From the course: Computer Architecture Essentials

Unlock this course with a free trial

Join today to access over 24,700 courses taught by industry experts.

Harvard vs. Von Neumann architectures

Harvard vs. Von Neumann architectures

From the course: Computer Architecture Essentials

Harvard vs. Von Neumann architectures

- [Host] Let me tell you about another important distinction to classify computer architectures. This distinction is based on the design of the memory subsystem and its connectivity to the CPU. So far, we have established that programs are composed of instructions, all of which are encoded as data and are stored in the memory subsystem. Well, back in the 1940s, computer designers faced a decision whether to handle executable instructions differently than general data or simply to handle instructions and general data equally, this led to the opposing Harvard and Von Neumann architectures. The Harvard architecture regards executable instructions as a special category of data that requires special treatment, thus storing programs in their own special memory. One implication is that the means to access this memory, its bus only needs to have one direction. This is a read only memory. The CPU has no way to write into it because the CPU is only supposed to obey its program. That's why this…

Contents