File Download
There are no files associated with this item.
Links for fulltext
(May Require Subscription)
- Publisher Website: 10.1016/j.cossms.2025.101226
- Scopus: eid_2-s2.0-105005860565
- Find via
Supplementary
-
Citations:
- Scopus: 0
- Appears in Collections:
Article: Current opinions on memristor-accelerated machine learning hardware
| Title | Current opinions on memristor-accelerated machine learning hardware |
|---|---|
| Authors | |
| Keywords | Accelerator AI hardware Machine learning Memristor Non-volatile memory |
| Issue Date | 1-Jul-2025 |
| Publisher | Elsevier |
| Citation | Current Opinion in Solid State and Materials Science, 2025, v. 37 How to Cite? |
| Abstract | The unprecedented advancement of artificial intelligence has placed immense demands on computing hardware, but traditional silicon-based semiconductor technologies are approaching their physical and economic limit, prompting the exploration of novel computing paradigms. Memristor offers a promising solution, enabling in-memory analog computation and massive parallelism, which leads to low latency and power consumption. This manuscript reviews the current status of memristor-based machine learning accelerators, highlighting the milestones achieved in developing prototype chips, that not only accelerate neural networks inference but also tackle other machine learning tasks. More importantly, it discusses our opinion on current key challenges that remain in this field, such as device variation, the need for efficient peripheral circuitry, and systematic co-design and optimization. We also share our perspective on potential future directions, some of which address existing challenges while others explore untouched territories. By addressing these challenges through interdisciplinary efforts spanning device engineering, circuit design, and systems architecture, memristor-based accelerators could significantly advance the capabilities of AI hardware, particularly for edge applications where power efficiency is paramount. |
| Persistent Identifier | http://hdl.handle.net/10722/365856 |
| ISSN | 2023 Impact Factor: 12.2 2023 SCImago Journal Rankings: 2.420 |
| DC Field | Value | Language |
|---|---|---|
| dc.contributor.author | Jiang, Mingrui | - |
| dc.contributor.author | Xu, Yichun | - |
| dc.contributor.author | Li, Zefan | - |
| dc.contributor.author | Li, Can | - |
| dc.date.accessioned | 2025-11-12T00:36:04Z | - |
| dc.date.available | 2025-11-12T00:36:04Z | - |
| dc.date.issued | 2025-07-01 | - |
| dc.identifier.citation | Current Opinion in Solid State and Materials Science, 2025, v. 37 | - |
| dc.identifier.issn | 1359-0286 | - |
| dc.identifier.uri | http://hdl.handle.net/10722/365856 | - |
| dc.description.abstract | <p>The unprecedented advancement of artificial intelligence has placed immense demands on computing hardware, but traditional silicon-based semiconductor technologies are approaching their physical and economic limit, prompting the exploration of novel computing paradigms. <a href="https://www.sciencedirect.com/topics/physics-and-astronomy/memristor" title="Learn more about Memristor from ScienceDirect's AI-generated Topic Pages">Memristor</a> offers a promising solution, enabling in-memory analog computation and massive <a href="https://www.sciencedirect.com/topics/engineering/parallelism" title="Learn more about parallelism from ScienceDirect's AI-generated Topic Pages">parallelism</a>, which leads to low latency and <a href="https://www.sciencedirect.com/topics/engineering/electric-power-utilization" title="Learn more about power consumption from ScienceDirect's AI-generated Topic Pages">power consumption</a>. This manuscript reviews the current status of memristor-based machine learning accelerators, highlighting the milestones achieved in developing prototype chips, that not only accelerate <a href="https://www.sciencedirect.com/topics/physics-and-astronomy/neural-network" title="Learn more about neural networks from ScienceDirect's AI-generated Topic Pages">neural networks</a> inference but also tackle other machine learning tasks. More importantly, it discusses our opinion on current key challenges that remain in this field, such as device variation, the need for efficient peripheral circuitry, and systematic co-design and optimization. We also share our perspective on potential future directions, some of which address existing challenges while others explore untouched territories. By addressing these challenges through interdisciplinary efforts spanning device engineering, circuit design, and systems architecture, memristor-based accelerators could significantly advance the capabilities of AI hardware, particularly for edge applications where power efficiency is paramount.<br></p> | - |
| dc.language | eng | - |
| dc.publisher | Elsevier | - |
| dc.relation.ispartof | Current Opinion in Solid State and Materials Science | - |
| dc.subject | Accelerator | - |
| dc.subject | AI hardware | - |
| dc.subject | Machine learning | - |
| dc.subject | Memristor | - |
| dc.subject | Non-volatile memory | - |
| dc.title | Current opinions on memristor-accelerated machine learning hardware | - |
| dc.type | Article | - |
| dc.identifier.doi | 10.1016/j.cossms.2025.101226 | - |
| dc.identifier.scopus | eid_2-s2.0-105005860565 | - |
| dc.identifier.volume | 37 | - |
| dc.identifier.eissn | 1879-0348 | - |
| dc.identifier.issnl | 1359-0286 | - |