4.1.1 Stored Programs and the Von Neumann Model
A modern computer can perform many different tasks without being rebuilt. The key idea is that the instructions controlling the machine are stored in memory and can be replaced by another program.
By the end of this section, you should be able to:
- Identify the processor, memory and input/output as the main parts of a simple Von Neumann system.
- Explain the stored-program concept in your own words.
- Explain how program instructions and working data can occupy the same memory.
- Describe how a processor normally works through instructions in sequence.
Changing the Task Without Rebuilding the Machine
Imagine a weather station that records a temperature every ten minutes. Later, its owner wants it to report only unusual readings. A hard-wired machine might require physical changes to its circuitry. A stored-program computer can instead be given a different sequence of instructions.
The hardware remains largely the same. What changes is the program held in memory. This is why one computer can run a spreadsheet, a simulation, a game or a control program at different times.
The Von Neumann Model
The Von Neumann model is a simplified way of describing a general-purpose computer. It focuses on how the processor, memory and input/output work together rather than on the physical layout of a particular device.
| Part of the model | Contribution to the system |
|---|---|
| Processor (CPU) | Retrieves instructions, interprets them and carries out the required operations. |
| Memory | Holds the current program and the data being used by that program. |
| Input | Supplies data or commands to the computer system. |
| Output | Communicates processed information or controls another device. |
Registers, buses and the detailed instruction cycle make this model work in practice. They are introduced in the following Unit 4 sections; this page concentrates on the underlying stored-program idea.
Working Through a Program
A stored program is made from separate instructions. In the simplest case, the processor starts at the first instruction and then continues with the next memory address.
Normal pattern:
fetch one instruction → interpret it → carry it out → move to the next instruction → repeat
This is called sequential execution. It describes the normal flow, not an unbreakable rule. A jump or branch instruction can direct the processor to continue from a different address, allowing selection and repetition.
Common Mistakes
- “Stored program” means permanently fixed. No—the important advantage is that another program can be loaded.
- Sequential execution means jumps are impossible. No—sequential order is the default, but control instructions can change it.
- Only data is stored in memory. In this model, memory holds both data and program instructions.
- The model is a picture of one exact processor. It is an abstract organisation that can be implemented in many ways.
Practice
- A ticket machine is updated so that it accepts a new type of travel card. Explain how the stored-program concept allows this change without replacing the processor.
- In the memory example above, state the value eventually stored at address 742 and explain how it is produced.
- Explain two features of the Von Neumann model.
- A student writes: “A Von Neumann processor must always execute every instruction in consecutive address order.” Explain why this statement is incomplete.
- Distinguish between a program instruction and program data, while also explaining what they have in common when held in memory.
Check your answers
- A revised sequence of instructions can be loaded into memory. The processor then fetches and executes the new program using the same general-purpose hardware.
- The value is 32: the program loads 23, adds 9 and stores the result.
- Suitable points include: one processor accesses memory; memory holds instructions and data; instructions are normally processed in sequence; input and output connect the system to its environment.
- Consecutive order is normal, but a jump or branch can replace the next address and redirect execution.
- An instruction encodes an operation; data is a value used or produced by the program. Both are represented as binary patterns in memory.
Review
| Idea | What you should remember |
|---|---|
| Stored program | Instructions are held in memory and can be replaced. |
| Shared memory | The same memory system can contain instructions and data. |
| Processor | Fetches, interprets and carries out stored instructions. |
| Sequential execution | Instructions normally follow address order unless control flow changes it. |
| General-purpose computer | The machine changes task by loading another program rather than being rewired. |