Efficient memory page management for NVDIMM-based big data processing environments

Seog Min Kwon; Hyokyung Bahn

doi:10.1109/ICISCE.2017.67

Efficient memory page management for NVDIMM-based big data processing environments

Seog Min Kwon, Hyokyung Bahn

Computer Science & Engineering

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

1 Scopus citations

Abstract

Due to the recent computing paradigm shift from desktop to cloud environments, NVDIMM-based memory architectures for big data processing are catching considerable interest. In this paper, we present an efficient page management scheme for large-scale NVDIMM memory. The proposed architecture carefully determines the page size of a large-scale memory system such that TLB hit ratio is improved significantly without increasing the page fault ratio, thereby allowing big-data processing efficiently, which was not easy in conventional DRAM-based memory systems. The main contribution of this paper is that it proposes a new paradigm of "NVDIMM-based large memory architecture" instead of conventional "DRAM-based memory system" and presents efficient management techniques for that architecture by considering both TLB performances and page fault rates.

Original language	English
Title of host publication	Proceedings - 2017 4th International Conference on Information Science and Control Engineering, ICISCE 2017
Editors	Ying Dai, Shaozi Li, Yun Cheng
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	283-287
Number of pages	5
ISBN (Electronic)	9781538630136
DOIs	https://doi.org/10.1109/ICISCE.2017.67
State	Published - 14 Nov 2017
Event	4th International Conference on Information Science and Control Engineering, ICISCE 2017 - Changsha, Hunan, China Duration: 21 Jul 2017 → 23 Jul 2017

Publication series

Name	Proceedings - 2017 4th International Conference on Information Science and Control Engineering, ICISCE 2017

Conference

Conference	4th International Conference on Information Science and Control Engineering, ICISCE 2017
Country/Territory	China
City	Changsha, Hunan
Period	21/07/17 → 23/07/17

Keywords

Big data
NVDIMM
Page fault
Page management
TLB

Access to Document

10.1109/ICISCE.2017.67

Cite this

Kwon, S. M., & Bahn, H. (2017). Efficient memory page management for NVDIMM-based big data processing environments. In Y. Dai, S. Li, & Y. Cheng (Eds.), Proceedings - 2017 4th International Conference on Information Science and Control Engineering, ICISCE 2017 (pp. 283-287). Article 8110294 (Proceedings - 2017 4th International Conference on Information Science and Control Engineering, ICISCE 2017). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICISCE.2017.67

Kwon, Seog Min ; Bahn, Hyokyung. / Efficient memory page management for NVDIMM-based big data processing environments. Proceedings - 2017 4th International Conference on Information Science and Control Engineering, ICISCE 2017. editor / Ying Dai ; Shaozi Li ; Yun Cheng. Institute of Electrical and Electronics Engineers Inc., 2017. pp. 283-287 (Proceedings - 2017 4th International Conference on Information Science and Control Engineering, ICISCE 2017).

@inproceedings{51e06aa32d2d47f38cc8d8c40e5746f6,

title = "Efficient memory page management for NVDIMM-based big data processing environments",

abstract = "Due to the recent computing paradigm shift from desktop to cloud environments, NVDIMM-based memory architectures for big data processing are catching considerable interest. In this paper, we present an efficient page management scheme for large-scale NVDIMM memory. The proposed architecture carefully determines the page size of a large-scale memory system such that TLB hit ratio is improved significantly without increasing the page fault ratio, thereby allowing big-data processing efficiently, which was not easy in conventional DRAM-based memory systems. The main contribution of this paper is that it proposes a new paradigm of {"}NVDIMM-based large memory architecture{"} instead of conventional {"}DRAM-based memory system{"} and presents efficient management techniques for that architecture by considering both TLB performances and page fault rates.",

keywords = "Big data, NVDIMM, Page fault, Page management, TLB",

author = "Kwon, {Seog Min} and Hyokyung Bahn",

note = "Funding Information: V. ACKNOWLEDGMENT This research was supported by the MSIP (Ministry of Science, ICT, and Future Planning), Korea, under the Seoul Accord Vitalization Program (IITP-2017-2012-0-00442) supervised by the IITP (Institute for Information & communications Technology Promotion). Hyokyung Bahn is the corresponding author of this paper. Publisher Copyright: {\textcopyright} 2017 IEEE.; 4th International Conference on Information Science and Control Engineering, ICISCE 2017 ; Conference date: 21-07-2017 Through 23-07-2017",

year = "2017",

month = nov,

day = "14",

doi = "10.1109/ICISCE.2017.67",

language = "English",

series = "Proceedings - 2017 4th International Conference on Information Science and Control Engineering, ICISCE 2017",

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

pages = "283--287",

editor = "Ying Dai and Shaozi Li and Yun Cheng",

booktitle = "Proceedings - 2017 4th International Conference on Information Science and Control Engineering, ICISCE 2017",

}

Kwon, SM & Bahn, H 2017, Efficient memory page management for NVDIMM-based big data processing environments. in Y Dai, S Li & Y Cheng (eds), Proceedings - 2017 4th International Conference on Information Science and Control Engineering, ICISCE 2017., 8110294, Proceedings - 2017 4th International Conference on Information Science and Control Engineering, ICISCE 2017, Institute of Electrical and Electronics Engineers Inc., pp. 283-287, 4th International Conference on Information Science and Control Engineering, ICISCE 2017, Changsha, Hunan, China, 21/07/17. https://doi.org/10.1109/ICISCE.2017.67

Efficient memory page management for NVDIMM-based big data processing environments. / Kwon, Seog Min; Bahn, Hyokyung.
Proceedings - 2017 4th International Conference on Information Science and Control Engineering, ICISCE 2017. ed. / Ying Dai; Shaozi Li; Yun Cheng. Institute of Electrical and Electronics Engineers Inc., 2017. p. 283-287 8110294 (Proceedings - 2017 4th International Conference on Information Science and Control Engineering, ICISCE 2017).

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

TY - GEN

T1 - Efficient memory page management for NVDIMM-based big data processing environments

AU - Kwon, Seog Min

AU - Bahn, Hyokyung

N1 - Funding Information: V. ACKNOWLEDGMENT This research was supported by the MSIP (Ministry of Science, ICT, and Future Planning), Korea, under the Seoul Accord Vitalization Program (IITP-2017-2012-0-00442) supervised by the IITP (Institute for Information & communications Technology Promotion). Hyokyung Bahn is the corresponding author of this paper. Publisher Copyright: © 2017 IEEE.

PY - 2017/11/14

Y1 - 2017/11/14

N2 - Due to the recent computing paradigm shift from desktop to cloud environments, NVDIMM-based memory architectures for big data processing are catching considerable interest. In this paper, we present an efficient page management scheme for large-scale NVDIMM memory. The proposed architecture carefully determines the page size of a large-scale memory system such that TLB hit ratio is improved significantly without increasing the page fault ratio, thereby allowing big-data processing efficiently, which was not easy in conventional DRAM-based memory systems. The main contribution of this paper is that it proposes a new paradigm of "NVDIMM-based large memory architecture" instead of conventional "DRAM-based memory system" and presents efficient management techniques for that architecture by considering both TLB performances and page fault rates.

AB - Due to the recent computing paradigm shift from desktop to cloud environments, NVDIMM-based memory architectures for big data processing are catching considerable interest. In this paper, we present an efficient page management scheme for large-scale NVDIMM memory. The proposed architecture carefully determines the page size of a large-scale memory system such that TLB hit ratio is improved significantly without increasing the page fault ratio, thereby allowing big-data processing efficiently, which was not easy in conventional DRAM-based memory systems. The main contribution of this paper is that it proposes a new paradigm of "NVDIMM-based large memory architecture" instead of conventional "DRAM-based memory system" and presents efficient management techniques for that architecture by considering both TLB performances and page fault rates.

KW - Big data

KW - NVDIMM

KW - Page fault

KW - Page management

KW - TLB

UR - http://www.scopus.com/inward/record.url?scp=85040577659&partnerID=8YFLogxK

U2 - 10.1109/ICISCE.2017.67

DO - 10.1109/ICISCE.2017.67

M3 - Conference contribution

AN - SCOPUS:85040577659

T3 - Proceedings - 2017 4th International Conference on Information Science and Control Engineering, ICISCE 2017

SP - 283

EP - 287

BT - Proceedings - 2017 4th International Conference on Information Science and Control Engineering, ICISCE 2017

A2 - Dai, Ying

A2 - Li, Shaozi

A2 - Cheng, Yun

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 4th International Conference on Information Science and Control Engineering, ICISCE 2017

Y2 - 21 July 2017 through 23 July 2017

ER -

Kwon SM, Bahn H. Efficient memory page management for NVDIMM-based big data processing environments. In Dai Y, Li S, Cheng Y, editors, Proceedings - 2017 4th International Conference on Information Science and Control Engineering, ICISCE 2017. Institute of Electrical and Electronics Engineers Inc. 2017. p. 283-287. 8110294. (Proceedings - 2017 4th International Conference on Information Science and Control Engineering, ICISCE 2017). doi: 10.1109/ICISCE.2017.67

Efficient memory page management for NVDIMM-based big data processing environments

Abstract

Publication series

Conference

Keywords

Access to Document

Fingerprint

Cite this