Investigation of communication overhead of SoC lookaside accelerators

Alperen Bolat; Fahad Siddiqui; Sakir Sezer; Kasim Tasdemir; Rafiullah Khan

doi:10.1109/SOCC58585.2023.10257113

Investigation of communication overhead of SoC lookaside accelerators

Alperen Bolat
, Fahad Siddiqui
, Sakir Sezer
, Kasim Tasdemir
, Rafiullah Khan

School of Electronics, Electrical Engineering and Computer Science

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

7 Citations (Scopus)

320 Downloads (Pure)

Abstract

Lookaside accelerators (LAAs) enable opportunities to asynchronously offload computationally intensive tasks from the general-purpose processors to the accelerator in order to improve the overall performance of a given system. However, the communication overhead involved in moving data between the LAA and the CPU can be significant and can impact anticipated performance, diminishing the benefits of a LAA. This paper examines the communication overhead between LAAs and CPUs, focusing on the communication setup and data transmission components. We develop a mathematical model to analyse and capture the problem and investigate the conditions under which hardware accelerators provide effective performance. To verify and validate the proposed model, we carried out studies using an ARM-based platform. Experimental results reveal that the length of data send to the LAA significantly impacts the overall offload benefit of the LAA when compared with software-only solutions. Communication overhead, specifically communication setup latency, has a diminishing impact on smaller data lengths. The presented mathematical model and experimental results clearly demonstrate that the targeted data length for offload is a critical parameter when designing LAAs.

Original language	English
Title of host publication	Proceedings of the IEEE 36th International System-on-Chip Conference, SOCC 2023
Editors	Jürgen Becker, Andrew Marshall, Tanja Harbaum, Amlan Ganguly, Fahad Siddiqui, Kieran McLaughlin
Publisher	Institute of Electrical and Electronics Engineers Inc.
Number of pages	6
ISBN (Electronic)	9798350300116
ISBN (Print)	9798350300123
DOIs	https://doi.org/10.1109/SOCC58585.2023.10257113
Publication status	Published - 22 Sept 2023
Event	36th IEEE International System-on-Chip Conference 2023 - Hyatt Regency, Santa Clara, United States Duration: 05 Sept 2023 → 08 Sept 2023 https://www.ieee-socc.org/

Publication series

Name	IEEE International SOC Conference: Proceedings
ISSN (Print)	2164-1676
ISSN (Electronic)	2164-1706

Conference

Conference	36th IEEE International System-on-Chip Conference 2023
Abbreviated title	SOCC 2023
Country/Territory	United States
City	Santa Clara
Period	05/09/2023 → 08/09/2023
Internet address	https://www.ieee-socc.org/

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 9 Industry, Innovation, and Infrastructure
SDG 11 Sustainable Cities and Communities
SDG 17 Partnerships for the Goals

Keywords

Hardware accelerator
Communication Overhead
Performance Evaluation
AMBA AXI
FPGA
System-on-Chip
Accelerated Computing

ASJC Scopus subject areas

Electrical and Electronic Engineering
Architecture
General Computer Science

Access to Document

10.1109/SOCC58585.2023.10257113Licence: Unspecified

Investigation of communication overhead of SoC lookaside accelerators
Copyright 2023 IEEE This work is made available online in accordance with the publisher’s policies. Please refer to any applicable terms of use of the publisher.
Accepted author manuscript, 23.1 MBLicence: Unspecified

Cite this

Bolat, A., Siddiqui, F., Sezer, S., Tasdemir, K., & Khan, R. (2023). Investigation of communication overhead of SoC lookaside accelerators. In J. Becker, A. Marshall, T. Harbaum, A. Ganguly, F. Siddiqui, & K. McLaughlin (Eds.), Proceedings of the IEEE 36th International System-on-Chip Conference, SOCC 2023 (IEEE International SOC Conference: Proceedings). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/SOCC58585.2023.10257113

Bolat, Alperen ; Siddiqui, Fahad ; Sezer, Sakir et al. / Investigation of communication overhead of SoC lookaside accelerators. Proceedings of the IEEE 36th International System-on-Chip Conference, SOCC 2023. editor / Jürgen Becker ; Andrew Marshall ; Tanja Harbaum ; Amlan Ganguly ; Fahad Siddiqui ; Kieran McLaughlin. Institute of Electrical and Electronics Engineers Inc., 2023. (IEEE International SOC Conference: Proceedings).

@inproceedings{f513416c89b145c6a20b5090c32f99c2,

title = "Investigation of communication overhead of SoC lookaside accelerators",

abstract = "Lookaside accelerators (LAAs) enable opportunities to asynchronously offload computationally intensive tasks from the general-purpose processors to the accelerator in order to improve the overall performance of a given system. However, the communication overhead involved in moving data between the LAA and the CPU can be significant and can impact anticipated performance, diminishing the benefits of a LAA. This paper examines the communication overhead between LAAs and CPUs, focusing on the communication setup and data transmission components. We develop a mathematical model to analyse and capture the problem and investigate the conditions under which hardware accelerators provide effective performance. To verify and validate the proposed model, we carried out studies using an ARM-based platform. Experimental results reveal that the length of data send to the LAA significantly impacts the overall offload benefit of the LAA when compared with software-only solutions. Communication overhead, specifically communication setup latency, has a diminishing impact on smaller data lengths. The presented mathematical model and experimental results clearly demonstrate that the targeted data length for offload is a critical parameter when designing LAAs. ",

keywords = "Hardware accelerator, Communication Overhead, Performance Evaluation, AMBA AXI, FPGA, System-on-Chip, Accelerated Computing",

author = "Alperen Bolat and Fahad Siddiqui and Sakir Sezer and Kasim Tasdemir and Rafiullah Khan",

year = "2023",

month = sep,

day = "22",

doi = "10.1109/SOCC58585.2023.10257113",

language = "English",

isbn = "9798350300123",

series = "IEEE International SOC Conference: Proceedings",

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

editor = "J{\"u}rgen Becker and Andrew Marshall and Tanja Harbaum and Amlan Ganguly and Fahad Siddiqui and Kieran McLaughlin",

booktitle = "Proceedings of the IEEE 36th International System-on-Chip Conference, SOCC 2023",

address = "United States",

note = "36th IEEE International System-on-Chip Conference 2023, SOCC 2023 ; Conference date: 05-09-2023 Through 08-09-2023",

url = "https://www.ieee-socc.org/",

}

Bolat, A, Siddiqui, F, Sezer, S, Tasdemir, K & Khan, R 2023, Investigation of communication overhead of SoC lookaside accelerators. in J Becker, A Marshall, T Harbaum, A Ganguly, F Siddiqui & K McLaughlin (eds), Proceedings of the IEEE 36th International System-on-Chip Conference, SOCC 2023. IEEE International SOC Conference: Proceedings, Institute of Electrical and Electronics Engineers Inc., 36th IEEE International System-on-Chip Conference 2023, Santa Clara, California, United States, 05/09/2023. https://doi.org/10.1109/SOCC58585.2023.10257113

Investigation of communication overhead of SoC lookaside accelerators. / Bolat, Alperen; Siddiqui, Fahad; Sezer, Sakir et al.
Proceedings of the IEEE 36th International System-on-Chip Conference, SOCC 2023. ed. / Jürgen Becker; Andrew Marshall; Tanja Harbaum; Amlan Ganguly; Fahad Siddiqui; Kieran McLaughlin. Institute of Electrical and Electronics Engineers Inc., 2023. (IEEE International SOC Conference: Proceedings).

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

TY - GEN

T1 - Investigation of communication overhead of SoC lookaside accelerators

AU - Bolat, Alperen

AU - Siddiqui, Fahad

AU - Sezer, Sakir

AU - Tasdemir, Kasim

AU - Khan, Rafiullah

PY - 2023/9/22

Y1 - 2023/9/22

N2 - Lookaside accelerators (LAAs) enable opportunities to asynchronously offload computationally intensive tasks from the general-purpose processors to the accelerator in order to improve the overall performance of a given system. However, the communication overhead involved in moving data between the LAA and the CPU can be significant and can impact anticipated performance, diminishing the benefits of a LAA. This paper examines the communication overhead between LAAs and CPUs, focusing on the communication setup and data transmission components. We develop a mathematical model to analyse and capture the problem and investigate the conditions under which hardware accelerators provide effective performance. To verify and validate the proposed model, we carried out studies using an ARM-based platform. Experimental results reveal that the length of data send to the LAA significantly impacts the overall offload benefit of the LAA when compared with software-only solutions. Communication overhead, specifically communication setup latency, has a diminishing impact on smaller data lengths. The presented mathematical model and experimental results clearly demonstrate that the targeted data length for offload is a critical parameter when designing LAAs.

AB - Lookaside accelerators (LAAs) enable opportunities to asynchronously offload computationally intensive tasks from the general-purpose processors to the accelerator in order to improve the overall performance of a given system. However, the communication overhead involved in moving data between the LAA and the CPU can be significant and can impact anticipated performance, diminishing the benefits of a LAA. This paper examines the communication overhead between LAAs and CPUs, focusing on the communication setup and data transmission components. We develop a mathematical model to analyse and capture the problem and investigate the conditions under which hardware accelerators provide effective performance. To verify and validate the proposed model, we carried out studies using an ARM-based platform. Experimental results reveal that the length of data send to the LAA significantly impacts the overall offload benefit of the LAA when compared with software-only solutions. Communication overhead, specifically communication setup latency, has a diminishing impact on smaller data lengths. The presented mathematical model and experimental results clearly demonstrate that the targeted data length for offload is a critical parameter when designing LAAs.

KW - Hardware accelerator

KW - Communication Overhead

KW - Performance Evaluation

KW - AMBA AXI

KW - FPGA

KW - System-on-Chip

KW - Accelerated Computing

U2 - 10.1109/SOCC58585.2023.10257113

DO - 10.1109/SOCC58585.2023.10257113

M3 - Conference contribution

AN - SCOPUS:85174627400

SN - 9798350300123

T3 - IEEE International SOC Conference: Proceedings

BT - Proceedings of the IEEE 36th International System-on-Chip Conference, SOCC 2023

A2 - Becker, Jürgen

A2 - Marshall, Andrew

A2 - Harbaum, Tanja

A2 - Ganguly, Amlan

A2 - Siddiqui, Fahad

A2 - McLaughlin, Kieran

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 36th IEEE International System-on-Chip Conference 2023

Y2 - 5 September 2023 through 8 September 2023

ER -

Bolat A, Siddiqui F, Sezer S, Tasdemir K, Khan R. Investigation of communication overhead of SoC lookaside accelerators. In Becker J, Marshall A, Harbaum T, Ganguly A, Siddiqui F, McLaughlin K, editors, Proceedings of the IEEE 36th International System-on-Chip Conference, SOCC 2023. Institute of Electrical and Electronics Engineers Inc. 2023. (IEEE International SOC Conference: Proceedings). doi: 10.1109/SOCC58585.2023.10257113

Investigation of communication overhead of SoC lookaside accelerators

Abstract

Publication series

Conference

UN SDGs

Keywords

ASJC Scopus subject areas

Access to Document

Fingerprint

Cite this