Salvato in:
Dettagli Bibliografici
Autori principali: Ferry, Corentin, Derumigny, Nicolas, Derrien, Steven, Rajopadhye, Sanjay
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:https://arxiv.org/abs/2401.12071
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866916101732958208
author Ferry, Corentin
Derumigny, Nicolas
Derrien, Steven
Rajopadhye, Sanjay
author_facet Ferry, Corentin
Derumigny, Nicolas
Derrien, Steven
Rajopadhye, Sanjay
contents 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 $7\times$ compared to un-optimized memory accesses.
format Preprint
id arxiv_https___arxiv_org_abs_2401_12071
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle An Irredundant and Compressed Data Layout to Optimize Bandwidth Utilization of FPGA Accelerators
Ferry, Corentin
Derumigny, Nicolas
Derrien, Steven
Rajopadhye, Sanjay
Hardware Architecture
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 $7\times$ compared to un-optimized memory accesses.
title An Irredundant and Compressed Data Layout to Optimize Bandwidth Utilization of FPGA Accelerators
topic Hardware Architecture
url https://arxiv.org/abs/2401.12071