Energy-Efficient Convolution Architecture Based on Rescheduled Dataflow
- 7 June 2018
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems I: Regular Papers
- Vol. 65 (12), 4196-4207
- https://doi.org/10.1109/tcsi.2018.2840092
Abstract
This paper presents a rescheduled dataflow of convolution and its hardware architecture that can enhance energy efficiency. For convolution involving a large amount of computations and memory accesses, previous accelerators employed parallel processing elements to meet real-time constraints. Though the previous approaches made a success in implementing complex convolution models, they load the same input features and filter weights from on-chip memories multiple times due to the iterative property of convolution operations, suffering from high energy consumption. To mitigate redundant memory accesses, a novel dataflow is proposed that computes convolution operations incrementally so as to reuse the loaded data as maximally as possible. In addition, several convolution accelerators supporting the rescheduled dataflow are investigated, and qualitative and quantitative analyses are performed to suggest a promising candidate for various convolution models. Simulation results show that the energy efficiency of the proposed accelerator outperforms that of the previous accelerator significantly.Keywords
Funding Information
- National Research Foundation of Korea (NRF-2017R1E1A1A01076992)
- IC Design Education Center
This publication has 18 references indexed in Scilit:
- An Energy-Efficient Precision-Scalable ConvNet Processor in 40-nm CMOSIEEE Journal of Solid-State Circuits, 2016
- Eyeriss: An Energy-Efficient Reconfigurable Accelerator for Deep Convolutional Neural NetworksIEEE Journal of Solid-State Circuits, 2016
- Cnvlutin: Ineffectual-Neuron-Free Deep Neural Network ComputingPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2016
- Cambricon: An Instruction Set Architecture for Neural NetworksPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2016
- Origami: A 803-GOp/s/W Convolutional Network AcceleratorIEEE Transactions on Circuits and Systems for Video Technology, 2016
- Cost-Effective Active Learning for Deep Image ClassificationIEEE Transactions on Circuits and Systems for Video Technology, 2016
- CnvlutinACM SIGARCH Computer Architecture News, 2016
- CambriconACM SIGARCH Computer Architecture News, 2016
- ShiDianNaoPublished by Association for Computing Machinery (ACM) ,2015
- DianNaoACM SIGPLAN Notices, 2014