Power and Performance in I/O for Scientific Applications

K. Coloma; A. Choudhary; A. Ching; W. K. Liao; S. W. Son; M. Kandemir; L. Ward

doi:10.1109/IPDPS.2005.347

Power and Performance in I/O for Scientific Applications

K. Coloma
, A. Choudhary
, A. Ching
, W. K. Liao
, S. W. Son
, M. Kandemir
, L. Ward

Research output: Contribution to conference › Paper › peer-review

1 Scopus citations

Abstract

The I/O patterns of large scale scientific applications can often be characterized as small, non-contiguous, and regular. From a performance and power perspective, this is perhaps the worse kind of I/O for a disk. Two approaches to mitigating the mechanical limitations of disks are write-back caches and software-directed power management. Previous distributed caches are plagued by synchronization and scalability issues. The Direct Access Cache: DAChe system is a user-level distributed cached that addresses both these problems. Past work on managing disk power during run time were effective, one should be able to improve on those results by adopting a proactive scheme.

Original language	English (US)
DOIs	https://doi.org/10.1109/IPDPS.2005.347
State	Published - 2005
Event	19th IEEE International Parallel and Distributed Processing Symposium, IPDPS 2005 - Denver, CO, United States Duration: Apr 4 2005 → Apr 8 2005

Other

Other	19th IEEE International Parallel and Distributed Processing Symposium, IPDPS 2005
Country/Territory	United States
City	Denver, CO
Period	4/4/05 → 4/8/05

All Science Journal Classification (ASJC) codes

General Engineering

Access to Document

10.1109/IPDPS.2005.347

Cite this

@conference{0520d40369dc4a13b174ebc0f3072b06,

title = "Power and Performance in I/O for Scientific Applications",

abstract = "The I/O patterns of large scale scientific applications can often be characterized as small, non-contiguous, and regular. From a performance and power perspective, this is perhaps the worse kind of I/O for a disk. Two approaches to mitigating the mechanical limitations of disks are write-back caches and software-directed power management. Previous distributed caches are plagued by synchronization and scalability issues. The Direct Access Cache: DAChe system is a user-level distributed cached that addresses both these problems. Past work on managing disk power during run time were effective, one should be able to improve on those results by adopting a proactive scheme.",

author = "K. Coloma and A. Choudhary and A. Ching and Liao, \{W. K.\} and Son, \{S. W.\} and M. Kandemir and L. Ward",

note = "Publisher Copyright: {\textcopyright} 2005 IEEE Computer Society. All rights reserved.; 19th IEEE International Parallel and Distributed Processing Symposium, IPDPS 2005 ; Conference date: 04-04-2005 Through 08-04-2005",

year = "2005",

doi = "10.1109/IPDPS.2005.347",

language = "English (US)",

}

TY - CONF

T1 - Power and Performance in I/O for Scientific Applications

AU - Coloma, K.

AU - Choudhary, A.

AU - Ching, A.

AU - Liao, W. K.

AU - Son, S. W.

AU - Kandemir, M.

AU - Ward, L.

PY - 2005

Y1 - 2005

N2 - The I/O patterns of large scale scientific applications can often be characterized as small, non-contiguous, and regular. From a performance and power perspective, this is perhaps the worse kind of I/O for a disk. Two approaches to mitigating the mechanical limitations of disks are write-back caches and software-directed power management. Previous distributed caches are plagued by synchronization and scalability issues. The Direct Access Cache: DAChe system is a user-level distributed cached that addresses both these problems. Past work on managing disk power during run time were effective, one should be able to improve on those results by adopting a proactive scheme.

AB - The I/O patterns of large scale scientific applications can often be characterized as small, non-contiguous, and regular. From a performance and power perspective, this is perhaps the worse kind of I/O for a disk. Two approaches to mitigating the mechanical limitations of disks are write-back caches and software-directed power management. Previous distributed caches are plagued by synchronization and scalability issues. The Direct Access Cache: DAChe system is a user-level distributed cached that addresses both these problems. Past work on managing disk power during run time were effective, one should be able to improve on those results by adopting a proactive scheme.

UR - https://www.scopus.com/pages/publications/105030927393

UR - https://www.scopus.com/pages/publications/105030927393#tab=citedBy

U2 - 10.1109/IPDPS.2005.347

DO - 10.1109/IPDPS.2005.347

M3 - Paper

AN - SCOPUS:105030927393

T2 - 19th IEEE International Parallel and Distributed Processing Symposium, IPDPS 2005

Y2 - 4 April 2005 through 8 April 2005

ER -

Power and Performance in I/O for Scientific Applications

Abstract

Other

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this