Noise is one of the biggest challenges in data acquisition and will be always present on every measurement made by any sensor, specially by Inertial Measurement Units (IMUs). Due to that, there are two approaches to solve this problem: analog or digital filtering. This second approach is usually the most taken, since very good precision can be achieved with mathematical operations performed by processors. However, depending on the application, the least energy consumption, time response or resources used are desirable. So, this paper presents the analysis of the well known Moving Average digital filter on an 8-bit AVR architecture microcontroller, because it is an interesting option as a result of its limited resources and its most likely probability of consuming less energy.