You and your friends have created a software prototype of a new chatbot you named HogGPT to converse with users about all things concerning the razorback nation. After some initial tests you have determined that the inference rate is way to slow and needs to run $1\mathrm{.}4$x faster. Based on your knowledge from digital design you believe you can implement an accelerator in an FPGA to accelerate the matrix vector multiply operations used during inference. You and you$$re your friends have profiled the code and determined that the matrix vector multiply code accounts for $30\%$ of the total code executed.