Synetgy: Algorithm-hardware Co-design for ConvNet Accelerators on Embedded FPGAs
Preprint
Abstract
Using FPGAs to accelerate ConvNets has attracted significant attention in recent years. However, FPGA accelerator design has not leveraged the latest progress of ConvNets. As a result, the key application characteristics such as frames-per-second (FPS) are ignored in favor of simply counting GOPs, and results on accuracy, which is critical to application success, are often not even reported. In this work, we adopt an algorithm-hardware co-design approach to develop a ConvNet accelerator called Synetgy and a novel ConvNet model called DiracDeltaNet$^{\dagger}$. Both the accelerator and ConvNet are tailored to FPGA requirements. DiracDeltaNet, as the name suggests, is a ConvNet with only $1\times 1$ convolutions while spatial convolutions are replaced by more efficient shift operations. DiracDeltaNet achieves competitive accuracy on ImageNet (88.7\% top-5), but with 42$\times$ fewer parameters and 48$\times$ fewer OPs than VGG16. We further quantize DiracDeltaNet's weights to 4-bit and activations to 4-bits, with less than 1\% accuracy loss. These quantizations exploit well the nature of FPGA hardware. In short, DiracDeltaNet's small model size, low computational OP count, low precision and simplified operators allow us to co-design a highly customized computing unit for an FPGA. We implement the computing units for DiracDeltaNet on an Ultra96 SoC system through high-level synthesis. Our accelerator's final top-5 accuracy of 88.1\% on ImageNet, is higher than all the previously reported embedded FPGA accelerators. In addition, the accelerator reaches an inference speed of 66.3 FPS on the ImageNet classification task, surpassing prior works with similar accuracy by at least 11.6$\times$.
Keywords
All Related Versions
This publication has 13 references indexed in Scilit:
- SqueezeSegV2: Improved Model Structure and Unsupervised Domain Adaptation for Road-Object Segmentation from a LiDAR Point Cloud2019
- ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture Design2018
- Towards Effective Low-Bitwidth Convolutional Neural Networks2018
- Learning Transferable Architectures for Scalable Image Recognition2018
- Shift: A Zero FLOP, Zero Parameter Alternative to Spatial Convolutions2018
- SqueezeNext: Hardware-Aware Neural Network Design2018
- SqueezeSeg: Convolutional Neural Nets with Recurrent CRF for Real-Time Road-Object Segmentation from 3D LiDAR Point Cloud2018
- FP-BNN: Binarized neural network on FPGANeurocomputing, 2018
- SqueezeDet: Unified, Small, Low Power Fully Convolutional Neural Networks for Real-Time Object Detection for Autonomous Driving2017
- Software-Hardware Codesign for Efficient Neural Network AccelerationIEEE Micro, 2017