SOT-MRAM-based true in-memory computing architecture for approximate multiplication

The in-memory computing (IMC) paradigm emerges as an effective solution to break the bottlenecks of conventional von Neumann architecture. In the current work, an approximate multiplier in spin-orbit torque magnetoresistive random access memory (SOT-MRAM) based true IMC (STIMC) architecture was pres...

Full description

Saved in:
Bibliographic Details
Main Authors: Min Song, Qilong Tang, Xintong Ouyang, Wei Duan, Yan Xu, Shuai Zhang, Long You
Format: Article
Language:English
Published: Elsevier 2025-06-01
Series:Chip
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2709472325000085
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850151738162544640
author Min Song
Qilong Tang
Xintong Ouyang
Wei Duan
Yan Xu
Shuai Zhang
Long You
author_facet Min Song
Qilong Tang
Xintong Ouyang
Wei Duan
Yan Xu
Shuai Zhang
Long You
author_sort Min Song
collection DOAJ
description The in-memory computing (IMC) paradigm emerges as an effective solution to break the bottlenecks of conventional von Neumann architecture. In the current work, an approximate multiplier in spin-orbit torque magnetoresistive random access memory (SOT-MRAM) based true IMC (STIMC) architecture was presented, where computations were performed natively within the cell array instead of in peripheral circuits. Firstly, basic Boolean logic operations were realized by utilizing the feature of unipolar SOT device. Two majority gate-based imprecise compressors and an ultra-efficient approximate multiplier were then built to reduce the energy and latency. An optimized data mapping strategy facilitating bit-serial operations with an extensive degree of parallelism was also adopted. Finally, the performance enhancements by performing our approximate multiplier in image smoothing were demonstrated. Detailed simulation results show that the proposed 8 × 8 approximate multiplier could reduce the energy and latency at least by 74.2% and 44.4% compared with the existing designs. Moreover, the scheme could achieve improved peak signal-to-noise ratio (PSNR) and structural similarity index metric (SSIM), ensuring high-quality image processing outcomes.
format Article
id doaj-art-8d0366d070f74a31b95d8729e259d312
institution OA Journals
issn 2709-4723
language English
publishDate 2025-06-01
publisher Elsevier
record_format Article
series Chip
spelling doaj-art-8d0366d070f74a31b95d8729e259d3122025-08-20T02:26:09ZengElsevierChip2709-47232025-06-014210013410.1016/j.chip.2025.100134SOT-MRAM-based true in-memory computing architecture for approximate multiplicationMin Song0Qilong Tang1Xintong Ouyang2Wei Duan3Yan Xu4Shuai Zhang5Long You6Key Laboratory of Intelligent Sensing System and Security of the Ministry of Education, Hubei Key Laboratory of Micro-Nanoelectronic Materials and Devices, School of Microelectronics, Hubei University, Wuhan 430062, ChinaKey Laboratory of Intelligent Sensing System and Security of the Ministry of Education, Hubei Key Laboratory of Micro-Nanoelectronic Materials and Devices, School of Microelectronics, Hubei University, Wuhan 430062, ChinaKey Laboratory of Intelligent Sensing System and Security of the Ministry of Education, Hubei Key Laboratory of Micro-Nanoelectronic Materials and Devices, School of Microelectronics, Hubei University, Wuhan 430062, ChinaSchool of Integrated Circuits, Huazhong University of Science and Technology, Wuhan 430074, ChinaSchool of Integrated Circuits, Huazhong University of Science and Technology, Wuhan 430074, ChinaSchool of Integrated Circuits, Huazhong University of Science and Technology, Wuhan 430074, ChinaSchool of Integrated Circuits, Huazhong University of Science and Technology, Wuhan 430074, China; Corresponding author.The in-memory computing (IMC) paradigm emerges as an effective solution to break the bottlenecks of conventional von Neumann architecture. In the current work, an approximate multiplier in spin-orbit torque magnetoresistive random access memory (SOT-MRAM) based true IMC (STIMC) architecture was presented, where computations were performed natively within the cell array instead of in peripheral circuits. Firstly, basic Boolean logic operations were realized by utilizing the feature of unipolar SOT device. Two majority gate-based imprecise compressors and an ultra-efficient approximate multiplier were then built to reduce the energy and latency. An optimized data mapping strategy facilitating bit-serial operations with an extensive degree of parallelism was also adopted. Finally, the performance enhancements by performing our approximate multiplier in image smoothing were demonstrated. Detailed simulation results show that the proposed 8 × 8 approximate multiplier could reduce the energy and latency at least by 74.2% and 44.4% compared with the existing designs. Moreover, the scheme could achieve improved peak signal-to-noise ratio (PSNR) and structural similarity index metric (SSIM), ensuring high-quality image processing outcomes.http://www.sciencedirect.com/science/article/pii/S2709472325000085Spin-orbit torque (SOT)Magnetoresistive random access memory (MRAM)In-memory computing (IMC)Approximate multiplierData mapping strategy
spellingShingle Min Song
Qilong Tang
Xintong Ouyang
Wei Duan
Yan Xu
Shuai Zhang
Long You
SOT-MRAM-based true in-memory computing architecture for approximate multiplication
Chip
Spin-orbit torque (SOT)
Magnetoresistive random access memory (MRAM)
In-memory computing (IMC)
Approximate multiplier
Data mapping strategy
title SOT-MRAM-based true in-memory computing architecture for approximate multiplication
title_full SOT-MRAM-based true in-memory computing architecture for approximate multiplication
title_fullStr SOT-MRAM-based true in-memory computing architecture for approximate multiplication
title_full_unstemmed SOT-MRAM-based true in-memory computing architecture for approximate multiplication
title_short SOT-MRAM-based true in-memory computing architecture for approximate multiplication
title_sort sot mram based true in memory computing architecture for approximate multiplication
topic Spin-orbit torque (SOT)
Magnetoresistive random access memory (MRAM)
In-memory computing (IMC)
Approximate multiplier
Data mapping strategy
url http://www.sciencedirect.com/science/article/pii/S2709472325000085
work_keys_str_mv AT minsong sotmrambasedtrueinmemorycomputingarchitectureforapproximatemultiplication
AT qilongtang sotmrambasedtrueinmemorycomputingarchitectureforapproximatemultiplication
AT xintongouyang sotmrambasedtrueinmemorycomputingarchitectureforapproximatemultiplication
AT weiduan sotmrambasedtrueinmemorycomputingarchitectureforapproximatemultiplication
AT yanxu sotmrambasedtrueinmemorycomputingarchitectureforapproximatemultiplication
AT shuaizhang sotmrambasedtrueinmemorycomputingarchitectureforapproximatemultiplication
AT longyou sotmrambasedtrueinmemorycomputingarchitectureforapproximatemultiplication