Kernel read utilization on this specific port is low.
This message is generated whenever the runtime environment considers that the utilization of a port is too low. This can be a combination of port data width and burst inference.
If the data width of a port is far less than 512 bits, consider packaging multiple values into a single transfer. The 512 bits can be efficiently packed and unpacked on the receiving side by using a separate process to read and write to the AXI interface.
These processes can also be used to infer burst reads and writes. Each of these processes should contain a loop with a simple variable incremented by one to read/write from the port and store the value internally.
Applying the DATAFLOW pragma to these processes surrounding the actual algorithm results in an efficient implementation.