
Random Access Memory (RAM) is the most recognized type of computer memory. It allows your computer to quickly browse the web and seamlessly transition between tasks like loading applications or editing documents. RAM is termed 'random access' because you can access any memory cell directly if you know the row and column where it is located.
In contrast, serial access memory (SAM) stores data in a series of memory cells that can only be accessed in a specific sequence (similar to a cassette tape). If the desired data isn't at the current location, each memory cell is examined one by one until the data is found. SAM is ideal for memory buffers, where data is typically stored in the order it will be accessed (such as texture buffer memory on a video card). RAM, on the other hand, allows data to be accessed in any sequence.
RAM functions as your computer's temporary memory. Like a microprocessor, a memory chip is an integrated circuit (IC) containing millions of transistors and capacitors. The most widely used form of computer memory, dynamic random access memory (DRAM), pairs a transistor with a capacitor to form a memory cell, which stores a single bit of data. The capacitor holds the data as either a 0 or a 1 (see How Bits and Bytes Work for more details). The transistor serves as a switch, allowing the control circuitry to either read or alter the state of the capacitor.
A capacitor functions like a tiny bucket that can store electrons. When storing a 1, the bucket is filled with electrons, while a 0 is stored when it is emptied. However, the capacitor's bucket has a leak — within a few milliseconds, a full bucket can become empty. To keep dynamic memory functioning, either the CPU or the memory controller must periodically recharge the capacitors holding a 1 before they discharge. This process involves reading the memory and immediately writing it back. The refresh operation occurs thousands of times per second automatically.
The capacitor in a dynamic RAM memory cell is akin to a leaky bucket. It needs constant refreshing or it will discharge to 0. This is the essence of dynamic RAM: it requires regular refresh cycles, or it loses the data it's holding. The downside of this constant refreshing is that it takes time and can slow down the memory's overall performance.
This article will guide you through everything you need to know about RAM — the types to consider, how to select the right one, and how to install it.
Memory Cells and DRAM
Memory consists of bits arranged in a grid-like structure, forming a two-dimensional layout.
In this illustration, red cells represent 1s, while white cells represent 0s. The animation shows how a column is selected, and the rows are charged to store data into the specific column.
Memory cells are engraved on a silicon wafer, organized in columns (bitlines) and rows (wordlines). The intersection of a bitline and a wordline defines the address of the memory cell.
DRAM operates by sending a charge through the relevant column (CAS) to activate the transistor at each bit. During a write operation, the row lines contain the value that the capacitor should store. When reading, the sense-amplifier detects the charge level in the capacitor. If it's over 50 percent, the read value is 1; otherwise, it's 0. A counter tracks the refresh sequence based on which rows are accessed and in what order. This entire process occurs in a time frame so short it's measured in nanoseconds (billionths of a second). A memory chip with a 70ns rating indicates that it takes 70 nanoseconds to completely read and recharge each cell.
Memory cells alone are insufficient without the support of additional specialized circuits to manage data flow. These circuits handle various tasks such as:
- Determining the location of each row and column (row address select and column address select)
- Tracking the refresh cycle (counter)
- Reading and restoring the data signal from a memory cell (sense amplifier)
- Controlling whether a cell should receive a charge (write enable)
Additional responsibilities of the memory controller involve tasks such as determining the memory type, speed, capacity, and performing error checking.
Static RAM operates on a different principle compared to DRAM. We'll explore this in the following section.
Static RAM
Static RAM utilizes a completely different technology. In static RAM, each bit of memory is stored in a type of flip-flop (see How Boolean Logic Works for further information on flip-flops). A flip-flop for a memory cell consists of four to six transistors and some wiring, but it doesn't require refreshing. This gives static RAM a significant speed advantage over dynamic RAM. However, the extra components mean that a static memory cell occupies much more space on a chip compared to a dynamic memory cell. As a result, less memory can be fitted onto a chip, which increases its cost.
Static RAM is fast but comes at a high cost, whereas dynamic RAM is more affordable but slower. As a result, static RAM is typically used to create the CPU's cache, which is sensitive to speed, while dynamic RAM serves as the larger system memory.
Memory chips in desktop computers initially used a pin configuration known as dual inline package (DIP). This configuration could either be soldered into holes on the computer's motherboard or inserted into a socket soldered to the motherboard. While this method worked for computers that typically had a few megabytes or less of RAM, the increasing demand for memory led to a higher number of chips needing space on the motherboard.
The solution was to place the memory chips and all related components onto a separate printed circuit board (PCB), which could then be inserted into a special connector (memory bank) on the motherboard. Most of these chips use a small outline J-lead (SOJ) pin configuration, though many manufacturers also use the thin small outline package (TSOP) configuration. The major difference between these newer pin types and the original DIP configuration is that SOJ and TSOP chips are surface-mounted to the PCB, meaning the pins are directly soldered onto the board's surface rather than inserted into holes or sockets.
Memory chips are typically found as part of a card called a module. When shopping for memory, you can often spot the individual memory chips on many of the modules.
In the next section, we'll explore some other common types of RAM.
Different RAM Types

Here are some common RAM types you might encounter:
- SRAM: Static RAM uses several transistors—usually four to six—for each memory cell but doesn’t require a capacitor. It's mainly used in cache memory.
- DRAM: Dynamic RAM has memory cells made up of paired transistors and capacitors, requiring frequent refreshing.
- FPM DRAM: Fast Page Mode DRAM is the original form of DRAM, processing data bit by bit. It has a maximum transfer rate of 176 Mbps to L2 cache.
- EDO DRAM: Extended Data-Out DRAM begins processing the next bit as soon as the first one’s address is located, making it faster than FPM DRAM by about 5-20%. It has a transfer rate of 264 Mbps to L2 cache.
- SDRAM: Synchronous DRAM uses burst mode to speed up performance by staying on the row with the requested data. It's 5% faster than EDO RAM and operates at transfer rates of 0.8-1.3 MT/s, developed in 1988.
- DDR SDRAM: This next-generation SDRAM has higher bandwidth and faster speeds than regular SDRAM. Released in 2000, it offers speeds of 2.1-3.2 MT/s. DDR has evolved into DDR2, DDR3, and DDR4, with DDR4 reaching speeds of 17-21.3 MT/s. JEDEC, the industry group, set these standards, with DDR5 released in 2020, offering a potential 87% performance boost over DDR4.
- RDRAM: Rambus DRAM is a distinct architecture designed by Rambus. It uses a unique high-speed data bus and operates at speeds of 800 MHz (1,600 Mbps) or more. These chips require heat spreaders due to their high temperatures. There are versions designed for notebooks called SO-RIMMs.
- Credit Card Memory: This is a proprietary DRAM module for notebooks, designed to plug into a dedicated slot.
- PCMCIA Memory Card: Another self-contained DRAM module for notebooks, these cards are not proprietary and work with any compatible system bus. They are less common today.
- CMOS RAM: CMOS RAM is used by computers and devices to store essential settings, like hard disk configurations, using a small battery to maintain data.
- VRAM: Video RAM, or Multiport DRAM (MPDRAM), is used in video adapters and 3D graphics accelerators. It has two access ports, allowing simultaneous access by the CPU and GPU. VRAM is vital for display resolution, color depth, and graphics processing, while SGRAM is a cheaper alternative with similar performance.
RAM Modules

The types of boards and connectors used for RAM in desktop computers have significantly changed over the years. Initially, these were proprietary, meaning only specific systems could use the memory boards developed by different manufacturers for their computers.
Next, came the introduction of SIMM, or Single In-Line Memory Module. These memory boards had a 30-pin connector and measured approximately x 0.75 inches (around 9 x 2 cm). Typically, SIMMs had to be installed in pairs with matching capacity and speed since the bus width exceeded the capacity of a single SIMM module.
For instance, to reach a total of 16 megabytes of RAM, you would install two 8MB SIMMs. Each SIMM could transfer 8 bits of data at a time, while the system bus could handle 16 bits simultaneously. Later SIMM boards, measuring 4.25 x 1 inch (around 11 x 2.5 cm), incorporated a 72-pin connector to improve bandwidth, allowing up to 256MB of RAM. SIMMs were widely used from the early 1980s through the early 2000s.
As processor speeds and bandwidth capabilities improved, the industry adopted the dual in-line memory module (DIMM) as a new standard. DIMMs come in various capacities and can be installed individually, rather than requiring pairs like previous memory types.
Certain laptop computers use RAM based on the small outline dual in-line memory module (SODIMM) configuration. These compact SODIMM cards are about 2 x 1 inch (5 x 2.5 cm) in size and feature either 144 or 200 pins. The capacity of these modules ranges from 2GB to 32GB. Some sub-notebooks use even smaller DIMMs called MicroDIMMs. The trend in the industry has been toward low-power DDR4 modules for thinner, lighter laptops because they consume less energy and occupy less space. However, these modules must be soldered directly into the laptop, making them non-replaceable by most users.
Memory available today is generally very reliable. Most systems simply rely on the memory controller to check for errors during startup. Memory chips with built-in error detection often use a method called parity, which adds an extra bit for every 8 bits of data. Let's break down how parity works, starting with even parity.
When data is written into the 8 bits of a byte, the memory chip calculates the total number of 1s. If the total number of 1s is odd, the parity bit is set to 1. If the total is even, the parity bit is set to 0. When the data is read from the memory, the chip recalculates the total number of 1s and compares it to the parity bit. If the total is odd and the parity bit is 1, the data is considered valid and sent to the CPU. However, if the total is odd and the parity bit is 0, an error is detected, and the data is discarded. Odd parity works in a similar manner, except the parity bit is set to 1 when the number of 1s is even in the byte.
The issue with parity is that while it can identify errors, it doesn't have the ability to correct them. If the data in a byte doesn't match its parity bit, the data is discarded and the system will try again. For computers used in critical environments, a higher level of fault tolerance is required. Many high-end servers employ error-correction code (ECC), which, like parity, adds extra bits to monitor each byte of data. The key difference is that ECC uses several bits for error checking, depending on the bus width, instead of just one. ECC memory uses a special algorithm to detect and correct single-bit errors, and can also detect when multiple bits fail, though such failures are rare and cannot be corrected, even with ECC.
Most computers today come equipped with nonparity memory chips. These chips don't have built-in error-checking mechanisms and rely on the memory controller to detect errors.
How Much RAM Is Necessary?
It's often said that you can never have too much money, and the same could be said about RAM, particularly if you're involved in graphic-heavy tasks or gaming. After the CPU, RAM plays the most significant role in computer performance. If your system lacks sufficient RAM, adding more could make a bigger impact than upgrading to a new CPU!
If your computer feels sluggish or frequently accesses the hard drive, it's time to upgrade your RAM. For Windows 10 users, Microsoft recommends 1GB for the 32-bit version and 2GB for the 64-bit version. If you're moving to Windows 11, 4GB is the minimum requirement. For Mac users running MacOS 11 (Big Sur), 4GB is also required.
Linux is well-known for its ability to run efficiently on machines with modest system resources, including RAM. Xubuntu, a popular Linux distribution designed for low resource usage, requires just 512MB of RAM. It uses the lightweight Xfce desktop environment, which is also compatible with other Linux distributions. Naturally, some Linux distributions demand higher system specifications.
Regardless of the operating system you use, keep in mind that minimum system requirements are designed for typical usage — browsing the web, word processing, standard home/office tasks, and light entertainment. If you engage in more demanding activities, such as computer-aided design (CAD), 3D modeling/animation, heavy data processing, or serious gaming, more RAM will be required. Additional RAM may also be necessary if your computer is functioning as a server (for hosting webpages, databases, applications, FTP, or networks).
Another factor to consider is how much VRAM your video card should have. Most modern video cards come equipped with at least 12-16MB of VRAM, which is usually sufficient for standard office work. However, if you plan to do any of the following, you should consider investing in a more powerful graphics card:
- Playing high-quality, realistic games
- Video capture and editing
- Creating 3D graphics
- Working in a high-resolution, full-color setting
- Designing detailed, full-color illustrations
When purchasing a video card, make sure that both your computer and monitor are compatible with the card you intend to buy.
Steps to Install RAM

Generally speaking, adding RAM is a simple and easy process. However, it's important to do your homework. Here’s what you need to check before getting started:
- How much RAM is already installed
- How much additional RAM you want to add
- Form factor of the RAM
- Type of RAM required
- Any necessary tools
- Warranty considerations
- Where to install the RAM
RAM is usually sold in densities with multiples of 2 gigabytes: 2, 4, 8, 16, 32. In other words, the module is the same standard size, but there can be different amounts of memory on the same board. For example, if your computer has 8GB and you want 16GB total RAM, you would want to buy a module with an 8GB density.
Once you know how much RAM you want, check to see what form factor (card type) you need to buy. You can find this in the manual that came with your computer or you can contact the manufacturer. It's important to realize that your options depend on the design of your computer. Most computers sold for normal home/office use have DIMM slots. High-end systems are moving to RIMM technology, which will eventually take over in standard desktop computers as well. Since DIMM and RIMM slots look a lot alike, be very careful to make sure you know which type your computer uses. Putting the wrong type of card in a slot can cause damage to your system and ruin the card.
You will also need to know what type of RAM is required. Some computers require very specific types of RAM to operate. For example, your computer may only work with 60ns-70ns parity EDO RAM. Most computers are not quite that restrictive, but they do have limitations. For optimal performance, the RAM you add to your computer must also match the existing RAM in speed, parity and type.
Additionally, some computers support dual-channel RAM configuration either as an option or as a requirement. Dual-channel RAM modules are installed in matched pairs, so if there is a 512MB RAM card installed, there is another 512MB card installed next to it. When dual channel is an optional configuration, installing RAM in matched pairs speeds up the performance of certain applications.
Your computer is only configured to accept so much memory. There is a limited number of memory slots, and depending on your machine, you may be limited to an 8GB density module even if the manufacturer makes a 16 or 32GB module. Or, in some cases, your computer may allow you to upgrade the RAM that was installed in the factory. If you had a machine that came with 4GB of replaceable RAM, but that machine could accept 16GB, you could buy two 8GB modules and swap out the 4GB module.
Many computer and memory manufacturers provide a helpful tool on their websites. By entering your computer's model, you can easily find out what type of memory you need to install. Check the system settings on your computer to see how much memory is already installed. Once you're aware of how many slots are available and the maximum capacity your machine can handle, you can choose how much additional memory to purchase. Some manufacturers solder the base memory in place, while others may allow you to swap out smaller RAM cards for larger ones.
Being aware of your computer’s configuration before buying memory is crucial to prevent unnecessary frustration. Realizing that the memory you purchased doesn’t fit or isn’t compatible after you've opened your computer can be incredibly frustrating.
Before opening your computer, double-check the end-user license agreement to ensure you won’t void the warranty by doing so. Some manufacturers seal the computer case and require that a certified technician installs the RAM. If you're comfortable with opening the case, first power down and unplug your computer. Ground yourself using an anti-static wrist strap or pad to discharge any static electricity. Depending on your machine, you may need a screwdriver or nut-driver to access the case. Some desktop systems have tool-less cases that use thumbscrews or simple latches. Laptops, on the other hand, can be trickier to open.
Installing the memory module typically doesn't require any special tools. The RAM fits into slots on the motherboard known as the memory bank. The memory module has a notch on one end, which prevents incorrect installation direction.
For SIMMs and some DIMMs, you start by placing the module at about a 45-degree angle into the slot, then gently press it forward until it is vertical to the motherboard. The small metal clips at each end should click into place. If the clips don’t lock, verify that the notch is in the correct position and ensure the card is firmly seated. Many DIMMs don’t have metal clips and instead rely on friction to keep them in place. Once again, ensure the module is properly seated in the slot. For more detailed instructions, refer to your motherboard's manual.
After the module is installed, close the computer case, reconnect the power, and turn the system on. During the POST ("Power On Self Test"), the computer should automatically detect the new memory, although it might require a couple of reboots for the system to fully register it. That’s all there is to it!