Nowadays, FPGA has become a very useful platform for multiple digital applications. Initially, the hardware programming languages like VHDL or Verilog were the only method for designing the FPGA. In this method, the designer should be able to transform the algorithm of the application into digital blocks. This consumes time and effort. Recently MATLAB realized the FPGA importance and decided to introduce a new tool for FPGA Design; this tool is MATLAB HDL Coder. The idea is to write a very easy MATLAB script and it will be converted to HDL using HDL Coder. Then this HDL code will go through the FPGA regular implementation path. This paper studies and compares, by example, the two methods. The comparison is done for Speed, FPGA utilization, and time for design/implementation. The digital unit under test was AES. The choice of this unit is based on having large input data, it makes many feedbacks, and It needs high speed. The test result doesn't recommend MATLAB HDL Coder for implementation but it recommends it to fast-proof ideas and fast prototypes. This is because the idea of just writing a simple script describing the algorithm results in a very complicated combinational circuit, which has a very low frequency. The recommended future research is to find a way to force the MATLAB script to be implemented in pipeline architecture. The expected result is to improve the performance in two directions utilization and frequency, but It'll lose the main advantage which is fast implementation.