Recent Developments in Low-Power AI Accelerators: A Survey
As machine learning and AI continue to rapidly develop, and with the ever-closer end of Moore’s law, new avenues and novel ideas in architecture design are being created and utilized. One avenue is accelerating AI as close to the user as possible, i.e., at the edge, to reduce latency and increase pe...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-11-01
|
| Series: | Algorithms |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1999-4893/15/11/419 |
| _version_ | 1827647555396173824 |
|---|---|
| author | Christoffer Åleskog Håkan Grahn Anton Borg |
| author_facet | Christoffer Åleskog Håkan Grahn Anton Borg |
| author_sort | Christoffer Åleskog |
| collection | DOAJ |
| description | As machine learning and AI continue to rapidly develop, and with the ever-closer end of Moore’s law, new avenues and novel ideas in architecture design are being created and utilized. One avenue is accelerating AI as close to the user as possible, i.e., at the edge, to reduce latency and increase performance. Therefore, researchers have developed low-power AI accelerators, designed specifically to accelerate machine learning and AI at edge devices. In this paper, we present an overview of low-power AI accelerators between 2019–2022. Low-power AI accelerators are defined in this paper based on their acceleration target and power consumption. In this survey, 79 low-power AI accelerators are presented and discussed. The reviewed accelerators are discussed based on five criteria: (i) power, performance, and power efficiency, (ii) acceleration targets, (iii) arithmetic precision, (iv) neuromorphic accelerators, and (v) industry vs. academic accelerators. CNNs and DNNs are the most popular accelerator targets, while Transformers and SNNs are on the rise. |
| first_indexed | 2024-03-09T19:20:40Z |
| format | Article |
| id | doaj.art-0ee230e333904a0a86613917c8653f78 |
| institution | Directory Open Access Journal |
| issn | 1999-4893 |
| language | English |
| last_indexed | 2024-03-09T19:20:40Z |
| publishDate | 2022-11-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Algorithms |
| spelling | doaj.art-0ee230e333904a0a86613917c8653f782023-11-24T03:23:13ZengMDPI AGAlgorithms1999-48932022-11-01151141910.3390/a15110419Recent Developments in Low-Power AI Accelerators: A SurveyChristoffer Åleskog0Håkan Grahn1Anton Borg2Department of Computer Science, Blekinge Institute of Technology, 371 79 Karlskrona, SwedenDepartment of Computer Science, Blekinge Institute of Technology, 371 79 Karlskrona, SwedenDepartment of Computer Science, Blekinge Institute of Technology, 371 79 Karlskrona, SwedenAs machine learning and AI continue to rapidly develop, and with the ever-closer end of Moore’s law, new avenues and novel ideas in architecture design are being created and utilized. One avenue is accelerating AI as close to the user as possible, i.e., at the edge, to reduce latency and increase performance. Therefore, researchers have developed low-power AI accelerators, designed specifically to accelerate machine learning and AI at edge devices. In this paper, we present an overview of low-power AI accelerators between 2019–2022. Low-power AI accelerators are defined in this paper based on their acceleration target and power consumption. In this survey, 79 low-power AI accelerators are presented and discussed. The reviewed accelerators are discussed based on five criteria: (i) power, performance, and power efficiency, (ii) acceleration targets, (iii) arithmetic precision, (iv) neuromorphic accelerators, and (v) industry vs. academic accelerators. CNNs and DNNs are the most popular accelerator targets, while Transformers and SNNs are on the rise.https://www.mdpi.com/1999-4893/15/11/419surveyhardware acceleratorlow-powerperformancemachine learningartificial intelligence |
| spellingShingle | Christoffer Åleskog Håkan Grahn Anton Borg Recent Developments in Low-Power AI Accelerators: A Survey Algorithms survey hardware accelerator low-power performance machine learning artificial intelligence |
| title | Recent Developments in Low-Power AI Accelerators: A Survey |
| title_full | Recent Developments in Low-Power AI Accelerators: A Survey |
| title_fullStr | Recent Developments in Low-Power AI Accelerators: A Survey |
| title_full_unstemmed | Recent Developments in Low-Power AI Accelerators: A Survey |
| title_short | Recent Developments in Low-Power AI Accelerators: A Survey |
| title_sort | recent developments in low power ai accelerators a survey |
| topic | survey hardware accelerator low-power performance machine learning artificial intelligence |
| url | https://www.mdpi.com/1999-4893/15/11/419 |
| work_keys_str_mv | AT christofferaleskog recentdevelopmentsinlowpoweraiacceleratorsasurvey AT hakangrahn recentdevelopmentsinlowpoweraiacceleratorsasurvey AT antonborg recentdevelopmentsinlowpoweraiacceleratorsasurvey |