ARTÍCULO
TITULO

A Framework for Ultra Low-Power Hardware Accelerators Using NNs for Embedded Time Series Classification

Daniel Reiser    
Peter Reichel    
Stefan Pechmann    
Maen Mallah    
Maximilian Oppelt    
Amelie Hagelauer    
Marco Breiling    
Dietmar Fey and Marc Reichenbach    

Resumen

In embedded applications that use neural networks (NNs) for classification tasks, it is important to not only minimize the power consumption of the NN calculation, but of the whole system. Optimization approaches for individual parts exist, such as quantization of the NN or analog calculation of arithmetic operations. However, there is no holistic approach for a complete embedded system design that is generic enough in the design process to be used for different applications, but specific in the hardware implementation to waste no energy for a given application. Therefore, we present a novel framework that allows an end-to-end ASIC implementation of a low-power hardware for time series classification using NNs. This includes a neural architecture search (NAS), which optimizes the NN configuration for accuracy and energy efficiency at the same time. This optimization targets a custom designed hardware architecture that is derived from the key properties of time series classification tasks. Additionally, a hardware generation tool is used that creates a complete system from the definition of the NN. This system uses local multi-level RRAM memory as weight and bias storage to avoid external memory access. Exploiting the non-volatility of these devices, such a system can use a power-down mode to save significant energy during the data acquisition process. Detection of atrial fibrillation (AFib) in electrocardiogram (ECG) data is used as an example for evaluation of the framework. It is shown that a reduction of more than 95% of the energy consumption compared to state-of-the-art solutions is achieved.

 Artículos similares

       
 
Priyank Kalgaonkar and Mohamed El-Sharkawy    
Artificial Intelligence (AI) combines computer science and robust datasets to mimic natural intelligence demonstrated by human beings to aid in problem-solving and decision-making involving consciousness up to a certain extent. From Apple?s virtual perso... ver más

 
Alexios Alexiou, Nikolaos Aretakis, Ioannis Kolias and Konstantinos Mathioudakis    
This paper presents a modular, flexible, extendable and fast-computational framework that implements a multidisciplinary, varying fidelity, multi-system approach for the conceptual and preliminary design of novel aero-engines. In its current status, the ... ver más
Revista: Aerospace

 
Rahmat Ullah and Tughrul Arslan    
Microwave imaging systems are currently being investigated for breast cancer, brain stroke and neurodegenerative disease detection due to their low cost, portable and wearable nature. At present, commonly used radar-based algorithms for microwave imaging... ver más
Revista: Algorithms

 
Hang Fan, Xuemin Zhang, Shengwei Mei, Kunjin Chen and Xinyang Chen    
Ultra-short-term wind power prediction is of great importance for the integration of renewable energy. It is the foundation of probabilistic prediction and even a slight increase in the prediction accuracy can exert significant improvement for the safe a... ver más
Revista: Applied Sciences

 
Salam Al-Obaidi, Patrick Bamonte, Massimo Luchini, Iacopo Mazzantini and Liberato Ferrara    
This paper provides the formulation and description of the framework and methodology for a Durability Assessment-based Design approach for structures made of the Ultra-High-Durability Concrete materials conceived, produced and investigated in the project... ver más
Revista: Infrastructures