Systematic design of pipelined recursive filters is presented. The pro
cedure is based on a multiplication algorithm which generates the resu
lt most significant digit first. Since the latency of such a multiplie
r is low, a reduced number of pipelining delays may be introduced in t
he recursion loop. This achieves a high sampling rate. The design proc
edure turns out to be systematic and very easy to apply. The implement
ation so obtained exhibits minimum hardware and insures minimum latenc
y. It is worth noting that its flexibility allows, on one hand, to cho
osing freely the number system radix and, on the other hand, interleav
ing two multiplier arrays into one. This is illustrated in this paper
by the realization of a second order all-pole filter, operating in rad
ix-4 representation and using only one array to perform two multiplica
tions. By this way, long interconnections are avoided and denser and m
ore regular layout is achieved. It turns out that the design procedure
may also be applied successfully to various types of realizations whe
re multiplications are required.