Defying Moore: Envisioning the Economics of a Semiconductor Revolution through 12nm Specialization.

This article investigates whether new artificial intelligence (AI) chips can be built at 12nm technology and exceed or at least match chips made at 5nm or lower. The authors cite the four trends in the semiconductor industry including the perceived end of Moore’s law when investigating technology sc...

Full description

Bibliographic Details
Published in:Communications of the ACM Vol. 68; no. 7; pp. 108 - 120
Main Authors: Davies, Michael, Sankaralingam, Karthikeyan
Format: Article
Published: Association for Computing Machinery Jul2025
Subjects:
Online Access:View this record in EBSCOhost
fields @attributes:
  recordID: 1
pdfLink:
plink: https://search.ebscohost.com/login.aspx?direct=true&db=hlh&AN=186249322&site=ehost-live
header:
  @attributes:
    shortDbName: hlh
    uiTerm: 186249322
    longDbName: Humanities International Complete
    uiTag: AN
  controlInfo:
    bkinfo:
    jinfo:
      jid:
        00010782
        ACM
      jtl: Communications of the ACM
      issn: 00010782
      maglogo: N
    pubinfo:
      dt: Jul2025
      vid: 68
      iid: 7
      pid: 68
      pub: Association for Computing Machinery
    artinfo:
      ui:
        186249322
        10.1145/3711920
      ppf: 108
      ppct: 12
      formats:
      tig:
        atl: Defying Moore: Envisioning the Economics of a Semiconductor Revolution through 12nm Specialization.
      aug:
        au:
          Davies, Michael
          Sankaralingam, Karthikeyan
        affil: University of Wisconsin-Madison, Computer Sciences, Madison, Wisconsin, United States
      su:
        Semiconductor design
        Semiconductor industry
        Technological innovations
        Computer architecture
        Moore's law
        Economics
        Graphics processing units
        Bandwidths
        Sustainability
      sug:
        subj:
          Semiconductor design
          Semiconductor industry
          Technological innovations
          Computer architecture
          Moore's law
          Economics
          Graphics processing units
          Bandwidths
          Sustainability
      ab: This article investigates whether new artificial intelligence (AI) chips can be built at 12nm technology and exceed or at least match chips made at 5nm or lower. The authors cite the four trends in the semiconductor industry including the perceived end of Moore’s law when investigating technology scaling and architectural innovation. This research presents tiled decoupled control and compute (TDCC) architecture with Galileo as an example and compares it to others with TDCC, including SIGMA, as well as Nvidia’s graphics processing unit (GPU).
      pubtype: Periodical
      doctype: Article
      src: R
    language: English
    refInfo:
    copyright:
      @attributes:
        flag: Y
      dt:
        @attributes:
          year: 2025
    holdings:
      @attributes:
        islocal: N