In this study, an advanced hardware design for heart sound signal processing based on ensemble empirical mode decomposition (EEMD) is developed and implemented. The EEMD method  is developed to alleviate a key drawback in the original empirical mode decomposition (EMD) algorithm. In a previous research, Huang et al.  developed an adaptive and efficient EMD method for nonlinear and nonstationary signal analysis. The physical meaning of a single intrinsic mode function (IMF) is obscure, and the original EMD algorithm cannot separate signals with different scales into appropriate IMFs. To overcome this major drawback, a noise-assisted data analysis (NADA) method called EEMD is developed. Heart sound signals are fed into the proposed system to simulate the EEMD-fixed-point performance. A comparison of the floating-point and fixed-point results exhibits satisfactory consistency and demonstrates that our design can accommodate wide variations of dynamic ranges and complicated calculations.