Workload-Driven Analysis on the Performance Characteristics of GPU-Accelerated DBMSes

Junyoung An; Young Kyoon Suh; Byungchul Tak

doi:10.1587/transinf.2022EDL8008

Workload-Driven Analysis on the Performance Characteristics of GPU-Accelerated DBMSes

Junyoung An, Young Kyoon Suh, Byungchul Tak

Kyungpook National University

Research output: Contribution to journal › Article › peer-review

1 Scopus citations

Abstract

This letter conducts an in-depth empirical analysis of the influence of various query characteristics on the performance of modern GPU DBMSes. Our analysis reveals that, although they can efficiently process concurrent queries, the GPU DBMSes we consider still should address various performance concerns including n-way joins, aggregates, and selective scans.

Original language	English
Pages (from-to)	1984-1989
Number of pages	6
Journal	IEICE Transactions on Information and Systems
Volume	E105D
Issue number	11
DOIs	https://doi.org/10.1587/transinf.2022EDL8008
State	Published - Nov 2022

Keywords

GPU DBMSes
performance evaluation
TPC-H

Access to Document

10.1587/transinf.2022EDL8008

Cite this

@article{96eed9c1679e472a92a19b8ae1c8442d,

title = "Workload-Driven Analysis on the Performance Characteristics of GPU-Accelerated DBMSes",

abstract = "This letter conducts an in-depth empirical analysis of the influence of various query characteristics on the performance of modern GPU DBMSes. Our analysis reveals that, although they can efficiently process concurrent queries, the GPU DBMSes we consider still should address various performance concerns including n-way joins, aggregates, and selective scans.",

keywords = "GPU DBMSes, performance evaluation, TPC-H",

author = "Junyoung An and Suh, {Young Kyoon} and Byungchul Tak",

note = "Publisher Copyright: Copyright {\textcopyright} 2022 The Institute of Electronics, Information and Communication Engineers.",

year = "2022",

month = nov,

doi = "10.1587/transinf.2022EDL8008",

language = "English",

volume = "E105D",

pages = "1984--1989",

journal = "IEICE Transactions on Information and Systems",

issn = "0916-8532",

publisher = "Maruzen Co., Ltd/Maruzen Kabushikikaisha",

number = "11",

}

TY - JOUR

T1 - Workload-Driven Analysis on the Performance Characteristics of GPU-Accelerated DBMSes

AU - An, Junyoung

AU - Suh, Young Kyoon

AU - Tak, Byungchul

PY - 2022/11

Y1 - 2022/11

N2 - This letter conducts an in-depth empirical analysis of the influence of various query characteristics on the performance of modern GPU DBMSes. Our analysis reveals that, although they can efficiently process concurrent queries, the GPU DBMSes we consider still should address various performance concerns including n-way joins, aggregates, and selective scans.

AB - This letter conducts an in-depth empirical analysis of the influence of various query characteristics on the performance of modern GPU DBMSes. Our analysis reveals that, although they can efficiently process concurrent queries, the GPU DBMSes we consider still should address various performance concerns including n-way joins, aggregates, and selective scans.

KW - GPU DBMSes

KW - performance evaluation

KW - TPC-H

UR - http://www.scopus.com/inward/record.url?scp=85141843906&partnerID=8YFLogxK

U2 - 10.1587/transinf.2022EDL8008

DO - 10.1587/transinf.2022EDL8008

M3 - Article

AN - SCOPUS:85141843906

SN - 0916-8532

VL - E105D

SP - 1984

EP - 1989

JO - IEICE Transactions on Information and Systems

JF - IEICE Transactions on Information and Systems

IS - 11

ER -

Workload-Driven Analysis on the Performance Characteristics of GPU-Accelerated DBMSes

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this