The virtual rotating array (VRA) beamforming is a post-processing algorithm used for localizing rotating sound sources. For arrays with arbitrary microphone configurations, spatial interpolation methods are required, among which the inverse distance weighting (IDW) interpolation performs optimally. However, its computational efficiency is unsatisfactory. To address the low computational efficiency of the algorithm for IDW interpolation in VRA beamforming, improvements are made in this paper. Firstly, the algorithm is optimized to reduce redundant computations within loops. Then, by leveraging the periodicity of virtual microphone coordinates, the computational logic of the algorithm is optimized. Through simulations on a benchmark case, the effectiveness and feasibility of the proposed improvement approach are demonstrated, and successfully accelerating the interpolation computations by a hundredfold.