Academic Journal
Προβολή σε EDS

VMOC: MLIR-based memory optimization for CNN deployment on vector accelerators.

Λεπτομέρειες βιβλιογραφικής εγγραφής
Τίτλος: VMOC: MLIR-based memory optimization for CNN deployment on vector accelerators.
Συγγραφείς: Chen, Xiaorong1 (AUTHOR) chenxiaorong23@nudt.edu.cn, Liu, Zhong1 (AUTHOR) zhongliu@nudt.edu.cn
Πηγή: Journal of Systems Architecture. Oct2026, Vol. 179, pN.PAG-N.PAG. 1p.
Θεματικοί όροι: Computer memory management, Code generators
Περίληψη: Vector accelerators are increasingly prominent platforms for deploying Convolutional Neural Networks (CNNs) due to their high energy efficiency and parallelism. On such architectures, efficient memory utilization is critical for successful deployment. However, existing graph fission methods exhibit several limitations on vector accelerators, including poor hardware adaptability, inaccurate fission cost modeling, low compilation efficiency, and heavy reliance on manual intervention. This paper proposes VMOC, a novel MLIR-based memory optimization method. VMOC automatically identifies memory hotspot subgraphs, constructs a Hardware-aware Fission Cost Model (HFCM) to quantify the memory benefits and execution overhead introduced by fission, and employs an Adaptive Graph Fission Algorithm (AGFA) to dynamically search for fission schemes that satisfy both memory constraints and positive benefit constraints. Meanwhile, a Graph Encapsulation-based Scheduling Optimization (GESO) is introduced to alleviate the problem of increased scheduling complexity caused by graph fission. These optimization strategies are integrated into the MLIR framework, enabling end-to-end automatic compilation from ONNX models to target executable code and generating high-performance code for vector accelerators. Experimental results demonstrate the effectiveness of VMOC on GoogleNet, ResNet18, VGG16, and YOLOv5s. VMOC achieves memory compression ratios of 52.05% ∼ 96.18%, with compilation speedups of 2.34 × ∼ 152.11 × over MAGIS and 12.63 × ∼ 100.93 × over TVM. On FT-M7032 and FT-XT4000 vector accelerators, it achieves computational efficiencies of 62.30% ∼ 74.05% and 50.25% ∼ 71.01%, consistently outperforming VCNN. These results suggest that VMOC has the potential to become an efficient and practical approach for the automatic deployment of CNNs on vector accelerators. [Display omitted] • A memory-hotspot method identifies critical operators affecting peak memory. • The HFCM model evaluates the benefits of fissioning under GSM capacity constraints. • The AGFA dynamically finds the minimum fission granularity under GSM constraints. • The GESO encapsulates local subgraphs into supernodes to suppress graph expansion. • An end-to-end MLIR framework efficiently deploys CNN models on vector accelerators. [ABSTRACT FROM AUTHOR]
Copyright of Journal of Systems Architecture is the property of Elsevier B.V. and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Βάση Δεδομένων: Business Source Index
FullText Text:
  Availability: 0
CustomLinks:
  – Url: https://www.doi.org/10.1016/j.sysarc.2026.103933?
    Name: ScienceDirect (all content) (s7799221)
    Category: fullText
    Text: View record from ScienceDirect
    MouseOverText: View record from ScienceDirect
  – Url: https://resolver.ebsco.com/c/fiv2js/result?sid=EBSCO:bsx&genre=article&issn=13837621&ISBN=&volume=179&issue=&date=20261001&spage=N.PAG&pages=&title=Journal of Systems Architecture&atitle=VMOC%3A%20MLIR-based%20memory%20optimization%20for%20CNN%20deployment%20on%20vector%20accelerators.&aulast=Chen%2C%20Xiaorong&id=DOI:10.1016/j.sysarc.2026.103933
    Name: Full Text Finder (for New FTF UI) (ns324271)
    Category: fullText
    Text: Full Text Finder
    MouseOverText: Full Text Finder
Header DbId: bsx
DbLabel: Business Source Index
An: 196022261
RelevancyScore: 1355
AccessLevel: 6
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 1355
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: VMOC: MLIR-based memory optimization for CNN deployment on vector accelerators.
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Chen%2C+Xiaorong%22">Chen, Xiaorong</searchLink><relatesTo>1</relatesTo> (AUTHOR)<i> chenxiaorong23@nudt.edu.cn</i><br /><searchLink fieldCode="AR" term="%22Liu%2C+Zhong%22">Liu, Zhong</searchLink><relatesTo>1</relatesTo> (AUTHOR)<i> zhongliu@nudt.edu.cn</i>
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="JN" term="%22Journal+of+Systems+Architecture%22">Journal of Systems Architecture</searchLink>. Oct2026, Vol. 179, pN.PAG-N.PAG. 1p.
– Name: Subject
  Label: Subject Terms
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Computer+memory+management%22">Computer memory management</searchLink><br /><searchLink fieldCode="DE" term="%22Code+generators%22">Code generators</searchLink>
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: Vector accelerators are increasingly prominent platforms for deploying Convolutional Neural Networks (CNNs) due to their high energy efficiency and parallelism. On such architectures, efficient memory utilization is critical for successful deployment. However, existing graph fission methods exhibit several limitations on vector accelerators, including poor hardware adaptability, inaccurate fission cost modeling, low compilation efficiency, and heavy reliance on manual intervention. This paper proposes VMOC, a novel MLIR-based memory optimization method. VMOC automatically identifies memory hotspot subgraphs, constructs a Hardware-aware Fission Cost Model (HFCM) to quantify the memory benefits and execution overhead introduced by fission, and employs an Adaptive Graph Fission Algorithm (AGFA) to dynamically search for fission schemes that satisfy both memory constraints and positive benefit constraints. Meanwhile, a Graph Encapsulation-based Scheduling Optimization (GESO) is introduced to alleviate the problem of increased scheduling complexity caused by graph fission. These optimization strategies are integrated into the MLIR framework, enabling end-to-end automatic compilation from ONNX models to target executable code and generating high-performance code for vector accelerators. Experimental results demonstrate the effectiveness of VMOC on GoogleNet, ResNet18, VGG16, and YOLOv5s. VMOC achieves memory compression ratios of 52.05% ∼ 96.18%, with compilation speedups of 2.34 × ∼ 152.11 × over MAGIS and 12.63 × ∼ 100.93 × over TVM. On FT-M7032 and FT-XT4000 vector accelerators, it achieves computational efficiencies of 62.30% ∼ 74.05% and 50.25% ∼ 71.01%, consistently outperforming VCNN. These results suggest that VMOC has the potential to become an efficient and practical approach for the automatic deployment of CNNs on vector accelerators. [Display omitted] • A memory-hotspot method identifies critical operators affecting peak memory. • The HFCM model evaluates the benefits of fissioning under GSM capacity constraints. • The AGFA dynamically finds the minimum fission granularity under GSM constraints. • The GESO encapsulates local subgraphs into supernodes to suppress graph expansion. • An end-to-end MLIR framework efficiently deploys CNN models on vector accelerators. [ABSTRACT FROM AUTHOR]
– Name: AbstractSuppliedCopyright
  Label:
  Group: Ab
  Data: <i>Copyright of Journal of Systems Architecture is the property of Elsevier B.V. and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract.</i> (Copyright applies to all Abstracts.)
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=bsx&AN=196022261
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1016/j.sysarc.2026.103933
    Languages:
      – Code: eng
        Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 1
        StartPage: N.PAG
    Subjects:
      – SubjectFull: Computer memory management
        Type: general
      – SubjectFull: Code generators
        Type: general
    Titles:
      – TitleFull: VMOC: MLIR-based memory optimization for CNN deployment on vector accelerators.
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Chen, Xiaorong
      – PersonEntity:
          Name:
            NameFull: Liu, Zhong
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 10
              Text: Oct2026
              Type: published
              Y: 2026
          Identifiers:
            – Type: issn-print
              Value: 13837621
          Numbering:
            – Type: volume
              Value: 179
          Titles:
            – TitleFull: Journal of Systems Architecture
              Type: main
ResultId 1