Von Neumann V Harvard Architecture

Von Neumann v Harvard Architecture

Von Neumann Architecture

Von Neumann Architecture:- Data and Program Memory are accessed on the same busses.

Harvard Architecture

Harvard Architecture:- Data and Program Memory are accessed on different busses.

Von Neumann v Harvard

Harvard Architecture has separate data and instruction busses, allowing transfers to be performed simultaneously on both busses.

A Von Neumann Architecture has only one bus, which is used for both data transfers and instruction fetches, and therefore data transfers and instruction fetches must be scheduled - they cannot be performed at the same time.

 Architectures

Most microcontrollers on the market today have two basic architectures. One architecture refers to the type of bus design used for the program and data memory, von Neumann vs. Harvard. The other classification describes the type of instruction set and the complexity of the CPU, CISC vs. RISC.

|Von Neumann Architecture |Harvard Architecture |

|Advantages |Advantages |

| | |

|Simple. Memory mapping technique is used to keep track of program and|Fast. There is less processor idle time since data and instruction|

|data memory addresses |fetch cycles may overlap. Fast response to interrupts |

| | |

|Busses can be accessed on external pins without a large pin count. |Greater flexibility. Data and instruction busses do not have to be |

|Routing of busses on a PCB is...