You are on page 1of 3

Weiye Kuang

September 15, 2019


Extended Definition

This definition is for someone who is interesting in the computer architecture and
the development of computer.

Von Neumann Architecture


Von Neumann Architecture is the most common computer architecture, first

published in 1945 . Nowadays, all computers more or less are based on this design.

In this definition, I will tell you what the Von Neumann architecture is.

Figure 1

The Von Neumann architecture is a computer architecture designed by John

von Neumann and others, consisting of four main sub-systems: Memory, Arithmetic

and Logic Unit (ALU), Control Unit, and Input / Output System (I/O). Each subsystem

has its own function. Memory, which contains many cells of a fixed size, is used to

store program and data when the computer is running. The ALU performs logical

operations and mathematical operations. The Control Unit fetches instruction,

decodes it and executes it when the computer is working. The Input / Output
systems is the basis of the communication between the computer and people. With

the help of the this system, people can input the information and the computer can

display its output.

How computer work.

At first, the instruction is fetched from the Memory and transmitted to the

Cache, which is accessible to the central processing unit (CPU). Then the Control Unit

received and decodes the instruction into a sequence of electric signals. Then, the

decoded instruction is moved to the relevant function unit of CPU to perform the

actions required by the instruction. Finally, once the program is stop, the results are

moved to the memory across the data bus and stored in the memory.

Figure 2

Compared to the Harvard architecture, another famous computer architecture,

in which machine instructions and data are stored in separate memory units and are

connected with different buses, Von Neumann architecture is much simpler and easy

to design. Since in Von Neumann architecture, the same memory and data bus are

used to store and transfer both instructions and data, it only needs one set of data

buses for both instructions fetch and data transfer. However, the shared buses and
memory units also lead to the design’s limitation, which is called Von Neumann

bottleneck. In Von Neumann architecture, only one bit of information can be

accessed at the same time. If you want a computer in Von Neumann architecture to

perform an operation on some data in the memory, the computer at first moves the

data across the bus into CPU. After the operation is finished, the computer transfers

the output into the memory across the same bus. Then the computer cannot fetch

new data unless it has written the output into the memory. So the CPU also has to

wait for needed data to move from memory. As a result, the processing speed is

limited.

In conclusion, the Von Neumann architecture is a common computer

architecture contains four subsystems. Compared to the other computer

architecture, it is much simpler to design, so it is used commonly in our desktops and

laptops.

Cited Work

Figure 1:

http://www.itportal.in/2012/07/difference-von-neumann-and-harvard-architecture.htm

Figure 2:

https://www.computerhope.com/jargon/m/machcycl.htm

You might also like