The task that I get is to improve the matrix multiplication performance through the FPGA, the first task is I have to make the lowest time of matrix multiplication. I have to optimize the code to be the most performances. It is measured by the time that calculated, the resource usage on the FPGA board such as Flip-flop, Lookup table, Digital Signal Processing, Block RAM and the matrix multiplication that can multiply on an arbitrary size of the matrix.
After the first task, I get the task that takes the code of the first task to implement to a neural network. The matrix multiplication is one of the processes of a neural network, fully connected, and make it calculate faster and more performance.