Data Throughput for Efficient Photonic Neural Network Accelerators

被引：0

作者：

Schwartz, Russell L. T. ^{[1
,2
]}

Jahannia, Belal ^{[1
,2
]}

Peserico, Nicola ^{[1
,2
]}

Dalir, Hamed ^{[1
,2
]}

Sorger, Volker J. ^{[1
,2
]}

机构：

[1] Univ Florida, Dept Elect & Comp Engn, Gainesville, FL 32611 USA

[2] Univ Florida, Florida Semicond Inst, Gainesville, FL 32611 USA

来源：

2024 IEEE SILICON PHOTONICS CONFERENCE, SIPHOTONICS | 2024年

关键词：

Photonic Tensor Core; Machine Learning; Neural Network; Silicon Photonics;

D O I：

10.1109/SiPhotonics60897.2024.10543636

中图分类号：

TM [电工技术]; TN [电子技术、通信技术];

学科分类号：

0808 ; 0809 ;

摘要：

Machine Learning has become a dominant technology, spurring the invention of photonic systems to implement neural network tasks. These photonic systems offer high throughput operations with low power but require large bandwidths of data to operate with maximum efficiency. Here we provide an analysis of the required data bandwidths, reaching nearly 1 Tbps for a single chip and necessitating the use of high bandwidth memory.

引用

页数：2

共 4 条

[1]

Ma Xiaoxuan, 2022, HIGH DENSITY INTEGRA

[2] Integrated Photonic Tensor Processing Unit for a Matrix Multiply: A Review [J].

Peserico, Nicola ;

Shastri, Bhavin J. ;

Sorger, Volker J. .

JOURNAL OF LIGHTWAVE TECHNOLOGY, 2023, 41 (12) :3704-3716

[3]

Samajdar A., 2018, arXiv

[4] Photonics for artificial intelligence and neuromorphic computing [J].

Shastri, Bhavin J. ;

Tait, Alexander N. ;

de Lima, T. Ferreira ;

Pernice, Wolfram H. P. ;

Bhaskaran, Harish ;

Wright, C. D. ;

Prucnal, Paul R. .

NATURE PHOTONICS, 2021, 15 (02) :102-114

← 1 →