I have some constant data which happen to be in floating-point format. However, I do not require much precision hence want to store it in a fixed-point format.
If I store it in a floating-point format, it takes 32 bit for one data point. I want to reduce the memory required by using a fixed point as ap_fixed<16,2>. However, HLS Cpp is not allowing me to do it. It does not compile the code and does not give a clear error too.
If anyone has any solution for this please let me know.