This rule checks buffer transfer size to the kernel and determines which ones could take advantage of PLRAM.
The data from host to device can be moved using global memory, or streaming. The global memory consists of two memory resources: DDR memory and PLRAM. Each of these memory types has an advantage over one other. PLRAM have the advantage of having less read/write LATENCY when using them.
If the buffer size is less than 512 Kb, Xilinx recommends using the PLRAM instead of DDR memory as a global memory option.