Energy Efficiency of Machine Learning in Embedded Systems Using Neuromorphic Hardware
Recently, the application of machine learning on embedded systems has drawn interest in both the research community and industry because embedded systems located at the edge can produce a faster response and reduce network load. However, software implementation of neural networks on Central Processi...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2020-06-01
|
| Series: | Electronics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2079-9292/9/7/1069 |
| _version_ | 1797563710399578112 |
|---|---|
| author | Minseon Kang Yongseok Lee Moonju Park |
| author_facet | Minseon Kang Yongseok Lee Moonju Park |
| author_sort | Minseon Kang |
| collection | DOAJ |
| description | Recently, the application of machine learning on embedded systems has drawn interest in both the research community and industry because embedded systems located at the edge can produce a faster response and reduce network load. However, software implementation of neural networks on Central Processing Units (CPUs) is considered infeasible in embedded systems due to limited power supply. To accelerate AI processing, the many-core Graphics Processing Unit (GPU) has been a preferred device to the CPU. However, its energy efficiency is not still considered to be good enough for embedded systems. Among other approaches for machine learning on embedded systems, neuromorphic processing chips are expected to be less power-consuming and overcome the memory bottleneck. In this work, we implemented a pedestrian image detection system on an embedded device using a commercially available neuromorphic chip, NM500, which is based on NeuroMem technology. The NM500 processing time and the power consumption were measured as the number of chips was increased from one to seven, and they were compared to those of a multicore CPU system and a GPU-accelerated embedded system. The results show that NM500 is more efficient in terms of energy required to process data for both learning and classification than the GPU-accelerated system or the multicore CPU system. Additionally, limits and possible improvement of the current NM500 are identified based on the experimental results. |
| first_indexed | 2024-03-10T18:47:19Z |
| format | Article |
| id | doaj.art-4316410a463a4430939c41f38c8f1eb8 |
| institution | Directory Open Access Journal |
| issn | 2079-9292 |
| language | English |
| last_indexed | 2024-03-10T18:47:19Z |
| publishDate | 2020-06-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Electronics |
| spelling | doaj.art-4316410a463a4430939c41f38c8f1eb82023-11-20T05:24:19ZengMDPI AGElectronics2079-92922020-06-0197106910.3390/electronics9071069Energy Efficiency of Machine Learning in Embedded Systems Using Neuromorphic HardwareMinseon Kang0Yongseok Lee1Moonju Park2Department of Computer Science & Engineering, Incheon National University, Incheon 22012, KoreaDepartment of Computer Science & Engineering, Incheon National University, Incheon 22012, KoreaDepartment of Computer Science & Engineering, Incheon National University, Incheon 22012, KoreaRecently, the application of machine learning on embedded systems has drawn interest in both the research community and industry because embedded systems located at the edge can produce a faster response and reduce network load. However, software implementation of neural networks on Central Processing Units (CPUs) is considered infeasible in embedded systems due to limited power supply. To accelerate AI processing, the many-core Graphics Processing Unit (GPU) has been a preferred device to the CPU. However, its energy efficiency is not still considered to be good enough for embedded systems. Among other approaches for machine learning on embedded systems, neuromorphic processing chips are expected to be less power-consuming and overcome the memory bottleneck. In this work, we implemented a pedestrian image detection system on an embedded device using a commercially available neuromorphic chip, NM500, which is based on NeuroMem technology. The NM500 processing time and the power consumption were measured as the number of chips was increased from one to seven, and they were compared to those of a multicore CPU system and a GPU-accelerated embedded system. The results show that NM500 is more efficient in terms of energy required to process data for both learning and classification than the GPU-accelerated system or the multicore CPU system. Additionally, limits and possible improvement of the current NM500 are identified based on the experimental results.https://www.mdpi.com/2079-9292/9/7/1069embedded systemartificial intelligencehardware accelerationneuromorphic processorpower consumption |
| spellingShingle | Minseon Kang Yongseok Lee Moonju Park Energy Efficiency of Machine Learning in Embedded Systems Using Neuromorphic Hardware Electronics embedded system artificial intelligence hardware acceleration neuromorphic processor power consumption |
| title | Energy Efficiency of Machine Learning in Embedded Systems Using Neuromorphic Hardware |
| title_full | Energy Efficiency of Machine Learning in Embedded Systems Using Neuromorphic Hardware |
| title_fullStr | Energy Efficiency of Machine Learning in Embedded Systems Using Neuromorphic Hardware |
| title_full_unstemmed | Energy Efficiency of Machine Learning in Embedded Systems Using Neuromorphic Hardware |
| title_short | Energy Efficiency of Machine Learning in Embedded Systems Using Neuromorphic Hardware |
| title_sort | energy efficiency of machine learning in embedded systems using neuromorphic hardware |
| topic | embedded system artificial intelligence hardware acceleration neuromorphic processor power consumption |
| url | https://www.mdpi.com/2079-9292/9/7/1069 |
| work_keys_str_mv | AT minseonkang energyefficiencyofmachinelearninginembeddedsystemsusingneuromorphichardware AT yongseoklee energyefficiencyofmachinelearninginembeddedsystemsusingneuromorphichardware AT moonjupark energyefficiencyofmachinelearninginembeddedsystemsusingneuromorphichardware |