<![CDATA[The process of deploying algorithms from a host computer to embedded systems such as Raspberry Pi imposes significant computational demands that are often overlooked in embedded system development workflows. Resource-intensive tasks including code generation, cross-compilation, and binary transfer can strain host system resources, yet their quantitative impact remains poorly characterized in the literature. Therefore, a systematic empirical evaluation is necessary to identify computational bottlenecks and guide hardware specification for MATLAB-based embedded development. This study aims to comparatively analyze the computational performance of a host computer before and after the MATLAB algorithm deployment process to a Wi-Fi-connected Raspberry Pi. The methodology employs a Pre-Test and Post-Test experimental design using the MATLAB Hardware Support Package for Raspberry Pi on an Intel Core i7-11800H system with 16 GB DDR4 RAM, running MATLAB R2023a on Windows 11, with CPU utilization, RAM usage, and processor temperature monitored via Task Manager and PowerShell WMI scripts. Experimental results show that RAM usage increased critically by 68% from 7.0 GB to 11.8 GB, while processor temperature rose 19°C from 49.05°C to 68.05°C due to compilation workload. CPU utilization, however, returned to 4% post-deployment, confirming its burst-type characteristic. These results demonstrate that RAM is the primary bottleneck, requiring a minimum of 16 GB with 32 GB recommended, alongside adequate thermal management to prevent throttling in iterative deployment cycle.]]>
Copyrights © 2026
