Affine Multibanking for High-Level Synthesis - ENS de Lyon - École normale supérieure de Lyon Accéder directement au contenu
Communication Dans Un Congrès Année : 2022

Affine Multibanking for High-Level Synthesis

Résumé

In the last decade, FPGAs appeared as a credible alternative for big data and high-performance computing applications. However, programming an FPGA is tedious: given a function to implement, the circuit must be designed from scratch by the developer. In this short paper, we address the compilation of data placement under parallelism and resource constraints. We propose an HLS algorithm able to partition the data across memory banks, so parallel accesses will target distinct banks to avoid data transfer serialization. Our algorithm is able to reduce the number of banks and the maximal bank size. Preliminary evaluation shows promising results.
Fichier principal
Vignette du fichier
impact22mb.pdf (547.06 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03862220 , version 1 (21-11-2022)

Identifiants

  • HAL Id : hal-03862220 , version 1

Citer

Ilham Lasfar, Christophe Alias, Matthieu Moy, Rémy Neveu, Alexis Carré. Affine Multibanking for High-Level Synthesis. IMPACT'22 - 12th International Workshop on Polyhedral Compilation Techniques, Jun 2022, Budapest, Hungary. ⟨hal-03862220⟩
47 Consultations
12 Téléchargements

Partager

Gmail Facebook X LinkedIn More