Evaluating energy-efficiency of DRAM channel interleaving schemes for multithreaded programs

Satoshi Imamura, Yuichiro Yasui, Koji Inoue, Takatsugu Ono, Hiroshi Sasaki, Katsuki Fujisawa

Research output: Contribution to journalArticle

Abstract

The power consumption of server platforms has been increasing as the amount of hardware resources equipped on them is increased. Especially, the capacity of DRAM continues to grow, and it is not rare that DRAM consumes higher power than processors on modern servers. Therefore, a reduction in the DRAM energy consumption is a critical challenge to reduce the system-level energy consumption. Although it is well known that improving row buffer locality (RBL) and bank-level parallelism (BLP) is effective to reduce the DRAM energy consumption, our preliminary evaluation on a real server demonstrates that RBL is generally low across 15 multithreaded benchmarks. In this paper, we investigate the memory access patterns of these benchmarks using a simulator and observe that cache line-grained channel interleaving schemes, which are widely applied to modern servers including multiple memory channels, hurt the RBL each of the benchmarks potentially possesses. In order to address this problem, we focus on a row-grained channel interleaving scheme and compare it with three cache line-grained schemes. Our evaluation shows that it reduces the DRAM energy consumption by 16.7%, 12.3%, and 5.5%on average (up to 34.7%, 28.2%, and 12.0%) compared to the other schemes, respectively.

Original languageEnglish
Pages (from-to)2247-2257
Number of pages11
JournalIEICE Transactions on Information and Systems
VolumeE101D
Issue number9
DOIs
Publication statusPublished - Sep 2018

Fingerprint

Dynamic random access storage
Energy efficiency
Servers
Energy utilization
Data storage equipment
Electric power utilization
Simulators
Hardware

All Science Journal Classification (ASJC) codes

  • Software
  • Hardware and Architecture
  • Computer Vision and Pattern Recognition
  • Electrical and Electronic Engineering
  • Artificial Intelligence

Cite this

Evaluating energy-efficiency of DRAM channel interleaving schemes for multithreaded programs. / Imamura, Satoshi; Yasui, Yuichiro; Inoue, Koji; Ono, Takatsugu; Sasaki, Hiroshi; Fujisawa, Katsuki.

In: IEICE Transactions on Information and Systems, Vol. E101D, No. 9, 09.2018, p. 2247-2257.

Research output: Contribution to journalArticle

@article{b31f8669f48245d4ae615c4d21da1c1f,
title = "Evaluating energy-efficiency of DRAM channel interleaving schemes for multithreaded programs",
abstract = "The power consumption of server platforms has been increasing as the amount of hardware resources equipped on them is increased. Especially, the capacity of DRAM continues to grow, and it is not rare that DRAM consumes higher power than processors on modern servers. Therefore, a reduction in the DRAM energy consumption is a critical challenge to reduce the system-level energy consumption. Although it is well known that improving row buffer locality (RBL) and bank-level parallelism (BLP) is effective to reduce the DRAM energy consumption, our preliminary evaluation on a real server demonstrates that RBL is generally low across 15 multithreaded benchmarks. In this paper, we investigate the memory access patterns of these benchmarks using a simulator and observe that cache line-grained channel interleaving schemes, which are widely applied to modern servers including multiple memory channels, hurt the RBL each of the benchmarks potentially possesses. In order to address this problem, we focus on a row-grained channel interleaving scheme and compare it with three cache line-grained schemes. Our evaluation shows that it reduces the DRAM energy consumption by 16.7{\%}, 12.3{\%}, and 5.5{\%}on average (up to 34.7{\%}, 28.2{\%}, and 12.0{\%}) compared to the other schemes, respectively.",
author = "Satoshi Imamura and Yuichiro Yasui and Koji Inoue and Takatsugu Ono and Hiroshi Sasaki and Katsuki Fujisawa",
year = "2018",
month = "9",
doi = "10.1587/transinf.2017EDP7296",
language = "English",
volume = "E101D",
pages = "2247--2257",
journal = "IEICE Transactions on Information and Systems",
issn = "0916-8532",
publisher = "一般社団法人電子情報通信学会",
number = "9",

}

TY - JOUR

T1 - Evaluating energy-efficiency of DRAM channel interleaving schemes for multithreaded programs

AU - Imamura, Satoshi

AU - Yasui, Yuichiro

AU - Inoue, Koji

AU - Ono, Takatsugu

AU - Sasaki, Hiroshi

AU - Fujisawa, Katsuki

PY - 2018/9

Y1 - 2018/9

N2 - The power consumption of server platforms has been increasing as the amount of hardware resources equipped on them is increased. Especially, the capacity of DRAM continues to grow, and it is not rare that DRAM consumes higher power than processors on modern servers. Therefore, a reduction in the DRAM energy consumption is a critical challenge to reduce the system-level energy consumption. Although it is well known that improving row buffer locality (RBL) and bank-level parallelism (BLP) is effective to reduce the DRAM energy consumption, our preliminary evaluation on a real server demonstrates that RBL is generally low across 15 multithreaded benchmarks. In this paper, we investigate the memory access patterns of these benchmarks using a simulator and observe that cache line-grained channel interleaving schemes, which are widely applied to modern servers including multiple memory channels, hurt the RBL each of the benchmarks potentially possesses. In order to address this problem, we focus on a row-grained channel interleaving scheme and compare it with three cache line-grained schemes. Our evaluation shows that it reduces the DRAM energy consumption by 16.7%, 12.3%, and 5.5%on average (up to 34.7%, 28.2%, and 12.0%) compared to the other schemes, respectively.

AB - The power consumption of server platforms has been increasing as the amount of hardware resources equipped on them is increased. Especially, the capacity of DRAM continues to grow, and it is not rare that DRAM consumes higher power than processors on modern servers. Therefore, a reduction in the DRAM energy consumption is a critical challenge to reduce the system-level energy consumption. Although it is well known that improving row buffer locality (RBL) and bank-level parallelism (BLP) is effective to reduce the DRAM energy consumption, our preliminary evaluation on a real server demonstrates that RBL is generally low across 15 multithreaded benchmarks. In this paper, we investigate the memory access patterns of these benchmarks using a simulator and observe that cache line-grained channel interleaving schemes, which are widely applied to modern servers including multiple memory channels, hurt the RBL each of the benchmarks potentially possesses. In order to address this problem, we focus on a row-grained channel interleaving scheme and compare it with three cache line-grained schemes. Our evaluation shows that it reduces the DRAM energy consumption by 16.7%, 12.3%, and 5.5%on average (up to 34.7%, 28.2%, and 12.0%) compared to the other schemes, respectively.

UR - http://www.scopus.com/inward/record.url?scp=85053835457&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85053835457&partnerID=8YFLogxK

U2 - 10.1587/transinf.2017EDP7296

DO - 10.1587/transinf.2017EDP7296

M3 - Article

AN - SCOPUS:85053835457

VL - E101D

SP - 2247

EP - 2257

JO - IEICE Transactions on Information and Systems

JF - IEICE Transactions on Information and Systems

SN - 0916-8532

IS - 9

ER -