Decentor-V: Lightweight ML Training on Low-Power RISC-V Edge Devices

2025/09/10 by Marcelo Ribeiro, Diogo Costa, Ribeiro, Marcelo +7
Computer Science · #FOS: Computer and information sciences #Hardware Architecture (cs.AR) #IoT and Edge/Fog Computing #Machine Learning (cs.LG) #Privacy-Preserving Technologies in Data #Stochastic Gradient Optimization Techniques

paper · pdf · doi:10.48550/arxiv.2509.18118

openalex publication_date 2025/09/10 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Modern IoT devices increasingly rely on machine learning solutions to process data locally. However, the lack of graphics processing units (GPUs) or dedicated accelerators on most platforms makes on-device training largely infeasible, often requiring cloud-based services to perform this task. This procedure often raises privacy-related concerns, and creates dependency on reliable and always-on connectivity. Federated Learning (FL) is a new trend that addresses these issues by enabling decentralized and collaborative training directly on devices, but it requires highly efficient optimization algorithms. L-SGD, a lightweight variant of stochastic gradient descent, has enabled neural network training on Arm Cortex-M Microcontroller Units (MCUs). This work extends L-SGD to RISC-V-based MCUs, an open and emerging architecture that still lacks robust support for on-device training. L-SGD was evaluated on both Arm and RISC-V platforms using 32-bit floating-point arithmetic, highlighting the performance impact of the absence of Floating-Point Units (FPUs) in RISC-V MCUs. To mitigate these limitations, we introduce an 8-bit quantized version of L-SGD for RISC-V, which achieves nearly 4x reduction in memory usage and a 2.2x speedup in training time, with negligible accuracy degradation.

Citations

• Lightweight Deep Learning for Resource-Constrained Environments: A Survey
• Training Machine Learning models at the Edge: A Survey
• Learn What You Need in Personalized Federated Learning
• LLM.int8(): 8-bit Matrix Multiplication for Transformers at Scale
• Gradient Descent, Stochastic Optimization, and Other Tales
• Performance Analysis and Comparison of Machine and Deep Learning Algorithms for IoT Data Classification
• Federated Learning: Challenges, Methods, and Future Directions
• CMSIS-NN: Efficient Neural Network Kernels for Arm Cortex-M CPUs
• An overview of gradient descent optimization algorithms
• Adam: A Method for Stochastic Optimization

Related

• Tin-Tin: Towards Tiny Learning on Tiny Devices with Integer-based Neural Network Training
• RockNet: Distributed Learning on Ultra-Low-Power Devices
• On-FPGA Training with Ultra Memory Reduction: A Low-Precision Tensor Method
• On-Device Training Under 256KB Memory
• PIM-Opt: Demystifying Distributed Optimization Algorithms on a Real-World Processing-In-Memory System