BLOwing Trees to the Ground: Layout Optimization of Decision Trees on Racetrack Memory

Modern distributed low power systems tend to integrate machine learning algorithms, which are directly executed on the distributed devices (on the edge). In resource constrained setups (e.g. battery driven sensor nodes), the execution of the machine learning models has to be optimized for execution...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:2021 58th ACM/IEEE Design Automation Conference (DAC) s. 1111 - 1116
Hlavní autoři: Hakert, Christian, Khan, Asif Ali, Chen, Kuan-Hsun, Hameed, Fazal, Castrillon, Jeronimo, Chen, Jian-Jia
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 05.12.2021
Témata:
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Abstract Modern distributed low power systems tend to integrate machine learning algorithms, which are directly executed on the distributed devices (on the edge). In resource constrained setups (e.g. battery driven sensor nodes), the execution of the machine learning models has to be optimized for execution time and energy consumption. Racetrack memory (RTM), an emerging non-volatile memory (NVM), promises to achieve these goals by offering unprecedented integration density, smaller access-latency and reduced energy consumption. However, in order to access data in RTM, it needs to be shifted to the access port first, resulting in latency and energy penalties. In this paper, we propose B.L.O. (Bidirectional Linear Ordering), a novel domain-specific approach for placing decision trees in RTMs. We reduce the total amount of shifts during inference by exploiting the tree structure and estimated access probabilities. We further apply the state-of-the-art methods to place data structures in RTM, without exploiting any domain-specific knowledge, to the decision trees and compare them to B. L.O. We formally prove that the B.L.O. solution has an approximation ratio of 4, i.e., its number of shifts is guaranteed to be at most 4 times the optimal number of shifts for a given decision tree. Throughout the experimental evaluation, we show that for the realistic use case B.L.O. empirically outperforms the state-of-the-art data placement method on average by 54.7% in terms of shifts, 19.2% in terms of runtime and 19.2% in terms of energy consumption.
AbstractList Modern distributed low power systems tend to integrate machine learning algorithms, which are directly executed on the distributed devices (on the edge). In resource constrained setups (e.g. battery driven sensor nodes), the execution of the machine learning models has to be optimized for execution time and energy consumption. Racetrack memory (RTM), an emerging non-volatile memory (NVM), promises to achieve these goals by offering unprecedented integration density, smaller access-latency and reduced energy consumption. However, in order to access data in RTM, it needs to be shifted to the access port first, resulting in latency and energy penalties. In this paper, we propose B.L.O. (Bidirectional Linear Ordering), a novel domain-specific approach for placing decision trees in RTMs. We reduce the total amount of shifts during inference by exploiting the tree structure and estimated access probabilities. We further apply the state-of-the-art methods to place data structures in RTM, without exploiting any domain-specific knowledge, to the decision trees and compare them to B. L.O. We formally prove that the B.L.O. solution has an approximation ratio of 4, i.e., its number of shifts is guaranteed to be at most 4 times the optimal number of shifts for a given decision tree. Throughout the experimental evaluation, we show that for the realistic use case B.L.O. empirically outperforms the state-of-the-art data placement method on average by 54.7% in terms of shifts, 19.2% in terms of runtime and 19.2% in terms of energy consumption.
Author Hameed, Fazal
Khan, Asif Ali
Castrillon, Jeronimo
Chen, Kuan-Hsun
Chen, Jian-Jia
Hakert, Christian
Author_xml – sequence: 1
  givenname: Christian
  surname: Hakert
  fullname: Hakert, Christian
  email: christian.hakert@tu-dortmund.de
  organization: Technische Universität Dortmund
– sequence: 2
  givenname: Asif Ali
  surname: Khan
  fullname: Khan, Asif Ali
  email: asif_ali.khan@tu-dresden.de
  organization: Technische Universität Dresden
– sequence: 3
  givenname: Kuan-Hsun
  surname: Chen
  fullname: Chen, Kuan-Hsun
  email: kuan-hsun.chen@tu-dortmund.de
  organization: Technische Universität Dortmund
– sequence: 4
  givenname: Fazal
  surname: Hameed
  fullname: Hameed, Fazal
  email: fazal.hameed@tu-dresden.de
  organization: Technische Universität Dresden
– sequence: 5
  givenname: Jeronimo
  surname: Castrillon
  fullname: Castrillon, Jeronimo
  email: jeronimo.castrillon@tu-dresden.de
  organization: Technische Universität Dresden
– sequence: 6
  givenname: Jian-Jia
  surname: Chen
  fullname: Chen, Jian-Jia
  email: jian-jia.chen@tu-dortmund.de
  organization: Technische Universität Dortmund
BookMark eNotj99KwzAchSMoqLNPIEJeoDNJ89e72ekUKgWZ3o40-VWDthlthtSnd7LdnPPdfAfOJTrtYw8I3VAyp5SY2-WipJooPmeE0bkRWlKpTlBmlKZSCl4wxck5ysYxNEQSofk-L9D7fVX_hP4DrweAEaeI0yfg1RB3vb_DlZ3iLuF6m0IXfm0KscexxUtwYfzng7SHV-sgDdZ94Rfo4jBdobPWfo-QHXuG3h4f1uVTXtWr53JR5ZZplfLWes0lcCgUNJ5KxlthnW4I5VY4brU23jHulDOFJ9RrLxiVvKDSEGm1K2bo-rAbAGCzHUJnh2lzfF_8AZ_iUkM
ContentType Conference Proceeding
DBID 6IE
6IH
CBEJK
RIE
RIO
DOI 10.1109/DAC18074.2021.9586167
DatabaseName IEEE Electronic Library (IEL) Conference Proceedings
IEEE Proceedings Order Plan (POP) 1998-present by volume
IEEE Xplore All Conference Proceedings
IEEE Electronic Library (IEL)
IEEE Proceedings Order Plans (POP) 1998-present
DatabaseTitleList
Database_xml – sequence: 1
  dbid: RIE
  name: IEEE Electronic Library (IEL)
  url: https://ieeexplore.ieee.org/
  sourceTypes: Publisher
DeliveryMethod fulltext_linktorsrc
EISBN 9781665432740
1665432748
EndPage 1116
ExternalDocumentID 9586167
Genre orig-research
GroupedDBID 6IE
6IH
ACM
ALMA_UNASSIGNED_HOLDINGS
CBEJK
RIE
RIO
ID FETCH-LOGICAL-a287t-fad846e4e37ebd1624f5ac8b014a5c4a889dc24c7c93d01d8d52164316906a8c3
IEDL.DBID RIE
ISICitedReferencesCount 3
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000766079700186&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
IngestDate Wed Aug 27 02:28:29 EDT 2025
IsPeerReviewed false
IsScholarly true
Language English
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-a287t-fad846e4e37ebd1624f5ac8b014a5c4a889dc24c7c93d01d8d52164316906a8c3
PageCount 6
ParticipantIDs ieee_primary_9586167
PublicationCentury 2000
PublicationDate 2021-Dec.-5
PublicationDateYYYYMMDD 2021-12-05
PublicationDate_xml – month: 12
  year: 2021
  text: 2021-Dec.-5
  day: 05
PublicationDecade 2020
PublicationTitle 2021 58th ACM/IEEE Design Automation Conference (DAC)
PublicationTitleAbbrev DAC
PublicationYear 2021
Publisher IEEE
Publisher_xml – name: IEEE
SSID ssib060584060
Score 2.1872492
Snippet Modern distributed low power systems tend to integrate machine learning algorithms, which are directly executed on the distributed devices (on the edge). In...
SourceID ieee
SourceType Publisher
StartPage 1111
SubjectTerms Energy consumption
Machine learning
Machine learning algorithms
Memory management
Nonvolatile memory
Power systems
Runtime
Title BLOwing Trees to the Ground: Layout Optimization of Decision Trees on Racetrack Memory
URI https://ieeexplore.ieee.org/document/9586167
WOSCitedRecordID wos000766079700186&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1NSwMxEA21ePCk0orf5ODRbTe72Xx409biobZFaumtZCcJiLgr263Sf2-yXSuCF29JIAQmIW_eZF4GoSsQEU18WgXRlgaUKhKkktmAWytCmYYmqbQwsyEfjcR8LicNdL3VwhhjquQz0_HN6i1f57DyobKuTAQjjO-gHc7ZRqv1fXb8657DprAW6ZBQdvu3PeK_enEkMCKdeu6vIioVhgz2_7f6AWr_iPHwZAszh6hhshaa3Q3Hn66Lp4UxS1zm2Dly2AeSMn2Dh2qdr0o8dtfBW62zxLnF_bqeTj3JNZ4UmLJQ8Ioffcbtuo2eB_fT3kNQl0gIlKM6ZWCVdg6EoSbmJtWERdQmCkTqiI9KgCohpIaIAgcZ65BooR1cMy9_lyFTAuIj1MzyzBwjDGB4lID1eEU9c2WcEaupYrHbOIhPUMvbZPG--QVjUZvj9O_hM7TnzV4lfiTnqFkWK3OBduGjfFkWl9XWfQFpfZnj
linkProvider IEEE
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1LSwMxEA61CnpSacW3OXh0281ukk28qbVU3D6QWnor2TxAxF3ZbpX-e5PtWhG8eEsCQyAT8s1M5psB4FKyABOXVoGUwR7GAnkJp8aLjGE-T3xNSi7MJI4GAzad8lENXK25MFrrMvlMt9yw_MtXmVy4UFmbE0YRjTbAJsE48Fdsre_b4_73LDr5FU0H-bzdublDrtiLdQMD1Kqkf7VRKVGku_u__fdA84eOB0droNkHNZ02wOQ2Hn7aKRznWs9hkUFrykEXSkrVNYzFMlsUcGgfhLeKaQkzAztVR51KyA6ehNRFLuQr7Luc22UTPHfvx3c9r2qS4Anr7BSeEcqaEBrrMNKJQjTAhgjJEuv6CCKxYIwrGWAZSR4qHymmLGBTR4DnPhVMhgegnmapPgRQSh0FRBqHWNj5rjSiyCgsaGhVJ8Mj0HBnMntf1cGYVcdx_PfyBdjujfvxLH4YPJ6AHaeCMg2EnIJ6kS_0GdiSH8XLPD8v1fgFmzCdKg
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=2021+58th+ACM%2FIEEE+Design+Automation+Conference+%28DAC%29&rft.atitle=BLOwing+Trees+to+the+Ground%3A+Layout+Optimization+of+Decision+Trees+on+Racetrack+Memory&rft.au=Hakert%2C+Christian&rft.au=Khan%2C+Asif+Ali&rft.au=Chen%2C+Kuan-Hsun&rft.au=Hameed%2C+Fazal&rft.date=2021-12-05&rft.pub=IEEE&rft.spage=1111&rft.epage=1116&rft_id=info:doi/10.1109%2FDAC18074.2021.9586167&rft.externalDocID=9586167