Understanding computer memory is very important because memory is what enables your PC to operate. Sure, your computer's CPU (central processing unit) is the computer's brain, but a brain is useless without memory.
Because memory is one of the most important parts of a computer, it has been redesigned and tweaked often over the years to enable it to store more data and work faster. This has resulted in a myriad of technical terms and contractions.
In this article, I'll try to clear up a few of these terms and give you a basic understanding of computer memory.
When you work with an application on your computer, the program is loaded into "main memory". Of course there are several other kinds of memory besides main memory. For example, there is "cache" and "flash BIOS" memory. In fact, a computer consists primarily of memory. Even the CPU has a type of memory referred to as "registers". But main memory is the focus of this article.
Memory is quite simply an array of "cells". Each cell holds one binary data bit, either 0 or 1. A row of 32 cells store a "word". A word may contain the bits to code one single number or one single character. Each word can be accessed by it's own unique "memory address". A computer doesn't have to access the addresses in any particular order. It can store or retrieve a word from any memory address at random, hence the term "random access memory" (RAM).
Because a computer needs millions of memory cells in order to operate efficiently, the cells have to be cheap. The cheapest memory cell of all is a very tiny capacitor. A "0" value would be represented by an absence of electric charge stored in the capacitor. A "1" is represented by an electric charge being stored in the cell.
A tiny capacitor can store an electric charge, but that charge dies off within a few milliseconds. To prevent data loss, the computer uses special circuitry that repeatedly scans all the memory cells and refreshes the charge in any cells that contain a "1". This circuitry must operate between accesses of the memory by the CPU and other components. This continuous recharging is known as "dynamic" refresh, hence the term "dynamic" RAM (DRAM).
To understand the way memory works, you need to know what a "bus" is. The CPU needs wires to connect to memory cells. But instead of wires, a circuit board uses copper "traces" inside or on the surface of the circuit board. It takes 32 parallel traces to access one word of data. We also need a few extra traces to carry the signals that control whether the CPU is reading from or writing to the memory. This set of parallel traces make up the "data bus".
In order to access a word of data, the CPU needs to provide the memory address of the word. The traces that perform this function make up the "address bus". The address bus requires 64 parallel traces for the address information.
There are also several other parallel groups of traces on the circuit board that handle various communication and control signals. Each of theses groups is referred to as a "bus". The most important signal racing around the system board is known as the "system clock signal".
The system clock provides the "pulse" of the computer system. It's a continuous stream of electrical pulses at a very high frequency. If you have a 1.8 GHz computer system, then your system clock pulses come at a rate of 1,800,000,000 per second. Memory can be accessed on the rising edge of the system clock pulse. Memory reading and writing is synchronized with the system clock, hence the term "synchronous" DRAM (SDRAM).
SDRAM is available in a 168-pin dual inline memory module (DIMM) package. It comes in two speeds: 100 MHz (PC100) and 133 MHz (PC133) and two voltages: 3.3 volt and 5 volts. Some SDRAM "sticks" contain a bit of extra circuitry called "error checking and correction" (ECC) that stores extra bits along with the data. These extra bits are used by the CPU to verify the validity of the data and correct any errors that it finds.
New systems use a type of memory that can access data on the rising and falling edges of the system clock pulses. This means your computer can access the data in memory at twice the normal speed, hence the term "Double-Data-Rate SDRAM (DDR-SDRAM).
DDR-SDRAM is available in a 168-pin DIMM. It comes in two speeds: 266 MHz (PC2100) and 333 MHz (PC2700). If your system board has more than one memory slot, and you have only one memory module, you should install the single DIMM in slot 1.
Some system boards made by Intel use Rambus DRAM (RDRAM). RDRAM uses a proprietary communications channel to access the memory cells. RAMBUS is available in a 184-pin RAMBUS In-line Memory Module (RIMM). There are two connectors on an Intel Rambus system board. Each connector must contain either a RIMM or a C-RIMM. A C-RIMM (continuity RIMM) does not contain memory. It's simply a module that provides a pass-through path for the signals.
Modern computer systems use DDR memory, which stands for
"Double Data Rate". DDR technology has improved over the
years, primarily in terms of speed. The DDR technology being
used as of this writing is DDR3.
If your PC is older, its main memory may use "fast page memory" (FPM) or "extended data output" (EDO) memory. Some of these memory sticks used a special "parity bit" to validate the data. These older types of memory are packaged in either a 30 pin single inline memory module (SIMM) or a 72 pin SIMM.
If you want to upgrade your PC's memory, be sure to use the correct type of memory for your system board. Refer to the manual that came with your system board to determine exactly what type and speed of memory is required for your computer.
Important: Memory modules of different types are typically
NOT interchangeable. When upgrading or adding memory to your
computer, be absolutely certain to buy the type of memory
stick(s) that your motherboard is designed to use.
Installing the wrong kind of memory can damage the memory
modules themselves, the motherboard, or both.
Stephen Bucaro is the owner of Bucaro TecHelp.
Visit him at: http://bucarotechelp.com
More Interesting Articles