Because of the parelelism inherent in this type of network it is surprisingly easy to produce a pipelined IDEA machine in hardware which can give huge bandwidth compared with a software implementation,the only limiting factor being the amount of hardware required in the FPGA to keep producing the single cycle modulo multipliers.