Runahead Threads to improve SMT performance

In this paper, we propose runahead threads (RaT) as a valuable solution for both reducing resource contention and exploiting memory-level parallelism in simultaneous multithreaded (SMT) processors. Our technique converts a resource intensive memory-bound thread to a speculative light thread under lo...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:2008 IEEE 14th International Symposium on High Performance Computer Architecture s. 149 - 158
Hlavní autori: Ramirez, T., Pajuelo, A., Santana, O.J., Valero, M.
Médium: Konferenčný príspevok.. Publikácia
Jazyk:English
Vydavateľské údaje: IEEE 01.02.2008
Institute of Electrical and Electronics Engineers (IEEE)
Predmet:
Arquitectura de computadors
Disseny
Informàtica
Multi-threading
Parallel processing
Parallel processing (Electronic computers)
Prefetching
Processament en paral·lel (Ordinadors)
Program processors
Programari
Proposals
Registers
Resource allocation
Resource management
Simultaneous multithreading processors
Software architecture
Storage management
Throughput
Àrees temàtiques de la UPC
ISBN:1424420709, 9781424420704
ISSN:1530-0897
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Abstract In this paper, we propose runahead threads (RaT) as a valuable solution for both reducing resource contention and exploiting memory-level parallelism in simultaneous multithreaded (SMT) processors. Our technique converts a resource intensive memory-bound thread to a speculative light thread under long-latency blocking memory operations. These speculative threads prefetch data and instructions with minimal resources, reducing critical resource conflicts between threads. We compare an SMT architecture using RaT to both state-of-the-art static fetch policies and dynamic resource control policies. In terms of throughput and fairness, our results show that RaT performs better than any other policy. The proposed mechanism improves average throughput by 37% regarding previous static fetch policies and by 28% compared to previous dynamic resource scheduling mechanisms. RaT also improves fairness by 36% and 30% respectively. In addition, the proposed mechanism permits register file size reduction of up to 60% in a SMT processor without performance degradation.
AbstractList In this paper, we propose runahead threads (RaT) as a valuable solution for both reducing resource contention and exploiting memory-level parallelism in simultaneous multithreaded (SMT) processors. Our technique converts a resource intensive memory-bound thread to a speculative light thread under long-latency blocking memory operations. These speculative threads prefetch data and instructions with minimal resources, reducing critical resource conflicts between threads. We compare an SMT architecture using RaT to both state-of-the-art static fetch policies and dynamic resource control policies. In terms of throughput and fairness, our results show that RaT performs better than any other policy. The proposed mechanism improves average throughput by 37% regarding previous static fetch policies and by 28% compared to previous dynamic resource scheduling mechanisms. RaT also improves fairness by 36% and 30% respectively. In addition, the proposed mechanism permits register file size reduction of up to 60% in a SMT processor without performance degradation.
In this paper, we propose Runahead Threads (RaT) as a valuable solution for both reducing resource contention and exploiting memory-level parallelism in Simultaneous Multithreaded (SMT) processors. Our technique converts a resource intensive memory-bound thread to a speculative light thread under long-latency blocking memory operations. These speculative threads prefetch data and instructions with minimal resources, reducing critical resource conflicts between threads. We compare an SMT architecture using RaT to both state-of-the-art static fetch policies and dynamic resource control policies. In terms of throughput and fairness, our results show that RaT performs better than any other policy. The proposed mechanism improves average throughput by 37% regarding previous static fetch policies and by 28% compared to previous dynamic resource scheduling mechanisms. RaT also improves fairness by 36% and 30% respectively. In addition, the proposed mechanism permits register file size reduction of up to 60% in a SMT processor without performance degradation. Peer Reviewed
Author Santana, O.J.
Pajuelo, A.
Valero, M.
Ramirez, T.
Author_xml – sequence: 1
  givenname: T.
  surname: Ramirez
  fullname: Ramirez, T.
  organization: Univ. Politelecnica de Catalunya, Barcelona
– sequence: 2
  givenname: A.
  surname: Pajuelo
  fullname: Pajuelo, A.
  organization: Univ. Politelecnica de Catalunya, Barcelona
– sequence: 3
  givenname: O.J.
  surname: Santana
  fullname: Santana, O.J.
  organization: Univ. de Las Palmas de Gran Canaria, Las Palmas de Gran Canaria
– sequence: 4
  givenname: M.
  surname: Valero
  fullname: Valero, M.
  organization: Univ. Politelecnica de Catalunya, Barcelona
– sequence: 5
  givenname: M.
  surname: Valero
  fullname: Valero, M.
  organization: Barcelona Supercomput. Center, Barcelona
BookMark eNpFUF1Lw0AQPLAF29ofIL7kXRL3PvYu91iCWqGiaHwO1-sdjZgPLq3gv--FFlxYhoGZZWbnZNJ2rSPklkJGKeiH9XuxyhhAngmJueR4ReZUMCEYKNATMqPIIYVcqymZjzoNwBGvyXIYviGOQC4Fzsj9x7E1e2d2SbkPEYbk0CV104fu1yWfr2XSu-C70JjWuhsy9eZncMsLLsjX02NZrNPN2_NLsdqklgmKqZDUCi2VwZwztfNotZTbSD1jMavXWlHjVW7oFtnYRmunqZIxE3LOgS8IPd-1w9FWwVkXrDlUnan_ybixK6uYFhxGz93ZUzvnqj7UjQl_1eU5_ATLr1P8
ContentType Conference Proceeding
Publication
Contributor Universitat Politècnica de Catalunya. CAP - Grup de Computació d'Altes Prestacions
Universitat Politècnica de Catalunya. Departament d'Arquitectura de Computadors
Contributor_xml – sequence: 1
  fullname: Universitat Politècnica de Catalunya. Departament d'Arquitectura de Computadors
– sequence: 2
  fullname: Universitat Politècnica de Catalunya. CAP - Grup de Computació d'Altes Prestacions
Copyright info:eu-repo/semantics/openAccess
Copyright_xml – notice: info:eu-repo/semantics/openAccess
DBID 6IE
6IL
CBEJK
RIE
RIL
XX2
DOI 10.1109/HPCA.2008.4658635
DatabaseName IEEE Electronic Library (IEL) Conference Proceedings
IEEE Xplore POP ALL
IEEE Xplore All Conference Proceedings
IEEE Electronic Library (IEL)
IEEE Proceedings Order Plans (POP All) 1998-Present
Recercat
DatabaseTitleList

Database_xml – sequence: 1
  dbid: RIE
  name: IEEE Electronic Library (IEL)
  url: https://ieeexplore.ieee.org/
  sourceTypes: Publisher
DeliveryMethod fulltext_linktorsrc
Discipline Computer Science
EndPage 158
ExternalDocumentID oai_recercat_cat_2072_294300
4658635
Genre orig-research
GroupedDBID 29O
6IE
6IF
6IH
6IK
6IL
6IM
6IN
AAJGR
AAWTH
ABLEC
ADZIZ
ALMA_UNASSIGNED_HOLDINGS
BEFXN
BFFAM
BGNUA
BKEBE
BPEOZ
CBEJK
CHZPO
IEGSK
IPLJI
M43
OCL
RIE
RIL
RNS
AARBI
IERZE
RIB
RIC
XX2
ID FETCH-LOGICAL-c2415-461c4967a58327df5c966b7a5f22586f9971af78a1b52110999e9176045533303
IEDL.DBID RIE
ISBN 1424420709
9781424420704
ISSN 1530-0897
IngestDate Fri Nov 07 13:45:06 EST 2025
Wed Aug 27 02:02:49 EDT 2025
IsDoiOpenAccess true
IsOpenAccess true
IsPeerReviewed false
IsScholarly true
LCCN 2008900355
Language English
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-c2415-461c4967a58327df5c966b7a5f22586f9971af78a1b52110999e9176045533303
OpenAccessLink https://recercat.cat/handle/2072/294300
PageCount 10
ParticipantIDs ieee_primary_4658635
csuc_recercat_oai_recercat_cat_2072_294300
PublicationCentury 2000
PublicationDate 2008-02
PublicationDateYYYYMMDD 2008-02-01
PublicationDate_xml – month: 02
  year: 2008
  text: 2008-02
PublicationDecade 2000
PublicationTitle 2008 IEEE 14th International Symposium on High Performance Computer Architecture
PublicationTitleAbbrev HPCA
PublicationYear 2008
Publisher IEEE
Institute of Electrical and Electronics Engineers (IEEE)
Publisher_xml – name: IEEE
– name: Institute of Electrical and Electronics Engineers (IEEE)
SSID ssj0000453645
ssj0002951
Score 1.9124254
Snippet In this paper, we propose runahead threads (RaT) as a valuable solution for both reducing resource contention and exploiting memory-level parallelism in...
In this paper, we propose Runahead Threads (RaT) as a valuable solution for both reducing resource contention and exploiting memory-level parallelism in...
SourceID csuc
ieee
SourceType Open Access Repository
Publisher
StartPage 149
SubjectTerms Arquitectura de computadors
Disseny
Informàtica
Multi-threading
Parallel processing
Parallel processing (Electronic computers)
Prefetching
Processament en paral·lel (Ordinadors)
Program processors
Programari
Proposals
Registers
Resource allocation
Resource management
Simultaneous multithreading processors
Software architecture
Storage management
Throughput
Àrees temàtiques de la UPC
Title Runahead Threads to improve SMT performance
URI https://ieeexplore.ieee.org/document/4658635
https://recercat.cat/handle/2072/294300
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV07a8MwEBZJ6NCpj6T0jYZObd3YjmTZYwkNWRpCm0I2c3pBFickdn9_T7biUOjSwWAZg81Jd9-9j5AHFVoEVkAjRwILGJdpAEkEQZTYNLFg9EiyetiEmM3S5TKbd8hzWwtjjKmTz8yLu61j-XqtKucqGzKESwTILukKIZpardafgqpJE1HzUjjOuO-V6vK1XPu_pqgrxjOe7Xs9-TXz4c4ozIbT-fi1SbH0X0NZrXaV-jV9pQafycn_fvuUDA5VfHTe4tMZ6ZjinJzsxzhQz9V98vRRFYAyWdMFbizoHS3XdFU7Gwz9fF_QzaG4YEC-Jm-L8TTwMxQC5bA5YEmkWJYI4Mi6Qluu0L6RuLTIyLgbWSYisCKFSHJnC6K-aNCCS5CcqAgivl2QXrEuzCWhMrYcgIXappzJkZao3ZgRNwykEQLCK_LoCJWjpDZbBWXuOle3C3chreM8dg3f8eW-I1S-aXpq5J5G138_viHHTZ6GSyO5Jb1yW5k7cqS-y9Vue18fhB9DxalF
linkProvider IEEE
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3NS8MwFH_MKehp6iZ-m4Mnta7tkn4cZTgmbmNohd1Kkiawyza21r_flzbrELx4KDSl0PKS937v-wHcS1cjsHI0cgSnDmUicnjgcccLdBRorrKeoOWwiXAyiWazeNqAp7oWRilVJp-pZ3NbxvKzpSyMq6xLES4RIPdgn1Hqe1W1Vu1RQeWkiqlZOezHzHZLNRlbpgFgVdbl4ymPt92e7JragKfnxt3htP9SJVna76G0lptC_pq_UsLPoPW_Hz-Gzq6Oj0xrhDqBhlqcQms7yIFYvm7D40ex4CiVM5Lg1vJsQ_IlmZfuBkU-xwlZ7coLOvA1eE36Q8dOUXCkQWeHBp6kcRByhswbZppJtHAELjWyMu5HHIce12HEPcGMNYgao0IbLkByoiqICHcGzcVyoc6BCF8zzqmb6YhR0csE6jeqxxTlQoUhdy_gwRAqRVmt1pLnqeldXS_MhbT2U9-0fMeX24ZQ6arqqpFaGl3-_fgODofJeJSO3ibvV3BUZW2YpJJraObrQt3AgfzO55v1bXkofgBLs6yM
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=proceeding&rft.title=2008+IEEE+14th+International+Symposium+on+High+Performance+Computer+Architecture&rft.atitle=Runahead+Threads+to+improve+SMT+performance&rft.au=Ramirez%2C+T.&rft.au=Pajuelo%2C+A.&rft.au=Santana%2C+O.J.&rft.au=Valero%2C+M.&rft.date=2008-02-01&rft.pub=IEEE&rft.isbn=9781424420704&rft.issn=1530-0897&rft.spage=149&rft.epage=158&rft_id=info:doi/10.1109%2FHPCA.2008.4658635&rft.externalDocID=4658635
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1530-0897&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1530-0897&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1530-0897&client=summon