Characterizing Sources of Ineffectual Computations in Deep Learning Networks
Autor: | Robert Mullins, Mostafa Mahmoud, Andreas Moshovos, Milos Nikolic, Yiren Zhao |
---|---|
Rok vydání: | 2019 |
Předmět: |
010302 applied physics
Artificial neural network Contextual image classification Computer science business.industry Deep learning Inference Memory bandwidth 02 engineering and technology Image segmentation Machine learning computer.software_genre 01 natural sciences Convolutional neural network Data type 020202 computer hardware & architecture 46 Information and Computing Sciences 0103 physical sciences 4611 Machine Learning 0202 electrical engineering electronic engineering information engineering Artificial intelligence business computer |
Zdroj: | ISPASS |
DOI: | 10.17863/cam.40387 |
Popis: | Hardware accelerators for inference with neural networks can take advantage of the properties of data they process. Performance gains and reduced memory bandwidth during inference have been demonstrated by using narrower data types [1] [2] and by exploiting the ability to skip and compress values that are zero [3]–[6]. Similarly useful properties have been identified at a lower-level such as varying precision requirements [7] and bit-level sparsity [8] [9]. To date, the analysis of these potential sources of superfluous computation and communication has been constrained to a small number of older Convolutional Neural Networks (CNNs) used for image classification. It is an open question as to whether they exist more broadly. This paper aims to determine whether these properties persist in: (1) more recent and thus more accurate and better performing image classification networks, (2) models for image applications other than classification such as image segmentation and low-level computational imaging, (3) Long-Short-Term-Memory (LSTM) models for non-image applications such as those for natural language processing, and (4) quantized image classification models. We demonstrate that such properties persist and discuss the implications and opportunities for future accelerator designs. |
Databáze: | OpenAIRE |
Externí odkaz: |