Towards Scalable and Efficient FPGA Stencil Accelerators - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2016

Towards Scalable and Efficient FPGA Stencil Accelerators

Abstract

In this paper we propose a design template for stencil computations targeting FPGA-based accelerators. The goal for our design is to provide scalable high throughput designs that can efficiently process iterative stencil programs with large size parameters, i.e., those whose data footprint is too large to fit on-chip. Our context is when we seek to use FP-GAs as accelerators attached to CPUs. Minimizing the area is not our primary goal. We propose a family of architectures based on hierarchical tiling, where the inner tiling is used to build coarse-grain data-path operators, increasing computational throughput, and the outer tiling is used to control the memory requirement , specifically data transfers to/from the accelerator. We present preliminary results for Jacobi-style stencils on 1D and 2D data, and are working on fully automating the flow.

Keywords

Fichier principal
Vignette du fichier
impact2016-deest.pdf (432.01 Ko) Télécharger le fichier
Origin Files produced by the author(s)

Dates and versions

hal-01425018 , version 1 (03-01-2017)

Identifiers

  • HAL Id : hal-01425018 , version 1

Cite

Gaël Deest, Nicolas Estibals, Tomofumi Yuki, Steven Derrien, Sanjay Rajopadhye. Towards Scalable and Efficient FPGA Stencil Accelerators. IMPACT'16 - 6th International Workshop on Polyhedral Compilation Techniques, held with HIPEAC'16, Jan 2016, Prague, Czech Republic. ⟨hal-01425018⟩
689 View
412 Download

Share

Gmail Mastodon Facebook X LinkedIn More