Skip to Main content Skip to Navigation
Conference papers

Instruction Set Design Methodology for In-Memory Computing through QEMU-based System Emulator

Kevin Mambu 1 Julie Dumas 1 Henri-Pierre Charles 1 Maha Kooli 1
1 LFIM - Laboratoire Fonctions Innovantes pour circuits Mixtes
UGA - Université Grenoble Alpes, DSCIN - Département Systèmes et Circuits Intégrés Numériques : DRT/LIST/DSCIN
Abstract : In-Memory Computing (IMC) is a promising paradigm to mitigate the von Neumann bottleneck. However its evaluation on complete applications in the context of full-scale systems is limited by the complexity of simulation frameworks as well is the disjunction between hardware exploration and compiler support. This paper proposes a global exploration flow in the scale of Instruction Set Architectures (ISA) to perform both modeling and the generation of compiler support to perform ISAlevel exploration. Our emulation methodology is based on QEMU and allows the modeling of cache hierarchies, while our compiler support is automatically generated and based on a specialized compiler. We evaluate three applications in the domains of image processing and linear algebra on a reference IMC architecture, and analyze the obtained results to validate our methodology.
Complete list of metadata

https://hal-cea.archives-ouvertes.fr/cea-03452244
Contributor : Contributeur Map Cea Connect in order to contact the contributor
Submitted on : Friday, November 26, 2021 - 6:30:09 PM
Last modification on : Saturday, December 4, 2021 - 3:51:55 AM

File

2021-RSP_camera_ready.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : cea-03452244, version 1

Citation

Kevin Mambu, Julie Dumas, Henri-Pierre Charles, Maha Kooli. Instruction Set Design Methodology for In-Memory Computing through QEMU-based System Emulator. 32nd International Workshop on Rapid System Prototyping (RSP), Oct 2021, (En ligne), France. ⟨cea-03452244⟩

Share

Metrics

Les métriques sont temporairement indisponibles