PETRA : How does the computer remember
where it's stored the value for a certain variable?
REMI : To answer your question, we first need
to dig deeper into computer memory.
One role of the computer's operating system
is to manage the memory of the computer.
The term memory refers to the hardware in the computer that
is involved in storing information for use in the computer.
We distinguish between two different types of memory.
RAM, short for Random Access Memory, is temporary memory
that is easy and quick to access and, for example, used to execute programs.
We also call this type of storage volatile memory.
On the other hand, there is nonvolatile, or lasting memory
for more permanent storage of data,
for example, used to store files on the hard drive.
In the technical specification of a computer's hardware,
you can see the distinction between the amount of storage
available via the RAM and the available hard drive space.
PETRA : The vast majority of computer programs
uses the computer's RAM during execution.
Values of variables, for example, are stored in --
meaningful written to and read from --
Random Access Memory.
For simplicity, we may represent the Random Access Memory
as a sequence of binary memory cells, each populated with a 0 or a 1.
We call each such cell 1 bit.
By grouping several of these cells or bits together,
we create what is called a word.
A word is the fundamental unit of data that
can be moved between the RAM and the computer processor.
The size of a word is thus expressed in number of bits --
that is, number of memory cells --
or in bytes.
Recall that 1 byte equals 8 bits.
REMI : Modern personal computers and modern processors more
ordinarily tend to use 8, 16, 32, or even 64 bits
to form one word, though other word sizes are possible.
These particular word sizes emerge historically
as a result of the hardware architecture,
but they have evolved over time.
Petra, why would one want to group memory cells into words?
PETRA : Well, this allows for the addressing of memory.
In fact, an address is assigned to each word.
We also call this a memory address.
The computer address is a whole number that
describes the location of the word in the computer memory.
For example, imagine the computer whose word length is 8 bits.
Suppose, in addition, that this computer can store a total of four words.
It's not a very big computer.
One could then address the computer's memory
by assigning an address to each of the four words,
just like you would assign an address to each of four houses on a street.
Each word gets a number as its address.
REMI : Look here.
The address 0 would be used for the first word : that is,
the first 8 memory cells.
The address 1 would be used for the second word :
the next 8 memory cells.
The address 2 would be used for the third word,
and the address 3 would be used for the fourth word.
In the C programming language, it is possible to get these memory addresses
during the execution of a program.
PETRA : For example, if I use a variable in my program
that stores an integer value, then not only could I find and use
this value during my program, but equally well,
I could obtain the memory location - that is,
the address where the value is stored.
REMI : The use of memory addresses allows low-level access to the computer
by allowing direct access to the computer's memory.
One could literally move through the computer's memory from word to word.
PETRA : This technique allows for optimization of memory usage
at a very precise level, or of performance
in terms of execution speed.
This is an extremely powerful tool for application developers.
REMI : So now you know that each variable has its own address,
and that memory is composed of bits --
cells -- that are organized into words.