Structure Harvard

The architecture of the type Harvard is a design of microprocessors which separates the data memory physically and the memory programs. The access to each of the two memories is carried out via two distinct buses.

History

The name of this structure comes from the name of the university where such an architecture was put in practice for the first time with the Mark I in 1944.

Other Architectures

The architecture of the type Von Neumann is opposed to that of Harvard because it uses a single structure to store at the same time the program and the data. With two distinct buses, the architecture known as of Harvard makes it possible simultaneously to transfer the data as well as the instructions to be carried out. Thus, the treatment unit will simultaneously have access to the instruction and the associated data. This model is faster than that of Von Neumann. However, the profit in performance is obtained with the detriment of the internal complexity of the structure.

Applications

Harvard architecture is often used in:

the numerical processors of signal (DSP);
the Microcontrôleur S, in particular the PEAK of Microchip and AVR of Atmel.

See too

Computer
Architecture of von Neumann
Architecture of the processors

Random links:

Structure Harvard

History

Other Architectures

Applications

See too

Related articles