Compressing deep neural networks on FPGAs to binary and ternary precision with HLS4ML

03/11/2020
by   Giuseppe Di Guglielmo, et al.
0

We present the implementation of binary and ternary neural networks in the hls4ml library, designed to automatically convert deep neural network models to digital circuits with FPGA firmware. Starting from benchmark models trained with floating point precision, we investigate different strategies to reduce the network's resource consumption by reducing the numerical precision of the network parameters to binary or ternary. We discuss the trade-off between model accuracy and resource consumption. In addition, we show how to balance between latency and accuracy by retaining full precision on a selected subset of network components. As an example, we consider two multiclass classification tasks: handwritten digit recognition with the MNIST data set and jet identification with simulated proton-proton collisions at the CERN Large Hadron Collider. The binary and ternary implementation has similar performance to the higher precision implementation while using drastically fewer FPGA resources.

READ FULL TEXT
research
12/22/2014

Training deep neural networks with low precision multiplications

Multipliers are the most space and power-hungry arithmetic operators of ...
research
03/25/2019

Performance-Efficiency Trade-off of Low-Precision Numerical Formats in Deep Neural Networks

Deep neural networks (DNNs) have been demonstrated as effective prognost...
research
10/23/2018

Deep Neural Network inference with reduced word length

Deep neural networks (DNN) are powerful models for many pattern recognit...
research
05/03/2018

Exploration of Numerical Precision in Deep Neural Networks

Reduced numerical precision is a common technique to reduce computationa...
research
07/11/2018

FINN-L: Library Extensions and Design Trade-off Analysis for Variable Precision LSTM Networks on FPGAs

It is well known that many types of artificial neural networks, includin...
research
05/19/2022

HyBNN and FedHyBNN: (Federated) Hybrid Binary Neural Networks

Binary Neural Networks (BNNs), neural networks with weights and activati...
research
05/30/2023

Implementation of a framework for deploying AI inference engines in FPGAs

The LCLS2 Free Electron Laser FEL will generate xray pulses to beamline ...

Please sign up or login with your details

Forgot password? Click here to reset