Skip to Main content Skip to Navigation
Conference papers

GANNoC: A Framework for Automatic Generation of NoC Topologies using Generative Adversarial Networks

Maxime Mirka 1 Maxime France Pillois 1 Gilles Sassatelli 1 Abdoulaye Gamatié 1
1 ADAC - ADAptive Computing
LIRMM - Laboratoire d'Informatique de Robotique et de Microélectronique de Montpellier
Abstract : We propose GANNoC, a framework for automatic generation of customized Network-on-Chip (NoC) topologies, which exploits generative adversarial networks (GANs) learning capabilities. We define the problem of NoC generation as a graph generation problem, and train a GAN to produce such graphs. We further present a Reward-WGAN (RWGAN) architecture, based on the Wasserstein GAN (WGAN). It is coupled to a reward network enabling to steer the resulting generative system towards topologies having desired properties. We illustrate this capability through a case study aimed at producing topologies with a specific number of physical connections. After training, the generative network produces unique topologies with a 36% improvement regarding the number of connections, when compared to those found in the training dataset. NoCs' performance assessment is carried out using the Ratatoskr 3D-NoC simulator with state-of-the-art characteristics. Results suggest interesting opportunities in learning correlations between intrinsic NoC features and resulting performance.
Complete list of metadata

https://hal-lirmm.ccsd.cnrs.fr/lirmm-03107918
Contributor : Maxime Mirka <>
Submitted on : Friday, January 22, 2021 - 3:51:29 PM
Last modification on : Monday, May 31, 2021 - 2:14:02 PM

File

RAPIDO_21_mmirka_halversion.pd...
Files produced by the author(s)

Identifiers

  • HAL Id : lirmm-03107918, version 2

Collections

Citation

Maxime Mirka, Maxime France Pillois, Gilles Sassatelli, Abdoulaye Gamatié. GANNoC: A Framework for Automatic Generation of NoC Topologies using Generative Adversarial Networks. 13th Workshop on Rapid Simulation and Performance Evaluation: Methods and Tools (RAPIDO 2021), Jan 2021, Budapest, Hungary. ⟨lirmm-03107918v2⟩

Share

Metrics

Record views

137

Files downloads

101