Skip to Main content Skip to Navigation
Journal articles

Tactics to Directly Map CNN graphs on Embedded FPGAs

Abstract : Deep Convolutional Neural Networks (CNNs) are the state-of-the-art in image classification. Since CNN feed forward propagation involves highly regular parallel computation, it benefits from a significant speed-up when running on fine grain parallel programmable logic devices. As a consequence, several studies have proposed FPGA-based accelerators for CNNs. However, because of the large computationalpower required by CNNs, none of the previous studies has proposed a direct mapping of the CNN onto the physical resources of an FPGA, allocating each processing actor to its own hardware instance.In this paper, we demonstrate the feasibility of the so called direct hardware mapping (DHM) and discuss several tactics we explore to make DHM usable in practice. As a proof of concept, we introduce the HADDOC2 open source tool, that automatically transforms a CNN description into a synthesizable hardware description with platform-independent direct hardware mapping.
Complete list of metadata

Cited literature [17 references]  Display  Hide  Download
Contributor : francois berry Connect in order to contact the contributor
Submitted on : Thursday, November 16, 2017 - 12:01:55 PM
Last modification on : Wednesday, February 24, 2021 - 4:16:02 PM
Long-term archiving on: : Saturday, February 17, 2018 - 1:24:49 PM


Files produced by the author(s)



Kamel Abdelouahab, Maxime Pelcat, Jocelyn Sérot, Cédric Bourrasset, François Berry. Tactics to Directly Map CNN graphs on Embedded FPGAs. IEEE Embedded Systems Letters, Institute of Electrical and Electronics Engineers, 2017, 9 (4), pp.113 - 116. ⟨10.1109/LES.2017.2743247⟩. ⟨hal-01626462⟩



Record views


Files downloads