Temporal Convolutional Networks (TCNs) are emerging lightweight Deep Learning models for Time Series analysis. We introduce an automated exploration approach and a library of optimized kernels to map TCNs on Parallel Ultra-Low Power (PULP) microcontrollers. Our approach minimizes latency and energy by exploiting a layer tiling optimizer to jointly find the tiling dimensions and select among alternative implementations of the causal and dilated 1D-convolution operations at the core of TCNs. We benchmark our approach on a commercial PULP device, achieving up to 103 imes lower latency and 20.3 imes lower energy than the Cube-AI toolkit executed on the STM32L4 and from 2.9 imes to 26.6 imes lower energy compared to commercial closed-source and academic open-source approaches on the same hardware target.
TCN Mapping Optimization for Ultra-Low Power Time-Series Edge Inference
Burrello A.;Dequino A.;Conti F.;Zanghieri M.;Benini L.;
2021
Abstract
Temporal Convolutional Networks (TCNs) are emerging lightweight Deep Learning models for Time Series analysis. We introduce an automated exploration approach and a library of optimized kernels to map TCNs on Parallel Ultra-Low Power (PULP) microcontrollers. Our approach minimizes latency and energy by exploiting a layer tiling optimizer to jointly find the tiling dimensions and select among alternative implementations of the causal and dilated 1D-convolution operations at the core of TCNs. We benchmark our approach on a commercial PULP device, achieving up to 103 imes lower latency and 20.3 imes lower energy than the Cube-AI toolkit executed on the STM32L4 and from 2.9 imes to 26.6 imes lower energy compared to commercial closed-source and academic open-source approaches on the same hardware target.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
