Loading…

Dynamic and adaptive cache prefetch policies

This paper begins an exploration of the applicability of traditional prefetching policies in multiprocessor architectures. In particular, the effectiveness of prefetching policies as a function of both the quality of the prefetching and the consumption of processor to memory bandwidth is an issue of...

Full description

Saved in:
Bibliographic Details
Published in:Conference Proceedings of the 2000 IEEE International Performance, Computing, and Communications Conference (Cat. No.00CH37086) Computing, and Communications Conference (Cat. No.00CH37086), 2000, p.509-515
Main Authors: Oliver, R.L., Teller, P.J.
Format: Article
Language:English
Subjects:
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
cited_by
cites
container_end_page 515
container_issue
container_start_page 509
container_title Conference Proceedings of the 2000 IEEE International Performance, Computing, and Communications Conference (Cat. No.00CH37086)
container_volume
creator Oliver, R.L.
Teller, P.J.
description This paper begins an exploration of the applicability of traditional prefetching policies in multiprocessor architectures. In particular, the effectiveness of prefetching policies as a function of both the quality of the prefetching and the consumption of processor to memory bandwidth is an issue of interest. Addressing this issue, the concept of a dynamic and adaptive cache (DAC), two new prefetch policies, and the design of an instruction DAC, called the DAC/sup 3/, which dynamically changes its prefetch policy at runtime, in response to process execution characteristics, are introduced. In addition, a detailed performance analysis of the DAC/sup 3/ and two new prefetch policies, which the DAC/sup 3/ uses, are presented; the performance of the DAC/sup 3/ is compared to that of the SSB prefetch instruction cache, which is based on Jouppi's sequential stream buffer design. This performance analysis is based on a new metric called CompositeCPI, which captures the usefulness of prefetches and their cost in terms of consumed memory bandwidth. The performance analysis indicates that, for the cache configurations and multiprogram workloads studied, the DAC/sup 3/ is superior to the SSB instruction prefetch cache.
doi_str_mv 10.1109/PCCC.2000.830357
format article
fullrecord <record><control><sourceid>proquest_6IE</sourceid><recordid>TN_cdi_ieee_primary_830357</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>830357</ieee_id><sourcerecordid>27514798</sourcerecordid><originalsourceid>FETCH-LOGICAL-i118t-d3689eebb8d02d31d1c0bf9985597862c653b6d8074bf9e217580e7aee0ef8ca3</originalsourceid><addsrcrecordid>eNotkM1LxDAUxAMquK57F089ebLdl6RpkqPUVRcW9KDnkiavbKRfNl1h_3sD9TTD8GN4bwi5o5BRCnr7UZZlxgAgUxy4kBfkBqSKTkutLskqMjJlRU6vySaE7wiCAAk5X5HH53NvOm8T07vEODPO_hcTa-wRk3HCBmd7TMah9dZjuCVXjWkDbv51Tb5edp_lW3p4f92XT4fUU6rm1PFCacS6Vg6Y49RRC3WjtRLxIFUwWwheF06BzGOMjEqhAKVBBGyUNXxNHpbecRp-ThjmqvPBYtuaHodTqJgUNI-vRfB-AT0iVuPkOzOdq2UE_gd57E7-</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>27514798</pqid></control><display><type>article</type><title>Dynamic and adaptive cache prefetch policies</title><source>IEEE Electronic Library (IEL) Conference Proceedings</source><creator>Oliver, R.L. ; Teller, P.J.</creator><creatorcontrib>Oliver, R.L. ; Teller, P.J.</creatorcontrib><description>This paper begins an exploration of the applicability of traditional prefetching policies in multiprocessor architectures. In particular, the effectiveness of prefetching policies as a function of both the quality of the prefetching and the consumption of processor to memory bandwidth is an issue of interest. Addressing this issue, the concept of a dynamic and adaptive cache (DAC), two new prefetch policies, and the design of an instruction DAC, called the DAC/sup 3/, which dynamically changes its prefetch policy at runtime, in response to process execution characteristics, are introduced. In addition, a detailed performance analysis of the DAC/sup 3/ and two new prefetch policies, which the DAC/sup 3/ uses, are presented; the performance of the DAC/sup 3/ is compared to that of the SSB prefetch instruction cache, which is based on Jouppi's sequential stream buffer design. This performance analysis is based on a new metric called CompositeCPI, which captures the usefulness of prefetches and their cost in terms of consumed memory bandwidth. The performance analysis indicates that, for the cache configurations and multiprogram workloads studied, the DAC/sup 3/ is superior to the SSB instruction prefetch cache.</description><identifier>ISSN: 1097-2641</identifier><identifier>ISBN: 0780359798</identifier><identifier>ISBN: 9780780359796</identifier><identifier>DOI: 10.1109/PCCC.2000.830357</identifier><language>eng</language><publisher>IEEE</publisher><subject>Amplitude modulation ; Bandwidth ; Computer science ; Costs ; Design optimization ; Multiprocessing systems ; Performance analysis ; Prefetching ; Runtime ; Throughput</subject><ispartof>Conference Proceedings of the 2000 IEEE International Performance, Computing, and Communications Conference (Cat. No.00CH37086), 2000, p.509-515</ispartof><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/830357$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>309,310,314,780,784,789,790,2058,4050,4051,27924,27925,54920</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/830357$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc></links><search><creatorcontrib>Oliver, R.L.</creatorcontrib><creatorcontrib>Teller, P.J.</creatorcontrib><title>Dynamic and adaptive cache prefetch policies</title><title>Conference Proceedings of the 2000 IEEE International Performance, Computing, and Communications Conference (Cat. No.00CH37086)</title><addtitle>PCCC</addtitle><description>This paper begins an exploration of the applicability of traditional prefetching policies in multiprocessor architectures. In particular, the effectiveness of prefetching policies as a function of both the quality of the prefetching and the consumption of processor to memory bandwidth is an issue of interest. Addressing this issue, the concept of a dynamic and adaptive cache (DAC), two new prefetch policies, and the design of an instruction DAC, called the DAC/sup 3/, which dynamically changes its prefetch policy at runtime, in response to process execution characteristics, are introduced. In addition, a detailed performance analysis of the DAC/sup 3/ and two new prefetch policies, which the DAC/sup 3/ uses, are presented; the performance of the DAC/sup 3/ is compared to that of the SSB prefetch instruction cache, which is based on Jouppi's sequential stream buffer design. This performance analysis is based on a new metric called CompositeCPI, which captures the usefulness of prefetches and their cost in terms of consumed memory bandwidth. The performance analysis indicates that, for the cache configurations and multiprogram workloads studied, the DAC/sup 3/ is superior to the SSB instruction prefetch cache.</description><subject>Amplitude modulation</subject><subject>Bandwidth</subject><subject>Computer science</subject><subject>Costs</subject><subject>Design optimization</subject><subject>Multiprocessing systems</subject><subject>Performance analysis</subject><subject>Prefetching</subject><subject>Runtime</subject><subject>Throughput</subject><issn>1097-2641</issn><isbn>0780359798</isbn><isbn>9780780359796</isbn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2000</creationdate><recordtype>article</recordtype><sourceid>6IE</sourceid><recordid>eNotkM1LxDAUxAMquK57F089ebLdl6RpkqPUVRcW9KDnkiavbKRfNl1h_3sD9TTD8GN4bwi5o5BRCnr7UZZlxgAgUxy4kBfkBqSKTkutLskqMjJlRU6vySaE7wiCAAk5X5HH53NvOm8T07vEODPO_hcTa-wRk3HCBmd7TMah9dZjuCVXjWkDbv51Tb5edp_lW3p4f92XT4fUU6rm1PFCacS6Vg6Y49RRC3WjtRLxIFUwWwheF06BzGOMjEqhAKVBBGyUNXxNHpbecRp-ThjmqvPBYtuaHodTqJgUNI-vRfB-AT0iVuPkOzOdq2UE_gd57E7-</recordid><startdate>2000</startdate><enddate>2000</enddate><creator>Oliver, R.L.</creator><creator>Teller, P.J.</creator><general>IEEE</general><scope>6IE</scope><scope>6IH</scope><scope>CBEJK</scope><scope>RIE</scope><scope>RIO</scope><scope>7SC</scope><scope>8FD</scope><scope>JQ2</scope><scope>L7M</scope><scope>L~C</scope><scope>L~D</scope></search><sort><creationdate>2000</creationdate><title>Dynamic and adaptive cache prefetch policies</title><author>Oliver, R.L. ; Teller, P.J.</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-i118t-d3689eebb8d02d31d1c0bf9985597862c653b6d8074bf9e217580e7aee0ef8ca3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2000</creationdate><topic>Amplitude modulation</topic><topic>Bandwidth</topic><topic>Computer science</topic><topic>Costs</topic><topic>Design optimization</topic><topic>Multiprocessing systems</topic><topic>Performance analysis</topic><topic>Prefetching</topic><topic>Runtime</topic><topic>Throughput</topic><toplevel>online_resources</toplevel><creatorcontrib>Oliver, R.L.</creatorcontrib><creatorcontrib>Teller, P.J.</creatorcontrib><collection>IEEE Electronic Library (IEL) Conference Proceedings</collection><collection>IEEE Proceedings Order Plan (POP) 1998-present by volume</collection><collection>IEEE Xplore All Conference Proceedings</collection><collection>IEEE Xplore (Online service)</collection><collection>IEEE Proceedings Order Plans (POP) 1998-present</collection><collection>Computer and Information Systems Abstracts</collection><collection>Technology Research Database</collection><collection>ProQuest Computer Science Collection</collection><collection>Advanced Technologies Database with Aerospace</collection><collection>Computer and Information Systems Abstracts – Academic</collection><collection>Computer and Information Systems Abstracts Professional</collection><jtitle>Conference Proceedings of the 2000 IEEE International Performance, Computing, and Communications Conference (Cat. No.00CH37086)</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Oliver, R.L.</au><au>Teller, P.J.</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Dynamic and adaptive cache prefetch policies</atitle><jtitle>Conference Proceedings of the 2000 IEEE International Performance, Computing, and Communications Conference (Cat. No.00CH37086)</jtitle><stitle>PCCC</stitle><date>2000</date><risdate>2000</risdate><spage>509</spage><epage>515</epage><pages>509-515</pages><issn>1097-2641</issn><isbn>0780359798</isbn><isbn>9780780359796</isbn><abstract>This paper begins an exploration of the applicability of traditional prefetching policies in multiprocessor architectures. In particular, the effectiveness of prefetching policies as a function of both the quality of the prefetching and the consumption of processor to memory bandwidth is an issue of interest. Addressing this issue, the concept of a dynamic and adaptive cache (DAC), two new prefetch policies, and the design of an instruction DAC, called the DAC/sup 3/, which dynamically changes its prefetch policy at runtime, in response to process execution characteristics, are introduced. In addition, a detailed performance analysis of the DAC/sup 3/ and two new prefetch policies, which the DAC/sup 3/ uses, are presented; the performance of the DAC/sup 3/ is compared to that of the SSB prefetch instruction cache, which is based on Jouppi's sequential stream buffer design. This performance analysis is based on a new metric called CompositeCPI, which captures the usefulness of prefetches and their cost in terms of consumed memory bandwidth. The performance analysis indicates that, for the cache configurations and multiprogram workloads studied, the DAC/sup 3/ is superior to the SSB instruction prefetch cache.</abstract><pub>IEEE</pub><doi>10.1109/PCCC.2000.830357</doi><tpages>7</tpages></addata></record>
fulltext fulltext_linktorsrc
identifier ISSN: 1097-2641
ispartof Conference Proceedings of the 2000 IEEE International Performance, Computing, and Communications Conference (Cat. No.00CH37086), 2000, p.509-515
issn 1097-2641
language eng
recordid cdi_ieee_primary_830357
source IEEE Electronic Library (IEL) Conference Proceedings
subjects Amplitude modulation
Bandwidth
Computer science
Costs
Design optimization
Multiprocessing systems
Performance analysis
Prefetching
Runtime
Throughput
title Dynamic and adaptive cache prefetch policies
url http://sfxeu10.hosted.exlibrisgroup.com/loughborough?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-30T04%3A57%3A37IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_6IE&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Dynamic%20and%20adaptive%20cache%20prefetch%20policies&rft.jtitle=Conference%20Proceedings%20of%20the%202000%20IEEE%20International%20Performance,%20Computing,%20and%20Communications%20Conference%20(Cat.%20No.00CH37086)&rft.au=Oliver,%20R.L.&rft.date=2000&rft.spage=509&rft.epage=515&rft.pages=509-515&rft.issn=1097-2641&rft.isbn=0780359798&rft.isbn_list=9780780359796&rft_id=info:doi/10.1109/PCCC.2000.830357&rft_dat=%3Cproquest_6IE%3E27514798%3C/proquest_6IE%3E%3Cgrp_id%3Ecdi_FETCH-LOGICAL-i118t-d3689eebb8d02d31d1c0bf9985597862c653b6d8074bf9e217580e7aee0ef8ca3%3C/grp_id%3E%3Coa%3E%3C/oa%3E%3Curl%3E%3C/url%3E&rft_id=info:oai/&rft_pqid=27514798&rft_id=info:pmid/&rft_ieee_id=830357&rfr_iscdi=true