We show that the UDP/IP core area can be reduced to 1/3 of the original size with an appropriate implementation, accomplished by a trade-off between parallelism/latency and area. Furthermore ...