View in EDS PDF Full Text

Performance Analysis of Two 8-Bit Floating-Point-based Piecewise Uniform Quantizers for a Laplacian Data Source.

Saved in:
Bibliographic Details
Title: Performance Analysis of Two 8-Bit Floating-Point-based Piecewise Uniform Quantizers for a Laplacian Data Source.
Authors: Nikolic, Jelena R.1 jelena.nikolic@elfak.ni.ac.rs, Peric, Zoran H.1 zoran.peric@elfak.ni.ac.rs, Jovanovic, Aleksandra Z.1 aleksandra.jovanovic@elfak.ni.ac.rs, Tomic, Stefan S.2 stefan.tomic@actvet.gov.ae, Peric, Sofija Z.1 stefan.tomic@actvet.gov.ae
Source: Electronics & Electrical Engineering. 2025, Vol. 31 Issue 1, p56-61. 6p.
Subjects: Floating-point arithmetic, Analogy, Exponents, Encoding
Abstract: In this paper, we employ the analogy between the representation of the floating-point (FP) format and the representation level distribution of the piecewise uniform quantizer (PWUQ) to assess the performance of FP-based solutions more thoroughly. We present theoretical derivations to assess the performance of the FP format and the PWUQ determined by this format for input data from the Laplacian source. We also provide a performance comparison of two selected 8-bit FP-based PWUQs. Beyond the typical evaluation of the applied FP format, through the accuracy degradation caused by the application of both FP8 solutions in neural network compression, we also use objective quantization measures. This approach offers insights into the robustness of these 8-bit FP-based solutions with respect to changes in input variance, which can be important when the input variance changes. The results demonstrate that the allocation of bits to encode the exponent and mantissa in the FP8 format is important, as it can significantly impact overall performance. [ABSTRACT FROM AUTHOR]
Copyright of Electronics & Electrical Engineering is the property of Electronics & Electrical Engineering 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.)
Database: Engineering Source
FullText Links:
  – Type: pdflink
Text:
  Availability: 0
Header DbId: egs
DbLabel: Engineering Source
An: 183767064
AccessLevel: 6
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Performance Analysis of Two 8-Bit Floating-Point-based Piecewise Uniform Quantizers for a Laplacian Data Source.
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Nikolic%2C+Jelena+R%2E%22">Nikolic, Jelena R.</searchLink><relatesTo>1</relatesTo><i> jelena.nikolic@elfak.ni.ac.rs</i><br /><searchLink fieldCode="AR" term="%22Peric%2C+Zoran+H%2E%22">Peric, Zoran H.</searchLink><relatesTo>1</relatesTo><i> zoran.peric@elfak.ni.ac.rs</i><br /><searchLink fieldCode="AR" term="%22Jovanovic%2C+Aleksandra+Z%2E%22">Jovanovic, Aleksandra Z.</searchLink><relatesTo>1</relatesTo><i> aleksandra.jovanovic@elfak.ni.ac.rs</i><br /><searchLink fieldCode="AR" term="%22Tomic%2C+Stefan+S%2E%22">Tomic, Stefan S.</searchLink><relatesTo>2</relatesTo><i> stefan.tomic@actvet.gov.ae</i><br /><searchLink fieldCode="AR" term="%22Peric%2C+Sofija+Z%2E%22">Peric, Sofija Z.</searchLink><relatesTo>1</relatesTo><i> stefan.tomic@actvet.gov.ae</i>
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="JN" term="%22Electronics+%26+Electrical+Engineering%22">Electronics & Electrical Engineering</searchLink>. 2025, Vol. 31 Issue 1, p56-61. 6p.
– Name: Subject
  Label: Subjects
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Floating-point+arithmetic%22">Floating-point arithmetic</searchLink><br /><searchLink fieldCode="DE" term="%22Analogy%22">Analogy</searchLink><br /><searchLink fieldCode="DE" term="%22Exponents%22">Exponents</searchLink><br /><searchLink fieldCode="DE" term="%22Encoding%22">Encoding</searchLink>
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: In this paper, we employ the analogy between the representation of the floating-point (FP) format and the representation level distribution of the piecewise uniform quantizer (PWUQ) to assess the performance of FP-based solutions more thoroughly. We present theoretical derivations to assess the performance of the FP format and the PWUQ determined by this format for input data from the Laplacian source. We also provide a performance comparison of two selected 8-bit FP-based PWUQs. Beyond the typical evaluation of the applied FP format, through the accuracy degradation caused by the application of both FP8 solutions in neural network compression, we also use objective quantization measures. This approach offers insights into the robustness of these 8-bit FP-based solutions with respect to changes in input variance, which can be important when the input variance changes. The results demonstrate that the allocation of bits to encode the exponent and mantissa in the FP8 format is important, as it can significantly impact overall performance. [ABSTRACT FROM AUTHOR]
– Name: AbstractSuppliedCopyright
  Label:
  Group: Ab
  Data: <i>Copyright of Electronics & Electrical Engineering is the property of Electronics & Electrical Engineering 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=egs&AN=183767064
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.5755/j02.eie.37430
    Languages:
      – Code: eng
        Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 6
        StartPage: 56
    Subjects:
      – SubjectFull: Floating-point arithmetic
        Type: general
      – SubjectFull: Analogy
        Type: general
      – SubjectFull: Exponents
        Type: general
      – SubjectFull: Encoding
        Type: general
    Titles:
      – TitleFull: Performance Analysis of Two 8-Bit Floating-Point-based Piecewise Uniform Quantizers for a Laplacian Data Source.
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Nikolic, Jelena R.
      – PersonEntity:
          Name:
            NameFull: Peric, Zoran H.
      – PersonEntity:
          Name:
            NameFull: Jovanovic, Aleksandra Z.
      – PersonEntity:
          Name:
            NameFull: Tomic, Stefan S.
      – PersonEntity:
          Name:
            NameFull: Peric, Sofija Z.
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 01
              Text: 2025
              Type: published
              Y: 2025
          Identifiers:
            – Type: issn-print
              Value: 13921215
          Numbering:
            – Type: volume
              Value: 31
            – Type: issue
              Value: 1
          Titles:
            – TitleFull: Electronics & Electrical Engineering
              Type: main
ResultId 1