Digital Electronic and Analog Photonic Acceleration for Point Cloud Classifiers

James Garofolo; Taichu Shi; Thomas Ferreira de Lima; Paul Prucnal; Ben Wu

Digital Electronic and Analog Photonic Acceleration for Point Cloud Classifiers

James Garofolo
, Taichu Shi
, Thomas Ferreira de Lima
, Paul Prucnal
, Ben Wu

Research output: Contribution to conference › Paper › peer-review

Abstract

We propose a method for training photonic neuromorphic accelerators using only software, and demonstrate its potential in autonomous driving.The method is modular and drop-in compatible with existing PyTorch training pipelines.

Original language	English (US)
State	Published - 2023
Externally published	Yes
Event	2023 Laser Science, LS 2023 - Tacoma, United States Duration: Oct 9 2023 → Oct 12 2023

Conference

Conference	2023 Laser Science, LS 2023
Country/Territory	United States
City	Tacoma
Period	10/9/23 → 10/12/23

All Science Journal Classification (ASJC) codes

Nuclear Energy and Engineering
Electronic, Optical and Magnetic Materials
Nuclear and High Energy Physics
Atomic and Molecular Physics, and Optics

Cite this

@conference{0269f8c4bb1f499f820885783e911d21,

title = "Digital Electronic and Analog Photonic Acceleration for Point Cloud Classifiers",

abstract = "We propose a method for training photonic neuromorphic accelerators using only software, and demonstrate its potential in autonomous driving.The method is modular and drop-in compatible with existing PyTorch training pipelines.",

author = "James Garofolo and Taichu Shi and \{de Lima\}, \{Thomas Ferreira\} and Paul Prucnal and Ben Wu",

note = "Publisher Copyright: {\textcopyright} Optica Publishing Group 2023, {\textcopyright} 2023 The Author(s); 2023 Laser Science, LS 2023 ; Conference date: 09-10-2023 Through 12-10-2023",

year = "2023",

language = "English (US)",

}

TY - CONF

T1 - Digital Electronic and Analog Photonic Acceleration for Point Cloud Classifiers

AU - Garofolo, James

AU - Shi, Taichu

AU - de Lima, Thomas Ferreira

AU - Prucnal, Paul

AU - Wu, Ben

PY - 2023

Y1 - 2023

N2 - We propose a method for training photonic neuromorphic accelerators using only software, and demonstrate its potential in autonomous driving.The method is modular and drop-in compatible with existing PyTorch training pipelines.

AB - We propose a method for training photonic neuromorphic accelerators using only software, and demonstrate its potential in autonomous driving.The method is modular and drop-in compatible with existing PyTorch training pipelines.

UR - https://www.scopus.com/pages/publications/85215511971

UR - https://www.scopus.com/pages/publications/85215511971#tab=citedBy

M3 - Paper

AN - SCOPUS:85215511971

T2 - 2023 Laser Science, LS 2023

Y2 - 9 October 2023 through 12 October 2023

ER -

Digital Electronic and Analog Photonic Acceleration for Point Cloud Classifiers

Abstract

Conference

All Science Journal Classification (ASJC) codes

Other files and links

Fingerprint

Cite this