![[2602.17508] Pareto Optimal Benchmarking of AI Models on ARM Cortex Processors for Sustainable Embedded Systems](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.17508] Pareto Optimal Benchmarking of AI Models on ARM Cortex Processors for Sustainable Embedded Systems
Summary
This article presents a benchmarking framework for optimizing AI models on ARM Cortex processors, focusing on energy efficiency and performance in embedded systems.
Why It Matters
As AI applications increasingly demand energy efficiency, this research provides a systematic approach to balance performance and sustainability. It offers valuable insights for developers working on embedded systems, ensuring that AI solutions are both effective and environmentally friendly.
Key Takeaways
- Introduces a benchmarking framework for AI models on ARM Cortex processors.
- Highlights the trade-offs between energy consumption and model accuracy using Pareto analysis.
- Identifies the M7 processor as optimal for short inference cycles and the M4 for energy efficiency.
- Demonstrates a near-linear correlation between FLOPs and inference time for reliable performance estimation.
- Guides developers in designing energy-efficient AI systems for real-world applications.
Computer Science > Artificial Intelligence arXiv:2602.17508 (cs) [Submitted on 19 Feb 2026] Title:Pareto Optimal Benchmarking of AI Models on ARM Cortex Processors for Sustainable Embedded Systems Authors:Pranay Jain, Maximilian Kasper, Göran Köber, Axel Plinge, Dominik Seuß View a PDF of the paper titled Pareto Optimal Benchmarking of AI Models on ARM Cortex Processors for Sustainable Embedded Systems, by Pranay Jain and 4 other authors View PDF HTML (experimental) Abstract:This work presents a practical benchmarking framework for optimizing artificial intelligence (AI) models on ARM Cortex processors (M0+, M4, M7), focusing on energy efficiency, accuracy, and resource utilization in embedded systems. Through the design of an automated test bench, we provide a systematic approach to evaluate across key performance indicators (KPIs) and identify optimal combinations of processor and AI model. The research highlights a nearlinear correlation between floating-point operations (FLOPs) and inference time, offering a reliable metric for estimating computational demands. Using Pareto analysis, we demonstrate how to balance trade-offs between energy consumption and model accuracy, ensuring that AI applications meet performance requirements without compromising sustainability. Key findings indicate that the M7 processor is ideal for short inference cycles, while the M4 processor offers better energy efficiency for longer inference tasks. The M0+ processor, while less efficient for...
