MCU Intelligent Upgrades: An Overview of AI-Enabled Low-Power Technologies
Abstract
1. Introduction
1.1. The Evolution of Low-Power Requirements for MCUs
1.2. The Development of AI Technology and Hardware Integration
1.3. Intelligent Upgrade of Low-Power Technology for MCUs
2. The Technical Approach of Combining MCU with AI
2.1. Lightweight AI Algorithms and Model Adaptation
2.2. AI Acceleration Hardware and Heterogeneous Architectures
3. AI-Enabled Low-Power Optimization Path for MCUs
3.1. Intelligent Task Scheduling
3.2. Intelligent Power Management
3.3. Inference Engine Optimization
3.4. Communication and Data Compression Optimization
4. MCU + AI Enables Low-Power Technology Applications
4.1. Deep Application of Smart Home Scenarios
4.2. Intelligent Innovation in Medical Devices
4.3. Digital Transformation of Smart Agriculture and Industry
5. Challenges and Prospects
Author Contributions
Funding
Data Availability Statement
Acknowledgments
Conflicts of Interest
References
- Zhang, H.Z.; He, M.X.; Shi, L.L.; Wang, P.F. Terahertz Thickness Measurement Based on Stochastic Optimization Algorithm. Spectrosc. Spectr. Anal. 2020, 40, 3066–3070. [Google Scholar] [CrossRef]
- Silva, S.N.; Goldbarg, M.A.S.d.S.; da Silva, L.M.D.; Fernandes, M.A.C. Real-Time Simulator for Dynamic Systems on FPGA. Electronics 2024, 13, 4056. [Google Scholar] [CrossRef]
- Krejci, J.; Babiuch, M.; Babjak, J.; Suder, J.; Wierbica, R. Implementation of an Embedded System into the Internet of Robotic Things. Micromachines 2023, 14, 113. [Google Scholar] [CrossRef]
- Cheshfar, M.; Maghami, M.H.; Amiri, P.; Garakani, H.G.; Lavagno, L. Comparative Survey of Embedded System Implementations of Convolutional Neural Networks in Autonomous Cars Applications. IEEE Access 2024, 12, 182410–182437. [Google Scholar] [CrossRef]
- Son, J.H.; Cho, Y.H.; Kim, J.H.; Joo, Y.S.; So, S.S. Design and Implementation of a sensor node for out-door environmental monitoring. Trans. Korean Inst. Electr. Eng. P 2007, 56, 117–122. [Google Scholar]
- Yang, H.; Yao, J.; Liu, J. A novel speech recognition system-on-chip. In Proceedings of the International Conference on Audio, Language and Image Processing, Shanghai, China, 7–9 July 2008; pp. 764–768. [Google Scholar]
- Xie, G.; Chen, Y.; Xiao, X.; Xu, C.; Li, R.; Li, K. Energy-Efficient Fault-Tolerant Scheduling of Reliable Parallel Applications on Heterogeneous Distributed Embedded Systems. IEEE Trans. Sustain. Comput. 2018, 3, 167–181. [Google Scholar] [CrossRef]
- Tiwari, A.; Ballal, P.; Lewis, F.L. Energy-efficient wireless sensor network design and implementation for condition-based maintenance. ACM Trans. Sens. Netw. (TOSN) 2007, 3, 1-es. [Google Scholar] [CrossRef]
- Kono, T.; Taito, Y.; Hidaka, H. Essential Roles, Challenges and Development of Embedded MCU Micro-Systems to Innovate Edge Computing for the IoT/AI Age. IEICE Trans. Electron. 2020, E103C, 132–143. [Google Scholar] [CrossRef]
- Zangi, U.; Feldman, N.; Hadas, T.; Dayag, N.; Shor, J.; Fish, A. 0.45 v and 18 μA/MHz MCU SOC with Advanced Adaptive Dynamic Voltage Control (ADVC). J. Low Power Electron. Appl. 2018, 8, 14. [Google Scholar] [CrossRef]
- Lee, H.; Kim, E.; Lee, Y.; Kim, H.; Lee, J.; Kim, M.; Yoo, H.-J.; Yoo, S. Toward all-day wearable health monitoring: An ultralow-power, reflective organic pulse oximetry sensing patch. Sci. Adv. 2018, 4, eaas9530. [Google Scholar] [CrossRef]
- Flamand, E.; Rossi, D.; Conti, F.; Loi, I.; Pullini, A.; Rotenberg, F.; Benini, L. GAP-8: A RISC-V SoC for AI at the Edge of the IoT. In Proceedings of the 29th Annual IEEE International Conference on Application-Specific Systems, Architectures and Processors (ASAP), Milan, Italy, 10–12 July 2018; pp. 69–72. [Google Scholar]
- Zhang, Y.; Yin, B.; Gomony, M.D.; Corporaal, H.; Trinitis, C.; Corradi, F. Hardware/Software Co-Design Optimization for Training Recurrent Neural Networks at the Edge. J. Low Power Electron. Appl. 2025, 15, 15. [Google Scholar] [CrossRef]
- Sharifi, S.; Coskun, A.K.; Rosing, T.S. Hybrid dynamic energy and thermal management in heterogeneous embedded multiprocessor SoCs. In Proceedings of the 2010 15th Asia and South Pacific Design Automation Conference (ASP-DAC), Taipei, Taiwan, 18–21 January 2010; pp. 873–878. [Google Scholar]
- Zhao, Z.; Xiao, Z.; Tao, J. MSDG: Multi-Scale Dynamic Graph Neural Network for Industrial Time Series Anomaly Detection. Sensors 2024, 24, 7218. [Google Scholar] [CrossRef]
- Sofianidis, I.; Konstantakos, V.; Nikolaidis, S. Energy Consumption Aspects on Embedded System for IoT Applications. In Proceedings of the 13th International Conference on Modern Circuits and Systems Technologies (MOCAST), Sofia, Bulgaria, 26–28 June 2024. [Google Scholar]
- Capogrosso, L.; Cunico, F.; Cheng, D.S.; Fummi, F.; Cristani, M. A Machine Learning-Oriented Survey on Tiny Machine Learning. IEEE Access 2024, 12, 23406–23426. [Google Scholar] [CrossRef]
- Jiang, B.-Y.; Zhou, F.-C.; Chai, Y. Application of neuromorphic resistive random access memory in image processing. Acta Phys. Sin. 2022, 71, 148504. [Google Scholar] [CrossRef]
- Gebregiorgis, A.; Hoang Anh Du, N.; Yu, J.; Bishnoi, R.; Taouil, M.; Catthoor, F.; Hamdioui, S. A Survey on Memory-centric Computer Architectures. ACM J. Emerg. Technol. Comput. Syst. 2022, 18, 79. [Google Scholar] [CrossRef]
- Li, Z.; Yang, B.; Li, Y. Development of an implantable animal activity rhythm detector and the low-power design. Sheng Wu Yi Xue Gong Cheng Xue Za Zhi J. Biomed. Eng. Shengwu Yixue Gongchengxue Zazhi 2011, 28, 1121–1125. [Google Scholar]
- Parvati, S.V.; Sathish, S.; Thenmozhi, K.; Amirtharajan, R.; Praveenkumar, P. IoT Accelerated Wi-Fi Bot controlled via Node MCU. In Proceedings of the 8th International Conference on Computer Communication and Informatics (ICCCI), Coimbatore, India, 4–6 January 2018. [Google Scholar]
- Araujo, D.; Marques, C.; Oliveira, H.; Piteira, J. A flexible and low-power IoT controller for agri-food field sensing applications. In Proceedings of the 2024 Smart Systems Integration Conference and Exhibition, Hamburg, Germany, 16–18 April 2024. [Google Scholar]
- Ayers, H.R. Making Small Embedded Systems Secure and Dependable. Ph.D. Thesis, Stanford University, Stanford, CA, USA, 2023. [Google Scholar]
- Gomes, P.M.; Brito, M.A. Low-Code Development Platforms: A Descriptive Study. In Proceedings of the 17th Iberian Conference on Information Systems and Technologies (CISTI), Madrid, Spain, 22–25 June 2022. [Google Scholar]
- Jung, V.J.-B.; Burrello, A.; Scherer, M.; Conti, F.; Benini, L. Optimizing the Deployment of Tiny Transformers on Low-Power MCUs. IEEE Trans. Comput. 2025, 74, 526–541. [Google Scholar] [CrossRef]
- Situ, L.; Zhang, C.; Guan, L.; Zuo, Z.; Wang, L.; Li, X.; Liu, P.; Shi, J. Physical Devices-Agnostic Hybrid Fuzzing of IoT Firmware. IEEE Internet Things J. 2023, 10, 20718–20734. [Google Scholar] [CrossRef]
- Vangal, S.; Paul, S.; Hsu, S.; Agarwal, A.; De, V. Near-Threshold Voltage Design Techniques for Heterogenous Manycore System-on-Chips. J. Low Power Electron. Appl. 2020, 10, 16. [Google Scholar] [CrossRef]
- Jisu, K.; Park, D.J. Collaborative Streamlined On-Chip Software Architecture on Heterogenous Multi-Cores for Low-Power Reactive Control in Automotive Embedded Processors. IEMEK J. Embed. Syst. Appl. 2022, 17, 375–382. [Google Scholar] [CrossRef]
- Torres, G.M.; Souza, J.V.L.D.; Aguiar, J.G.F.; Sousa, J.O.d. Performance and Power Consumption of Real-Time Task Scheduling for Embedded Systems Devices. In Proceedings of the 2025 IEEE International Conference on Consumer Electronics (ICCE), Las Vegas, NV, USA, 11–14 January 2025; pp. 1–6. [Google Scholar]
- Zhou, F.; Zhao, L.; Ding, X.; Wang, S. Enhanced DDPG algorithm for latency and energy-efficient task scheduling in MEC systems. Discov. Internet Things 2025, 5, 40. [Google Scholar] [CrossRef]
- Nayak, S.; Patgiri, R.; Waikhom, L.; Ahmed, A. A review on edge analytics: Issues, challenges, opportunities, promises, future directions, and applications. Digit. Commun. Netw. 2024, 10, 783–804. [Google Scholar] [CrossRef]
- Sun, X.; Xu, C.; Li, C. Minimizing Peak Memory Footprint of Inference on IoTs Devices by Efficient Recomputation. In Proceedings of the 19th International Conference on Advanced Intelligent Computing Technology and Applications (ICIC), Zhengzhou, China, 10–13 August 2023; pp. 15–26. [Google Scholar]
- Boyle, L.; Moosmann, J.; Baumann, N.; Heo, S.; Magno, M. DSORT-MCU: Detecting Small Objects in Real Time on Microcontroller Units. IEEE Sens. J. 2024, 24, 40231–40239. [Google Scholar] [CrossRef]
- Tang, X.; Shang, D. Design of High-performance while Energy-efficient Microprocessor with Novel Asynchronous Techniques. In Proceedings of the 35th IEEE International Conference on Application-Specific Systems, Architectures and Processors (ASAP), City Univ Hong Kong, Dept Elect Engn, Hong Kong, China, 24–26 July 2024; pp. 247–248. [Google Scholar]
- Younesi, A.; Fazli, M.A.; Ejlali, A. A Novel Levy Walk-based Framework for Scheduling Power-intensive Mobile Edge Computing Tasks. J. Grid Comput. 2024, 22, 69. [Google Scholar] [CrossRef]
- Abadade, Y.; Temouden, A.; Bamoumen, H.; Benamar, N.; Chtouki, Y.; Hafid, A.S. A Comprehensive Survey on TinyML. IEEE Access 2023, 11, 96892–96922. [Google Scholar] [CrossRef]
- Shi, W.; Cao, J.; Zhang, Q.; Li, Y.; Xu, L. Edge Computing: Vision and Challenges. IEEE Internet Things J. 2016, 3, 637–646. [Google Scholar] [CrossRef]
- Kim, D.; Roh, C.; Baek, D.; Choi, S.-G. Low-Power Preprocessing System at MCU-Based Application Nodes for Reducing Data Transmission. Electronics 2024, 13, 2932. [Google Scholar] [CrossRef]
- Zhang, J.; Xie, X.; Peng, G.; Liu, L.; Yang, H.; Guo, R.; Cao, J.; Yang, J. A Real-Time and Privacy-Preserving Facial Expression Recognition System Using an AI-Powered Microcontroller. Electronics 2024, 13, 2791. [Google Scholar] [CrossRef]
- Oyeronke, A. Trusted Execution Environments for Edge Devices: Architecture and Attacks. Available online: https://www.researchgate.net/publication/393449260_ (accessed on 6 April 2025).
- Samanta, R.; Saha, B.; Ghosh, S.K. TinyML-On-The-Fly: Real-Time Low-Power and Low-Cost MCU-Embedded On-Device Computer Vision for Aerial Image Classification. In Proceedings of the IEEE Space, Aerospace and Defence Conference (SPACE), Bangalore, India, 22–23 July 2024; pp. 194–198. [Google Scholar]
- Kong, H.; Liu, D.; Huai, S.; Luo, X.; Subramaniam, R.; Makaya, C.; Lin, Q.; Liu, W. Edge Compress: Coupling Multidimensional Model Compression and Dynamic Inference for EdgeAI. IEEE Trans. Comput. Aided Des. Integr. Circuits Syst. 2023, 42, 4657–4670. [Google Scholar] [CrossRef]
- Song, Z.R.; Fu, B.Q.; Wu, F.Y.; Jiang, Z.M.; Jiang, L.; Jing, N.F.; Liang, X.Y. DRQ: Dynamic Region-based Quantization for Deep Neural Network Acceleration. In Proceedings of the 47th ACM/IEEE Annual International Symposium on Computer Architecture (ISCA), Electr Network, Valencia, Spain, 3 May–3 June 2020; pp. 1010–1021. [Google Scholar]
- Liu, Z.H.; Wang, Y.H.; Han, K.; Ma, S.W.; Gao, W. Instance-Aware Dynamic Neural Network Quantization. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), New Orleans, LA, USA, 18–24 June 2022; pp. 12424–12433. [Google Scholar]
- Xie, Y.-L.; Lin, X.-R.; Lee, C.-Y.; Lin, C.-W. Design and Implementation of an ARM-Based AI Module for Ectopic Beat Classification Using Custom and Structural Pruned Lightweight CNN. IEEE Sens. J. 2024, 24, 19834–19844. [Google Scholar] [CrossRef]
- Krasteva, V.; Stoyanov, T.; Jekova, I. Implementing Deep Neural Networks on ARM-Based Microcontrollers: Application for Ventricular Fibrillation Detection. Appl. Sci. 2025, 15, 1965. [Google Scholar] [CrossRef]
- Sakr, F.; Bellotti, F.; Berta, R.; De Gloria, A.; Doyle, J. Memory-Efficient CMSIS-NN with Replacement Strategy. In Proceedings of the 8th International Conference on Future Internet of Things and Cloud, Rome, Italy, 23–25 August 2021; pp. 299–303. [Google Scholar]
- Liu, C.; Jobst, M.; Jobst, M.; Guo, L.Y.; Shi, X.Y.; Partzsch, J.; Mayr, C.; Assoc Computing, M. Deploying Machine Learning Models to Ahead-of-Time Runtime on Edge Using MicroTVM. In Proceedings of the IEEE/ACM International Workshop on Compilers, Deployment, and Tooling for Edge AI (CODAI), Hamburg, Germany, 21 September 2023; pp. 37–40. [Google Scholar]
- Diab, M.S.; Rodriguez-Villegas, E. Performance Evaluation of Embedded Image Classification Models Using Edge Impulse for Application on Medical Images. In Proceedings of the 44th Annual International Conference of the IEEE-Engineering-in-Medicine-and-Biology-Society (EMBC), Glasgow, Scotland, 11–15 July 2022; pp. 2639–2642. [Google Scholar]
- Arm Limited. Available online: https://community.arm.com/arm-community-blogs/b/ai-blog/posts/pruning-clustering-arm-ethos-u-npu (accessed on 22 July 2023).
- Sá, P.; Bessa Loureiro, R.; Lisboa, F.; Peixoto, R.; Nascimento, L.; Bonfim, Y.; Cruz, G.; Ramos, T.; Montes, C.; Pagano, T.; et al. Efficient Deployment of Machine Learning Models on Microcontrollers: A Comparative Study of Quantization and Pruning Strategies. In Proceedings of the IX Simpósio Internacional de Inovação e Tecnologia, Salvador, Brazil, 25–27 October 2023; pp. 181–188. [Google Scholar]
- Joshua, C.; Karkala, S.; Krishnapatnam, M.; Aggarwal, A.; Zahir, Z.; Pandhare, V.; Shah, V. Cross-Platform Optimization of ONNX Models for Mobile and Edge Deployment. Int. J. Wirel. Netw. Broadband Technol. 2025. Available online: https://www.researchgate.net/publication/392623112 (accessed on 24 September 2025).
- Pradana, E.Y.; Aji, S.A.; Abdulrrozaq, M.A.; Alasiry, A.H.; Risnumawan, A.; Pitowarno, E. Optimizing YOLOv8 for Real-Time Performance in Humanoid Soccer Robots with OpenVINO. In Proceedings of the 26th International Electronics Symposium (IES) on Shaping the Future—Society 5.0 and Beyond, Politeknik Elektronika Negeri Surabaya, Surabaya, Indonesia, 6–8 August 2024; pp. 304–309. [Google Scholar]
- Shafique, M.A.; Munir, A.; Kong, J. Deep Learning Performance Characterization on GPUs for Various Quantization Frameworks. AI 2023, 4, 926–948. [Google Scholar] [CrossRef]
- Guo, J.; Wang, K.; Sun, J.; Jia, Y. Research and Implementation of Low-Power Anomaly Recognition Method for Intelligent Manhole Covers. Electronics 2023, 12, 1926. [Google Scholar] [CrossRef]
- Dere, M.D.; Jo, J.-H.; Lee, B. Event-Driven Edge Deep Learning Decoder for Real-Time Gesture Classification and Neuro-Inspired Rehabilitation Device Control. IEEE Trans. Instrum. Meas. 2023, 72, 4011612. [Google Scholar] [CrossRef]
- Huang, W.; Wang, D.X.; Ouyang, X.C.; Wan, J.H.; Liu, J.; Li, T.R. Multimodal federated learning: Concept, methods, applications and future directions. Inf. Fusion 2024, 112, 102576. [Google Scholar] [CrossRef]
- Adhikary, S.; Dutta, S. FedTinyWolf—A Memory Efficient Federated Embedded Learning Mechanism. IEEE Embed. Syst. Lett. 2024, 16, 513–516. [Google Scholar] [CrossRef]
- Song, J.; Xie, G.; Li, R.; Chen, X. An Efficient Scheduling Algorithm for Energy Consumption Constrained Parallel Applications on Heterogeneous Distributed Systems. In Proceedings of the 2017 IEEE International Symposium on Parallel and Distributed Processing with Applications and 2017 IEEE International Conference on Ubiquitous Computing and Communications (ISPA/IUCC), Guangzhou, China, 12–15 December 2017; pp. 32–39. [Google Scholar]
- Emilio, M.D.P. BrainChip’s Akida NPU: Redefining AI Processing with Event-Based Architecture. Available online: https://www.embedded.com/brainchips-akida-npu-redefining-ai-processing-with-event-based-architecture (accessed on 1 October 2024).
- Corporation, R.E. Renesas Sets New MCU Performance Bar with 1-GHz RA8P1 Devices with AI Acceleration. Available online: https://www.renesas.com/en/about/newsroom/renesas-sets-new-mcu-performance-bar-1-ghz-ra8p1-devices-ai-acceleration?srsltid=AfmBOoo7t4p8f49vQPZ8sh3YrOiwjPHn47OYsyw8jTGXtqDBoRP0UxQA (accessed on 1 July 2025).
- Farahani, A.; Beithollahi, H.; Fathi, M.; Barangi, R. CNNX: A Low Cost, CNN Accelerator for Embedded System in Vision at Edge. Arab. J. Sci. Eng. 2023, 48, 1537–1545. [Google Scholar] [CrossRef]
- Zhang, D.; Rashid, M.T.; Li, X.K.; Vance, N.; Wang, D. HeteroEdge: Taming The Heterogeneity of Edge Computing System in Social Sensing. In Proceedings of the ACM/IEEE International Conference on Internet of Things Design and Implementation (IoTDI), Montreal, QC, Canada, 15–18 April 2019; pp. 37–48. [Google Scholar]
- Garofalo, A.; Tortorella, Y.; Perotti, M.; Valente, L.; Nadalini, A.; Benini, L.; Rossi, D.; Conti, F. DARKSIDE: A Heterogeneous RISC-V Compute Cluster for Extreme-Edge On-Chip DNN Inference and Training. IEEE Open J. Solid-State Circuits Soc. 2022, 2, 231–243. [Google Scholar] [CrossRef]
- Yang, C.H. AI Acceleration with RISC-V for Edge Computing. In Proceedings of the 2020 International Symposium on VLSI Design, Automation and Test (VLSI-DAT), Hsinchu, Taiwan, 10–13 August 2020. [Google Scholar]
- El Zarif, N.; Hemmat, M.A.; Dupuis, T.; David, J.P.; Savaria, Y. Polara-Keras2c: Supporting Vectorized AI Models on RISC-V Edge Devices. IEEE Access 2024, 12, 171836–171852. [Google Scholar] [CrossRef]
- Manor, E.; Greenberg, S. Custom Hardware Inference Accelerator for TensorFlow Lite for Microcontrollers. IEEE Access 2022, 10, 73484–73493. [Google Scholar] [CrossRef]
- Lee, S.; Lee, K.B.; Joo, S.; Ahn, H.K.; Lee, J.; Kim, D.; Ham, B.; Jung, S.O. SIF-NPU: A 28nm 3.48 TOPS/W 0.25 TOPS/mm2 CNN Accelerator with Spatially Independent Fusion for Real-Time UHD Super-Resolution. In Proceedings of the ESSCIRC 2022—IEEE 48th European Solid State Circuits Conference (ESSCIRC), Milan, Italy, 19–22 September 2022; pp. 97–100. [Google Scholar]
- Qi, M.; Ren, Y.; Sun, T.; Xu, R.; Lv, Z.; Zhou, Y.; Han, S.-T. Self-powered artificial vibrissal system with anemotaxis behavior. Sci. Adv. 2025, 11, eadt3068. [Google Scholar] [CrossRef]
- Texas Instruments Inc. TMS320F28P55x Real-Time Microcontrollers: SPRSP85C[Z]. Revised June 2025. Available online: https://www.ti.com/lit/ds/sprsp85c/sprsp85c.pdf (accessed on 1 April 2024).
- Kena, X.; Wei, H.; Mengke, S.; Wenjun, L.; Tianpei, L. A real-time task scheduling algorithm for multicore embedded systems. In Proceedings of the 2015 Chinese Automation Congress (CAC), Wuhan, China, 27–29 November 2015; pp. 1165–1170. [Google Scholar]
- Zhu, W.; Wu, W.; Yang, X.; Zeng, G. TSSA: Task structure-aware scheduling of energy-constrained parallel applications on heterogeneous distributed embedded platforms. J. Syst. Archit. 2022, 132, 102741. [Google Scholar] [CrossRef]
- Zhang, M.; Liu, L.Q.; Li, C.Z.; Wang, H.F.; Li, M. A Particle Swarm Optimization Method for AI Stream Scheduling in Edge Environments. Symmetry 2022, 14, 2565. [Google Scholar] [CrossRef]
- Rizvandi, N.B.; Taheri, J.; Zomaya, A.Y. Some observations on optimal frequency selection in DVFS-based energy consumption minimization. J. Parallel Distrib. Comput. 2011, 71, 1154–1164. [Google Scholar] [CrossRef]
- Fettes, Q.; Clark, M.; Bunescu, R.; Karanth, A.; Louri, A. Dynamic Voltage and Frequency Scaling in NoCs with Supervised and Reinforcement Learning Techniques. IEEE Trans. Comput. 2019, 68, 375–389. [Google Scholar] [CrossRef]
- Huang, H.; Lin, M.; Yang, L.T.; Zhang, Q. Autonomous Power Management with Double-Q Reinforcement Learning Method. IEEE Trans. Ind. Inform. 2020, 16, 1938–1946. [Google Scholar] [CrossRef]
- Chen, F.; Yu, H.; Jiang, W.; Ha, Y. Quality Optimization of Adaptive Applications via Deep Reinforcement Learning in Energy Harvesting Edge Devices. IEEE Trans. Comput. Aided Des. Integr. Circuits Syst. 2022, 41, 4873–4886. [Google Scholar] [CrossRef]
- Chen, T.Q.; Moreau, T.; Jiang, Z.H.; Zheng, L.M.; Yan, E.; Cowan, M.; Shen, H.; Wang, L.; Hu, Y.; Ceze, L.; et al. TVM: An Automated End-to-End Optimizing Compiler for Deep Learning. In Proceedings of the 13th USENIX Symposium on Operating Systems Design and Implementation (OSDI), Carlsbad, CA, USA, 8–10 October 2018; pp. 579–594. [Google Scholar]
- Sun, W.B.; Katsifodimos, A.; Hai, R. Accelerating machine learning queries with linear algebra query processing. Distrib. Parallel Databases 2025, 43, 8. [Google Scholar] [CrossRef]
- Meddeb, M.; Cagnazzo, M.; Pesquet-Popescu, B. Region-of-interest-based rate control scheme for high-efficiency video coding. APSIPA Trans. Signal Inf. Process. 2014, 3, e16. [Google Scholar] [CrossRef]
- Devindi, I.; Liyanage, S.; Jayarathna, T.; Alawatugoda, J.; Ragel, R. A novel ECG compression algorithm using Pulse-Width Modulation integrated quantization for low-power real-time monitoring. Sci. Rep. 2024, 14, 17162. [Google Scholar] [CrossRef] [PubMed]
- Abdel-Kader, R.F.; El-Sayad, N.E.; Rizk, R.Y. Efficient Noise Reduction System in Industrial IoT Data Streams. In Proceedings of the International Conference on Advanced Intelligent Systems and Informatics 2021, Cairo, Egypt, 11–13 December 2022; pp. 219–232. [Google Scholar]
- Ali, I.; Wassif, K.; Bayomi, H. Dimensionality reduction for images of IoT using machine learning. Sci. Rep. 2024, 14, 7205. [Google Scholar] [CrossRef]
- Ungureanu, V.-I.; Negirla, P.; Korodi, A. Image-Compression Techniques: Classical and “Region-of-Interest-Based” Approaches Presented in Recent Papers. Sensors 2024, 24, 791. [Google Scholar] [CrossRef]
- Pioli, L.; de Macedo, D.D.J.; Costa, D.G.; Dantas, M.A.R. Intelligent Edge-powered Data Reduction: A Systematic Literature Review. ACM Comput. Surv. 2024, 56, 234. [Google Scholar] [CrossRef]
- Pazmiño Ortiz, L.A.; Maldonado Soliz, I.F.; Guevara Balarezo, V.K. Advancing TinyML in IoT: A Holistic System-Level Perspective for Resource-Constrained AI. Future Internet 2025, 17, 257. [Google Scholar] [CrossRef]
- Crafton, B.; Wan, Z.; Spetalnick, S.; Yoon, J.-H.; Wu, W.; Tokunaga, C.; De, V.; Raychowdhury, A. Improving compute in-memory ECC reliability with successive correction. In Proceedings of the 59th ACM/IEEE Design Automation Conference, San Francisco, CA, USA, 10–14 July 2022; pp. 745–750. [Google Scholar]
- Shalan, M.; Hasan, M.R.; Bai, Y.; Li, J. Enhancing Smart Home Security: Blockchain-Enabled Federated Learning with Knowledge Distillation for Intrusion Detection. Smart Cities 2025, 8, 35. [Google Scholar] [CrossRef]
- Distributors, M.-H. Case Studies: Successful HVAC System Implementations in 2025. Available online: https://www.marhy.com/2025-hvac-success-smart-efficient-and-sustainable (accessed on 1 May 2025).
- GigaDevice. How the High-Performance GD32G5 Series MCUs Are Driving Transformation in Digital Energy, Motor Control, and Optical Communications. Available online: https://www.gigadevice.com.cn/about/news-and-event/blog/high-performance-gd32g5-series-mcus (accessed on 1 June 2025).
- Huang, Z.; Herbozo Contreras, L.F.; Leung, W.H.; Yu, L.; Truong, N.D.; Nikpour, A.; Kavehei, O. Efficient Edge-AI Models for Robust ECG Abnormality Detection on Resource-Constrained Hardware. J. Cardiovasc. Transl. Res. 2024, 17, 879–892. [Google Scholar] [CrossRef] [PubMed]
- Zhou, H.; Huang, F.; Lou, W.; Gu, Q.; Ye, Z.; Hu, H.; Zhang, X. Yield prediction through UAV-based multispectral imaging and deep learning in rice breeding trials. Agric. Syst. 2025, 223, 104214. [Google Scholar] [CrossRef]
- Pedroso, A.F.; Khera, R. Leveraging AI-enhanced digital health with consumer devices for scalable cardiovascular screening, prediction, and monitoring. npj Cardiovasc. Health 2025, 2, 34. [Google Scholar] [CrossRef]
- Antoun, I.; Abdelrazik, A.; Eldesouky, M.; Li, X.; Layton, G.R.; Zakkar, M.; Somani, R.; Ng, G.A. Artificial intelligence in atrial fibrillation: Emerging applications, research directions and ethical considerations. Front. Cardiovasc. Med. 2025, 12, 1596574. [Google Scholar] [CrossRef] [PubMed]
- Zhao, Y.F.; Chaw, J.K.; Ang, M.C.; Tew, Y.; Shi, X.Y.; Liu, L.; Cheng, X. A safe-enhanced fully closed-loop artificial pancreas controller based on deep reinforcement learning. PLoS ONE 2025, 20, e0317662. [Google Scholar] [CrossRef] [PubMed]
- Venkateswara, S.M.; Padmanabhan, J. Deep learning based agricultural pest monitoring and classification. Sci. Rep. 2025, 15, 8684. [Google Scholar] [CrossRef] [PubMed]
- Li, L.M.; Zhang, L.; Zhang, Y.F. The Intelligent Design of Solar LED Street Lamps Based on MCU. In Proceedings of the 2nd International Conference on Materials and Products Manufacturing Technology (ICMPMT 2012), Guangzhou, China, 22–23 September 2012; pp. 2005–2008. [Google Scholar]
- Ng, N.; Souza, A.; Diggavi, S.; Suri, N.; Abdelzaher, T.; Towsley, D.; Shenoy, P. Collaborative Inference in Resource-Constrained Edge Networks: Challenges and Opportunities. In Proceedings of the 2024 Military Communications Conference, Washington, DC, USA, 28 October–1 November 2024. [Google Scholar]
- Wang, L.F.; Li, W.Z.; Zhou, Z.D.; An, J.J.; Ye, W.; Li, Z.; Gao, H.H.; Hu, H.Y.; Liu, J.; Chen, X.M.; et al. A near-threshold memristive computing-in-memory engine for edge intelligence. Nat. Commun. 2025, 16, 5897. [Google Scholar] [CrossRef]
- Ding, C.; Ren, Y.; Liu, Z.; Wong, N. Transforming memristor noises into computational innovations. Commun. Mater. 2025, 6, 149. [Google Scholar] [CrossRef]
- Immonen, R.; Hamalainen, T. Tiny Machine Learning for Resource-Constrained Microcontrollers. J. Sens. 2022, 2022, 7437023. [Google Scholar] [CrossRef]
- Lin, J.; Chen, W.-M.; Lin, Y.; Cohn, J.; Gan, C.; Han, S. MCUNet: Tiny Deep Learning on IoT Devices. In Proceedings of the 34th Conference on Neural Information Processing Systems (NeurIPS), Electr Network, Vancouver, BC, Canada, 6–12 December 2020. [Google Scholar]
- Wang, P. Neural Network Optimization Framework for NPU-MCU Heterogeneous Platforms. Appl. Comput. Eng. 2025, 145, 43–50. [Google Scholar] [CrossRef]
- Liu, Y.; Qu, H.; Chen, S.; Feng, X. Energy efficient task scheduling for heterogeneous multicore processors in edge computing. Sci. Rep. 2025, 15, 11819. [Google Scholar] [CrossRef]
Stage | Key Scenes | Key Technology | Popular MCU |
---|---|---|---|
Stage 1 | Traditional industrial equipment, low-end home appliances | Reduce static current | 8-bit |
Stage 2 | Wearable devices, IoT sensors | Sleep–wake mechanism, DVFS technology | 16-bit, 32-bit |
Stage 3 | Smart IoT, Industry 4.0, New Energy Vehicles | System architecture optimization, AI integration | 32-bit |
Toolchain | Quantization Support | Pruning Support |
---|---|---|
LiteRT [46] | INT8 Quantization (Dynamic/Static) supports mixed precision (FP16/INT8) | Structured pruning (removing redundant channels/layers) combined with dynamic range Quantization to optimize model size |
CMSIS-NN [47] | INT8/INT16 Quantization Q Format Conversion | Structural pruning must be manually implemented in conjunction with CMSIS-DSP, relying on the sparsity of quantized weights for optimization. |
MicroTVM [48] | INT8 Quantization supports dynamic shapes | Structured pruning (based on TVM sparse IR) requires integration with model compression tools |
Edge Impulse EON [49] | 8-bit Quantization (automatically generated) | Relies on EON Tuner for architecture search optimization and does not directly support pruning |
Arm Ethos-U [50] | INT8 Quantization supports mixed precision | Structured pruning (requires preprocessing tools), implemented via the Vela compiler for weight clustering and compression |
Compression Method | Toolchain Support |
---|---|
Quantify | TensorFlow Lite [51] (PTQ/QAT) ONNX Runtime [52] (INT8) Intel OpenVINO [53] NVIDIA TensorRT [54] (INT8/FP16) |
Structured Pruning | TensorFlow Lite (TF-MOT) NVIDIA TensorRT Arm Ethos-U |
Unstructured pruning | TVM ONNX Runtime |
Mixed Precision Optimization | NVIDIA TensorRT Arm Ethos-U |
Model Reconstruction | LiteRT TVM |
Solution | Advantages | Disadvantages | Performance Metrics | Source |
---|---|---|---|---|
RL-based DVFS | Adjust speed and voltage according to real-time demand to save active power consumption | Switching delay and voltage stability; achieving complex | Energy consumption reduced by 5–18%, operating time decreased by 17.9% | [76,77] |
Predictive Task Scheduling | Resource utilization has been significantly optimized; Task response delays have been substantially reduced | Additional computational and data overhead; Data privacy and security risks | Task execution time reduced by 31.6%, scheduling time reduced by 40% | [72,73] |
Lightweight Model | Reduce computational load and memory usage to extend battery life | Accuracy decreases, requiring more complex preprocessing or model tuning | Model size reduced by a factor of 19, number of parameters reduced by a factor of 13,960 | [45,86] |
NPU Acceleration | Dedicated MAC array accelerates inference with low latency | Increase chip area and static power consumption; Initialization/switching overhead | Inference speed increased by 724 times | [67] |
CIM Architecture | Significantly reduce data transport energy consumption and improve MAC efficiency | New process and analog circuit challenges; affected by PVT variations | Energy efficiency reaches the level of 1000 TOPS per watt | [87] |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Zhang, T.; Huang, B.; Liu, X.; Fan, J.; Li, J.; Yue, Z.; Wang, Y. MCU Intelligent Upgrades: An Overview of AI-Enabled Low-Power Technologies. J. Low Power Electron. Appl. 2025, 15, 60. https://doi.org/10.3390/jlpea15040060
Zhang T, Huang B, Liu X, Fan J, Li J, Yue Z, Wang Y. MCU Intelligent Upgrades: An Overview of AI-Enabled Low-Power Technologies. Journal of Low Power Electronics and Applications. 2025; 15(4):60. https://doi.org/10.3390/jlpea15040060
Chicago/Turabian StyleZhang, Tong, Bosen Huang, Xiewen Liu, Jiaqi Fan, Junbo Li, Zhao Yue, and Yanfang Wang. 2025. "MCU Intelligent Upgrades: An Overview of AI-Enabled Low-Power Technologies" Journal of Low Power Electronics and Applications 15, no. 4: 60. https://doi.org/10.3390/jlpea15040060
APA StyleZhang, T., Huang, B., Liu, X., Fan, J., Li, J., Yue, Z., & Wang, Y. (2025). MCU Intelligent Upgrades: An Overview of AI-Enabled Low-Power Technologies. Journal of Low Power Electronics and Applications, 15(4), 60. https://doi.org/10.3390/jlpea15040060