Tuna: A Static Analysis Approach to Optimizing Deep Neural Networks

04/29/2021
by   Yao Wang, et al.
0

We introduce Tuna, a static analysis approach to optimizing deep neural network programs. The optimization of tensor operations such as convolutions and matrix multiplications is the key to improving the performance of deep neural networks. Many deep learning model optimization mechanisms today use dynamic analysis, which relies on experimental execution on a target device to build a data-driven cost model of the program. The reliance on dynamic profiling not only requires access to target hardware at compilation time but also incurs significant cost in machine resources. We introduce an approach that profiles the program by constructing features based on the target hardware characteristics in order. We use static analysis of the relative performance of tensor operations to optimize the deep learning program. Experiments show that our approach can achieve up to 11x performance compared to dynamic profiling based methods with the same compilation time.

READ FULL TEXT
research
06/04/2020

Nimble: Efficiently Compiling Dynamic Neural Networks for Model Inference

Modern deep neural networks increasingly make use of features such as dy...
research
09/11/2014

Building Program Vector Representations for Deep Learning

Deep learning has made significant breakthroughs in various fields of ar...
research
06/11/2020

Ansor : Generating High-Performance Tensor Programs for Deep Learning

High-performance tensor programs are crucial to guarantee efficient exec...
research
11/10/2019

Using Deep Neural Networks for Estimating Loop Unrolling Factor

Optimizing programs requires deep expertise. On one hand, it is a tediou...
research
06/17/2020

Dynamic Tensor Rematerialization

Checkpointing enables training deep learning models under restricted mem...
research
06/21/2022

Boosting Performance Optimization with Interactive Data Movement Visualization

Optimizing application performance in today's hardware architecture land...
research
04/06/2023

Hardware-Aware Static Optimization of Hyperdimensional Computations

Binary spatter code (BSC)-based hyperdimensional computing (HDC) is a hi...

Please sign up or login with your details

Forgot password? Click here to reset