Performance evaluation of versal AI engines for radar signal processing

Sammanfattning

This thesis aimed to evaluate the artificial intelligence engines found on the VersalVCK 190 evaluation board. To perform the evaluation, a non-homogeneous detectorbased on the generalised inner product was used. This algorithm was implementedin its entirety in MATLAB and on a graphics processing unit, which was used forcomparison. For the Versal board however, only parts of the algorithm were implementedin the design. Further, the design was never run on the Versal board,instead, simulations were used to achieve the results presented in this thesis.This thesis found no real advantages of using the artificial intelligence engines forthe parts of the generalised inner product algorithm that were implemented. Forcomputations of matrices larger than 24 × 24, the graphics processing unit outperformedthe artificial intelligence engines in terms of speed. Although this reportdid not find any real advantages for using the artificial intelligence engines, furtherstudies are required, especially regarding the power consumption in comparison tothe throughput, before any final conclusions can be drawn regarding the artificialintelligence engines.