We provide benchmark performance results of a vision-based navigation algorithm for vehicle guidance, navigation and control (GN&C). Vision-based Terrain Relative Navigation was designed for absolute navigation in a lunar orbit, motivated by the need to provide a precision position/velocity estimation algorithm for precisely locating a lunar orbiter or lander in its orbit, without the benefit of a geo-positioning satellite (GPS) system. The concept and the methods however can be extended to vision-based navigation in any GPS-denied areas for which an accurate terrain database is available. The algorithm is written in C and was previously benchmarked on the BAE RAD750 radiation-hardened processor (133 MHz system clock). After parallelization of portions of the algorithm for the multi-core architecture, the benchmarks were repeated on the Maestro processor (260 MHz system clock) for comparison. In addition, we provide performance results for the (FFTW) 1-D FFT and examine comparisons of the Maestro architecture with the RAD750. Future work concerning the incorporation of Maestro into fault tolerant processing architectures for space is considered.