PetaOps/W edge-AI µProcessors: Myth or reality?

Manil Dev Gomony, Floran de Putter, Anteneh Gebregiorgis, Gianna Paulin, Linyan Mei, Vikram Jain, Said Hamdioui, Victor Sanchez, Tobias Grosser, Marc Geilen, Marian Verhelst, Friedemann Zenke, Frank Gurkaynak, Barry de Bruin, Sander Stuijk, Simon Davidson, Sayandip De, Mounir Ghogho, Alexandra Jimborean, Sherif EissaLuca Benini, Dimitrios Soudris, Rajendra Bishnoi, Sam Ainsworth, Federico Corradi, Ouassim Karrakchou, Tim Güneysu, Henk Corporaal

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

1 Downloads (Pure)

Abstract

With the rise of deep learning (DL), our world braces for artificial intelligence (AI) in every edge device, creating an urgent need for edge-AI SoCs. This SoC hardware needs to support high throughput, reliable and secure AI processing at ultra-low power (ULP), with a very short time to market. With its strong legacy in edge solutions and open processing platforms, the EU is well-positioned to become a leader in this SoC market. However, this requires AI edge processing to become at least 100 times more energy-efficient, while offering sufficient flexibility and scalability to deal with AI as a fast-moving target. Since the design space of these complex SoCs is huge, advanced tooling is needed to make their design tractable. The CONVOLVE project (currently in Inital stage) addresses these roadblocks. It takes a holistic approach with innovations at all levels of the design hierarchy. Starting with an overview of SOTA DL processing support and our project methodology, this paper presents 8 important design choices largely impacting the energy efficiency and flexibility of DL hardware. Finding good solutions is key to making smart-edge computing a reality.
Original languageEnglish
Title of host publication2023 Design, Automation and Test in Europe Conference and Exhibition, DATE 2023 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers
Number of pages6
ISBN (Electronic)978-3-9819263-7-8
DOIs
Publication statusPublished - 2 Jun 2023
Event2023 Design, Automation & Test in Europe Conference & Exhibition (DATE) - Antwerp, Belgium
Duration: 17 Apr 202319 Apr 2023

Conference

Conference2023 Design, Automation & Test in Europe Conference & Exhibition (DATE)
Abbreviated titleDATE 2023
Country/TerritoryBelgium
CityAntwerp
Period17/04/2319/04/23

Keywords

  • Technological innovation
  • Program processors
  • Scalability
  • Time to market
  • Throughput
  • Hardware
  • Energy efficiency
  • compiler stack
  • approximate computing
  • SoC
  • edge-AI
  • dynamic DL
  • DSE
  • memristor
  • ULP

Fingerprint

Dive into the research topics of 'PetaOps/W edge-AI µProcessors: Myth or reality?'. Together they form a unique fingerprint.

Cite this