SySCIM: SystemC-AMS Simulation of Memristive Computation In-Memory

Seyed Hossein Hashemi Shadmehri; Ali Banagozar; Mehdi Kamal; Sander Stuijk; Ali Afzali-Kusha; Massoud Pedram; Henk Corporaal

doi:10.23919/DATE54114.2022.9774749

SySCIM: SystemC-AMS Simulation of Memristive Computation In-Memory

Seyed Hossein Hashemi Shadmehri, Ali Banagozar, Mehdi Kamal, Sander Stuijk, Ali Afzali-Kusha, Massoud Pedram, Henk Corporaal

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

3 Citations (Scopus)

Abstract

Computation-in-memory (CIM) is one of the most appealing computing paradigms, especially for implementing artificial neural networks. Non-volatile memories like ReRAMs, PCMs, etc., have proven to be promising candidates for the realization of CIM processors. However, these devices and their driving circuits are subject to non-idealities. This paper presents a comprehensive platform, named SysCIM, for simulating memristor-based CIM systems. SySCIM considers the impact of the non-idealities of the CIM components, including memristor device, memristor crossbar (interconnects), analog-to-digital converter, and transimpedance amplifier, on the vector-matrix multiplication performed by the CIM unit. The CIM modules are described in SystemC and SystemC-AMS to reach a higher simulation speed while maintaining high simulation accuracy. Experiments under different crossbar sizes show SySCIM performs simulations up to 117 x faster than HSPICE with less than 4% accuracy loss. The modular design of SySCIM provides researchers with an easy design-space exploration tool to investigate the effects of various non-idealities.

Original language	English
Title of host publication	Proceedings of the 2022 Design, Automation & Test in Europe Conference & Exhibition (DATE 2022)
Editors	Cristiana Bolchini, Ingrid Verbauwhede, Ioana Vatajelu
Publisher	Institute of Electrical and Electronics Engineers
Pages	1467-1472
Number of pages	6
ISBN (Electronic)	978-3-9819263-6-1
DOIs	https://doi.org/10.23919/DATE54114.2022.9774749
Publication status	Published - 19 May 2022
Event	2022 Design, Automation and Test in Europe Conference and Exhibition, DATE 2022 - Virtual, Online, Belgium Duration: 14 Mar 2022 → 23 Mar 2022

Conference

Conference	2022 Design, Automation and Test in Europe Conference and Exhibition, DATE 2022
Country/Territory	Belgium
Period	14/03/22 → 23/03/22

Keywords

computation-in-memory
memristor
reliability
simulation

Access to Document

10.23919/DATE54114.2022.9774749

Cite this

Shadmehri, S. H. H., Banagozar, A., Kamal, M., Stuijk, S., Afzali-Kusha, A., Pedram, M., & Corporaal, H. (2022). SySCIM: SystemC-AMS Simulation of Memristive Computation In-Memory. In C. Bolchini, I. Verbauwhede, & I. Vatajelu (Eds.), Proceedings of the 2022 Design, Automation & Test in Europe Conference & Exhibition (DATE 2022) (pp. 1467-1472). Article 9774749 Institute of Electrical and Electronics Engineers. https://doi.org/10.23919/DATE54114.2022.9774749

Shadmehri, Seyed Hossein Hashemi ; Banagozar, Ali ; Kamal, Mehdi et al. / SySCIM : SystemC-AMS Simulation of Memristive Computation In-Memory. Proceedings of the 2022 Design, Automation & Test in Europe Conference & Exhibition (DATE 2022). editor / Cristiana Bolchini ; Ingrid Verbauwhede ; Ioana Vatajelu. Institute of Electrical and Electronics Engineers, 2022. pp. 1467-1472

@inproceedings{f07e214824d6489cace1b60b9003ff33,

title = "SySCIM: SystemC-AMS Simulation of Memristive Computation In-Memory",

abstract = "Computation-in-memory (CIM) is one of the most appealing computing paradigms, especially for implementing artificial neural networks. Non-volatile memories like ReRAMs, PCMs, etc., have proven to be promising candidates for the realization of CIM processors. However, these devices and their driving circuits are subject to non-idealities. This paper presents a comprehensive platform, named SysCIM, for simulating memristor-based CIM systems. SySCIM considers the impact of the non-idealities of the CIM components, including memristor device, memristor crossbar (interconnects), analog-to-digital converter, and transimpedance amplifier, on the vector-matrix multiplication performed by the CIM unit. The CIM modules are described in SystemC and SystemC-AMS to reach a higher simulation speed while maintaining high simulation accuracy. Experiments under different crossbar sizes show SySCIM performs simulations up to 117 x faster than HSPICE with less than 4% accuracy loss. The modular design of SySCIM provides researchers with an easy design-space exploration tool to investigate the effects of various non-idealities.",

keywords = "computation-in-memory, memristor, reliability, simulation",

author = "Shadmehri, {Seyed Hossein Hashemi} and Ali Banagozar and Mehdi Kamal and Sander Stuijk and Ali Afzali-Kusha and Massoud Pedram and Henk Corporaal",

year = "2022",

month = may,

day = "19",

doi = "10.23919/DATE54114.2022.9774749",

language = "English",

pages = "1467--1472",

editor = "Cristiana Bolchini and Ingrid Verbauwhede and Ioana Vatajelu",

booktitle = "Proceedings of the 2022 Design, Automation & Test in Europe Conference & Exhibition (DATE 2022)",

publisher = "Institute of Electrical and Electronics Engineers",

address = "United States",

note = "2022 Design, Automation and Test in Europe Conference and Exhibition, DATE 2022 ; Conference date: 14-03-2022 Through 23-03-2022",

}

Shadmehri, SHH, Banagozar, A, Kamal, M, Stuijk, S, Afzali-Kusha, A, Pedram, M & Corporaal, H 2022, SySCIM: SystemC-AMS Simulation of Memristive Computation In-Memory. in C Bolchini, I Verbauwhede & I Vatajelu (eds), Proceedings of the 2022 Design, Automation & Test in Europe Conference & Exhibition (DATE 2022)., 9774749, Institute of Electrical and Electronics Engineers, pp. 1467-1472, 2022 Design, Automation and Test in Europe Conference and Exhibition, DATE 2022, Belgium, 14/03/22. https://doi.org/10.23919/DATE54114.2022.9774749

SySCIM: SystemC-AMS Simulation of Memristive Computation In-Memory. / Shadmehri, Seyed Hossein Hashemi; Banagozar, Ali; Kamal, Mehdi et al.
Proceedings of the 2022 Design, Automation & Test in Europe Conference & Exhibition (DATE 2022). ed. / Cristiana Bolchini; Ingrid Verbauwhede; Ioana Vatajelu. Institute of Electrical and Electronics Engineers, 2022. p. 1467-1472 9774749.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

TY - GEN

T1 - SySCIM

T2 - 2022 Design, Automation and Test in Europe Conference and Exhibition, DATE 2022

AU - Shadmehri, Seyed Hossein Hashemi

AU - Banagozar, Ali

AU - Kamal, Mehdi

AU - Stuijk, Sander

AU - Afzali-Kusha, Ali

AU - Pedram, Massoud

AU - Corporaal, Henk

PY - 2022/5/19

Y1 - 2022/5/19

N2 - Computation-in-memory (CIM) is one of the most appealing computing paradigms, especially for implementing artificial neural networks. Non-volatile memories like ReRAMs, PCMs, etc., have proven to be promising candidates for the realization of CIM processors. However, these devices and their driving circuits are subject to non-idealities. This paper presents a comprehensive platform, named SysCIM, for simulating memristor-based CIM systems. SySCIM considers the impact of the non-idealities of the CIM components, including memristor device, memristor crossbar (interconnects), analog-to-digital converter, and transimpedance amplifier, on the vector-matrix multiplication performed by the CIM unit. The CIM modules are described in SystemC and SystemC-AMS to reach a higher simulation speed while maintaining high simulation accuracy. Experiments under different crossbar sizes show SySCIM performs simulations up to 117 x faster than HSPICE with less than 4% accuracy loss. The modular design of SySCIM provides researchers with an easy design-space exploration tool to investigate the effects of various non-idealities.

AB - Computation-in-memory (CIM) is one of the most appealing computing paradigms, especially for implementing artificial neural networks. Non-volatile memories like ReRAMs, PCMs, etc., have proven to be promising candidates for the realization of CIM processors. However, these devices and their driving circuits are subject to non-idealities. This paper presents a comprehensive platform, named SysCIM, for simulating memristor-based CIM systems. SySCIM considers the impact of the non-idealities of the CIM components, including memristor device, memristor crossbar (interconnects), analog-to-digital converter, and transimpedance amplifier, on the vector-matrix multiplication performed by the CIM unit. The CIM modules are described in SystemC and SystemC-AMS to reach a higher simulation speed while maintaining high simulation accuracy. Experiments under different crossbar sizes show SySCIM performs simulations up to 117 x faster than HSPICE with less than 4% accuracy loss. The modular design of SySCIM provides researchers with an easy design-space exploration tool to investigate the effects of various non-idealities.

KW - computation-in-memory

KW - memristor

KW - reliability

KW - simulation

UR - http://www.scopus.com/inward/record.url?scp=85130796285&partnerID=8YFLogxK

U2 - 10.23919/DATE54114.2022.9774749

DO - 10.23919/DATE54114.2022.9774749

M3 - Conference contribution

AN - SCOPUS:85130796285

SP - 1467

EP - 1472

BT - Proceedings of the 2022 Design, Automation & Test in Europe Conference & Exhibition (DATE 2022)

A2 - Bolchini, Cristiana

A2 - Verbauwhede, Ingrid

A2 - Vatajelu, Ioana

PB - Institute of Electrical and Electronics Engineers

Y2 - 14 March 2022 through 23 March 2022

ER -

Shadmehri SHH, Banagozar A, Kamal M, Stuijk S, Afzali-Kusha A, Pedram M et al. SySCIM: SystemC-AMS Simulation of Memristive Computation In-Memory. In Bolchini C, Verbauwhede I, Vatajelu I, editors, Proceedings of the 2022 Design, Automation & Test in Europe Conference & Exhibition (DATE 2022). Institute of Electrical and Electronics Engineers. 2022. p. 1467-1472. 9774749 doi: 10.23919/DATE54114.2022.9774749

SySCIM: SystemC-AMS Simulation of Memristive Computation In-Memory

Abstract

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this