There has been an increasing interest in powering electronic systems using solar energy. Designers are seeking for techniques to manufacture solar panels using low cost material in a massive scale. This will likely lead to wide process variation and hence unreliable performance. This paper considers the impact of the process variations on the output power of large Photovoltaic (PV) module by modeling each PV cell as a current source whose short circuit current is a normal random variable. The probability distribution of the overall output power of an NxM PV module is analytically derived. The proposed statistical analysis technique will enable the designer to predict the maximum output power of abstract PV module for a given confidence level. This analysis also reveals that, when the size and the manufacturing technology are given, the efficiency of a PV module is determined by its topology. The proposed power prediction model can be applied to find the optimal structure of the PV module that maximizes the energy harvesting rate at the given confidence level.