A Δ-Based Spike Sorting SoC with End-to-End Implementation of Event-Driven Binary Autoencoder Neural Network in Analog CIM Achieving 94.54% Accuracy and 3.11μW/ch

Edward Choi; Vincent Lukito; Injun Choi; Seoyoung Lee; Ik Joon Chang; Sohmyung Ha; Minkyu Je

doi:10.1109/VLSITechnologyandCir46783.2024.10631554

A Δ-Based Spike Sorting SoC with End-to-End Implementation of Event-Driven Binary Autoencoder Neural Network in Analog CIM Achieving 94.54% Accuracy and 3.11μW/ch

Edward Choi
, Vincent Lukito
, Injun Choi
, Seoyoung Lee
, Ik Joon Chang
, Sohmyung Ha
, Minkyu Je

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

2 Citations (Scopus)

Abstract

The proposed Δ -based spike sorting (SS) SoC is the first on-chip implementation of an analog computing-in-memory (CIM) binary autoencoder neural network (B-AENN) feature extraction with enhanced spike detection adopting Δ -spikes, resulting in the highest on-chip SS classification accuracy of 94.54%. It also allows to reduce the digital data transmission rate by 48.8× compared to prior SS systems.

Original language	English
Title of host publication	2024 IEEE Symposium on VLSI Technology and Circuits, VLSI Technology and Circuits 2024
Publisher	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)	9798350361469
DOIs	https://doi.org/10.1109/VLSITechnologyandCir46783.2024.10631554
Publication status	Published - 2024
Event	2024 IEEE Symposium on VLSI Technology and Circuits, VLSI Technology and Circuits 2024 - Honolulu, United States Duration: 16 Jun 2024 → 20 Jun 2024

Publication series

Name	Digest of Technical Papers - Symposium on VLSI Technology
ISSN (Print)	0743-1562

Conference

Conference	2024 IEEE Symposium on VLSI Technology and Circuits, VLSI Technology and Circuits 2024
Country/Territory	United States
City	Honolulu
Period	16/06/24 → 20/06/24

Bibliographical note

Publisher Copyright:
© 2024 IEEE.

Access to Document

10.1109/VLSITechnologyandCir46783.2024.10631554

Cite this

Choi, E., Lukito, V., Choi, I., Lee, S., Chang, I. J., Ha, S., & Je, M. (2024). A Δ-Based Spike Sorting SoC with End-to-End Implementation of Event-Driven Binary Autoencoder Neural Network in Analog CIM Achieving 94.54% Accuracy and 3.11μW/ch. In 2024 IEEE Symposium on VLSI Technology and Circuits, VLSI Technology and Circuits 2024 (Digest of Technical Papers - Symposium on VLSI Technology). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/VLSITechnologyandCir46783.2024.10631554

Choi, Edward ; Lukito, Vincent ; Choi, Injun et al. / A Δ-Based Spike Sorting SoC with End-to-End Implementation of Event-Driven Binary Autoencoder Neural Network in Analog CIM Achieving 94.54% Accuracy and 3.11μW/ch. 2024 IEEE Symposium on VLSI Technology and Circuits, VLSI Technology and Circuits 2024. Institute of Electrical and Electronics Engineers Inc., 2024. (Digest of Technical Papers - Symposium on VLSI Technology).

@inproceedings{2d80f3d8a507481d8077eedea643cf6e,

title = "A Δ-Based Spike Sorting SoC with End-to-End Implementation of Event-Driven Binary Autoencoder Neural Network in Analog CIM Achieving 94.54\% Accuracy and 3.11μW/ch",

abstract = "The proposed Δ -based spike sorting (SS) SoC is the first on-chip implementation of an analog computing-in-memory (CIM) binary autoencoder neural network (B-AENN) feature extraction with enhanced spike detection adopting Δ -spikes, resulting in the highest on-chip SS classification accuracy of 94.54\%. It also allows to reduce the digital data transmission rate by 48.8× compared to prior SS systems.",

author = "Edward Choi and Vincent Lukito and Injun Choi and Seoyoung Lee and Chang, \{Ik Joon\} and Sohmyung Ha and Minkyu Je",

note = "Publisher Copyright: {\textcopyright} 2024 IEEE.; 2024 IEEE Symposium on VLSI Technology and Circuits, VLSI Technology and Circuits 2024 ; Conference date: 16-06-2024 Through 20-06-2024",

year = "2024",

doi = "10.1109/VLSITechnologyandCir46783.2024.10631554",

language = "English",

series = "Digest of Technical Papers - Symposium on VLSI Technology",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

booktitle = "2024 IEEE Symposium on VLSI Technology and Circuits, VLSI Technology and Circuits 2024",

address = "United States",

}

Choi, E, Lukito, V, Choi, I, Lee, S, Chang, IJ, Ha, S & Je, M 2024, A Δ-Based Spike Sorting SoC with End-to-End Implementation of Event-Driven Binary Autoencoder Neural Network in Analog CIM Achieving 94.54% Accuracy and 3.11μW/ch. in 2024 IEEE Symposium on VLSI Technology and Circuits, VLSI Technology and Circuits 2024. Digest of Technical Papers - Symposium on VLSI Technology, Institute of Electrical and Electronics Engineers Inc., 2024 IEEE Symposium on VLSI Technology and Circuits, VLSI Technology and Circuits 2024, Honolulu, United States, 16/06/24. https://doi.org/10.1109/VLSITechnologyandCir46783.2024.10631554

A Δ-Based Spike Sorting SoC with End-to-End Implementation of Event-Driven Binary Autoencoder Neural Network in Analog CIM Achieving 94.54% Accuracy and 3.11μW/ch. / Choi, Edward; Lukito, Vincent; Choi, Injun et al.
2024 IEEE Symposium on VLSI Technology and Circuits, VLSI Technology and Circuits 2024. Institute of Electrical and Electronics Engineers Inc., 2024. (Digest of Technical Papers - Symposium on VLSI Technology).

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

TY - GEN

T1 - A Δ-Based Spike Sorting SoC with End-to-End Implementation of Event-Driven Binary Autoencoder Neural Network in Analog CIM Achieving 94.54% Accuracy and 3.11μW/ch

AU - Choi, Edward

AU - Lukito, Vincent

AU - Choi, Injun

AU - Lee, Seoyoung

AU - Chang, Ik Joon

AU - Ha, Sohmyung

AU - Je, Minkyu

PY - 2024

Y1 - 2024

N2 - The proposed Δ -based spike sorting (SS) SoC is the first on-chip implementation of an analog computing-in-memory (CIM) binary autoencoder neural network (B-AENN) feature extraction with enhanced spike detection adopting Δ -spikes, resulting in the highest on-chip SS classification accuracy of 94.54%. It also allows to reduce the digital data transmission rate by 48.8× compared to prior SS systems.

AB - The proposed Δ -based spike sorting (SS) SoC is the first on-chip implementation of an analog computing-in-memory (CIM) binary autoencoder neural network (B-AENN) feature extraction with enhanced spike detection adopting Δ -spikes, resulting in the highest on-chip SS classification accuracy of 94.54%. It also allows to reduce the digital data transmission rate by 48.8× compared to prior SS systems.

UR - https://www.scopus.com/pages/publications/85203585636

U2 - 10.1109/VLSITechnologyandCir46783.2024.10631554

DO - 10.1109/VLSITechnologyandCir46783.2024.10631554

M3 - Conference contribution

AN - SCOPUS:85203585636

T3 - Digest of Technical Papers - Symposium on VLSI Technology

BT - 2024 IEEE Symposium on VLSI Technology and Circuits, VLSI Technology and Circuits 2024

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2024 IEEE Symposium on VLSI Technology and Circuits, VLSI Technology and Circuits 2024

Y2 - 16 June 2024 through 20 June 2024

ER -

Choi E, Lukito V, Choi I, Lee S, Chang IJ, Ha S et al. A Δ-Based Spike Sorting SoC with End-to-End Implementation of Event-Driven Binary Autoencoder Neural Network in Analog CIM Achieving 94.54% Accuracy and 3.11μW/ch. In 2024 IEEE Symposium on VLSI Technology and Circuits, VLSI Technology and Circuits 2024. Institute of Electrical and Electronics Engineers Inc. 2024. (Digest of Technical Papers - Symposium on VLSI Technology). doi: 10.1109/VLSITechnologyandCir46783.2024.10631554

A Δ-Based Spike Sorting SoC with End-to-End Implementation of Event-Driven Binary Autoencoder Neural Network in Analog CIM Achieving 94.54% Accuracy and 3.11μW/ch

Abstract

Publication series

Conference

Bibliographical note

Access to Document

Other files and links

Fingerprint

Cite this