DeepAI AI Chat
Log In Sign Up

HOBFLOPS CNNs: Hardware Optimized Bitsliced Floating-Point Operations Convolutional Neural Networks

by   James Garland, et al.
Trinity College Dublin

Convolutional neural network (CNN) inference is commonly performed with 8-bit integer values. However, higher precision floating-point inference is required. Existing processors support 16- or 32 bit FP but do not typically support custom precision FP. We propose hardware optimized bit-sliced floating-point operators (HOBFLOPS), a method of generating efficient custom-precision emulated bitsliced software FP arithmetic, for CNNs. We compare HOBFLOPS8-HOBFLOPS16 performance against SoftFP16 on Arm Neon and Intel architectures. HOBFLOPS allows researchers to prototype arbitrary-levels of FP arithmetic precision for CNN accelerators. Furthermore, HOBFLOPS fast custom-precision FP CNNs in software may be valuable in cases where memory bandwidth is limited.


page 13

page 17


Customizable Precision of Floating-Point Arithmetic with Bitslice Vector Types

Customizing the precision of data can provide attractive trade-offs betw...

Numerical Stability of DeepGOPlus Inference

Convolutional neural networks (CNNs) are currently among the most widely...

FPIRM: Floating-point Processing in Racetrack Memories

Convolutional neural networks (CNN) have become a ubiquitous algorithm w...

Rethinking Floating Point Overheads for Mixed Precision DNN Accelerators

In this paper, we propose a mixed-precision convolution unit architectur...

Tuning of Mixture-of-Experts Mixed-Precision Neural Networks

Deep learning has become a useful data analysis method, however mainstre...

A MAC-less Neural Inference Processor Supporting Compressed, Variable Precision Weights

This paper introduces two architectures for the inference of convolution...

Understanding Cache Boundness of ML Operators on ARM Processors

Machine Learning compilers like TVM allow a fast and flexible deployment...