The number of bits used in addressing memory is the size of the memory space or address space. This is a linear range of addresses that we can put things in. The number of bits in each address governs how many possible locations you can access; for example, a 16-bit memory space would have locations (from to , and any addresses above this simply do not exist. This is why you can’t add RAM; you simply can’t access it!
”Addresses” refer to abstractions in terms of the smallest available unit of memory. Memory might be:
- Byte-addressable
- Word-addressable (4 bytes)
- Double word addressable (8 bytes)
For example, a word-addressable 8-bit memory space would mean there are locations, and each location has 8 bytes of memory, so there are a total of bytes of memory.
Smaller units
It’s possible to read or write smaller pieces of memory than the fundamental memory unit – you would simply read in the whole unit, then use CPU instructions to read or modify the appropriate bits – it just takes extra steps and therefore extra time.
It is predicted that memory will end at 64-bit memory spaces that modern computers use, as a byte-addressable 64-bit memory space would have 17 billion gigabytes. This is a lot of memory for one computer and would need hundreds of years to be searched by a modern computer.