random-access memory (MRAM) is a promising memory technology due to its high density, nonvolatility, and high endurance. However, achieving high memory fidelity incurs significant write-energy costs, which should be reduced for large-scale deployment of MRAMs. In this paper, we formulate an optimization problem for maximizing the memory fidelity given energy constraints, and propose a biconvex optimization approach to solve it. The basic idea is to allocate non-uniform write pulses depending on the importance of each bit position. The fidelity measure we consider is minimum mean squared error (MSE), for which we propose an iterative waterfilling algorithm. Although the iterative algorithm does not guarantee global optimality, we can choose a proper starting point that decreases the MSE exponentially and guarantees fast convergence. For an 8-bit accessed word, the proposed algorithm reduces the MSE by a factor of 21.

show abstract

A case study on integrating energy‐efficient technologies for display devices

Rahman,

Tunny,

Bhuiyan

et al. 2024

Electronics Letters

View full text Add to dashboard Cite

For liquid crystal display‐based devices, the memory sub‐system and backlight consume a notable amount of energy, posing a critical concern, especially for portable devices. This work introduces an integrated display system (IDS) aimed at combining technologies for energy‐efficient memory usage and backlight power optimization, which have been addressed independently until now in the existing literature. Two modifications of approximated memory are proposed to find an optimal scheme for energy‐efficient memory usage as part of the proposed IDS schemes. Additionally, three distinct IDS schemes are proposed and compared to identify an optimum solution, ensuring energy efficiency maximization while maintaining image quality for each image. Experimental results demonstrate that 42.19% of energy consumption in memory access and 37.57% of the backlight power consumption can be reduced, while maintaining the image quality in terms of two different metrics.

show abstract

On the Optimal Refresh Power Allocation for Energy-Efficient Memories

Cited by 6 publications

References 31 publications

Optimizing Write Fidelity of MRAMs via Iterative Water-filling Algorithm

Optimizing Write Fidelity of MRAMs via Iterative Water-filling Algorithm

Optimizing the Write Fidelity of MRAMs

A case study on integrating energy‐efficient technologies for display devices

Contact Info

Product

Resources

About