This work consists of a systematization of properties analysis of block digital processing. As a consequence, alternative implementation are studied. It is first shown a resume of some aspects of block implementation, i. e., the processing of many samples of the excitation simultaneouly, to examine the stability, properties of some matrixes of interest and facility of calculating the sensibility of such implementation. It is studied the transpose network and comparisons are made with the normal structure analysing its properties. Other realizations are proposed examining aspects such as: number of processors, memory, computing time and delay. It is suggested that a digital filtering should be implemented as a tandem connection of an all pole and FIR filter to allow a simple calculus of the sensibility of the impulsional and frequency responses to the coeficients quantization of the transfer function. Finally, a simulation on a microcomputer in a high level language (BASIC) is given to serve as an example of the block implementation.