Course contentsShow
← Computer Science
Lesson 1335 of 2,87231. Memory ManagementFree lesson

Contiguous Memory Allocation

Examine the early approach where each process occupies a single contiguous block of physical memory.

Contiguous Memory Allocation

You'll learn: How early operating systems gave each process a single, unbroken block of physical memory.


In the early days of memory management, operating systems used a straightforward approach: when a process needed memory, the OS allocated one continuous chunk of physical RAM from start to finish. Think of it like renting a hotel room block—you get rooms 301 through 310 all in a row, not scattered across different floors.

How It Works

With contiguous memory allocation, each process occupies consecutive memory addresses in physical RAM. When the OS loads a program, it finds a "hole" (free region) large enough to hold the entire process and places it there. The process runs entirely within that fixed, adjacent range of addresses.

The base and limit registers you learned earlier enforce this boundary: the base points to the starting address, and the limit defines how far the process can reach. Every address the process generates is checked against these bounds.

The Simplicity and the Problem

This model is beautifully simple—no complex mapping tables, just straightforward arithmetic. Finding where a process lives means knowing its base address.

However, this approach creates external fragmentation: as processes start and terminate, memory becomes a patchwork of occupied and free regions. You might have 100 MB free total, but scattered across ten 10 MB holes—none big enough for a process needing 50 MB continuously.

Another issue: processes must fit entirely in memory. If your available contiguous space is smaller than the process size, you're stuck, even if total free memory is sufficient.

Key Takeaway: Contiguous memory allocation gives each process a single, unbroken block of physical memory—simple and fast, but prone to fragmentation and inflexible sizing constraints.