SMOF: Streaming Modern CNNs on FPGAs with Smart Off-Chip Eviction

arXiv:2403.18921v1 Announce Type: cross
Abstract: Convolutional Neural Networks (CNNs) have demonstrated their effectiveness in numerous vision tasks. However, their high processing requirements necessitate efficient hardware acceleration to meet the application's performance targets. In the space of FPGAs, streaming-based dataflow architectures are often adopted by users, as significant performance gains can be achieved through layer-wise pipelining and reduced off-chip memory access by retaining data on-chip. However, modern topologies, such as the UNet, YOLO, and X3D models, utilise long skip connections, requiring …

abstract application architectures arxiv chip cnns convolutional neural networks cs.ar cs.cv cs.lg dataflow fpgas hardware however modern networks neural networks performance processing requirements smart space s performance streaming targets tasks type vision