English

An Irredundant and Compressed Data Layout to Optimize Bandwidth Utilization of FPGA Accelerators

Hardware Architecture 2024-01-23 v1

Abstract

Memory bandwidth is known to be a performance bottleneck for FPGA accelerators, especially when they deal with large multi-dimensional data-sets. A large body of work focuses on reducing of off-chip transfers, but few authors try to improve the efficiency of transfers. This paper addresses the later issue by proposing (i) a compiler-based approach to accelerator's data layout to maximize contiguous access to off-chip memory, and (ii) data packing and runtime compression techniques that take advantage of this layout to further improve memory performance. We show that our approach can decrease the I/O cycles up to 7×7\times compared to un-optimized memory accesses.

Keywords

Cite

@article{arxiv.2401.12071,
  title  = {An Irredundant and Compressed Data Layout to Optimize Bandwidth Utilization of FPGA Accelerators},
  author = {Corentin Ferry and Nicolas Derumigny and Steven Derrien and Sanjay Rajopadhye},
  journal= {arXiv preprint arXiv:2401.12071},
  year   = {2024}
}

Comments

11 pages, 11 figures, 2 tables

R2 v1 2026-06-28T14:23:42.199Z