RAM (Random Access Memory): what it is, how it works and what it does

Definition

RAM (Random Access Memory) is a type of volatile working memory that temporarily holds the data and instructions the processor needs to execute at any given moment. Unlike non-volatile memory (such as NAND flash or disk), its contents are lost when electrical power is removed.

The term «random access» indicates that any memory cell can be read or written in a time that is essentially independent of its physical position, unlike sequential media such as magnetic tape.

How It Works

When a system boots, the firmware loads the operating system and active processes into RAM. From that point on, the processor reads from and writes to it at speeds far exceeding those of mass storage. Each access cycle involves:

1. The memory controller receives an address.
2. The corresponding cell is enabled (row and column selected).
3. The data is read or written.

When the device is powered off, the energy sustaining each cell's state disappears and the information is lost; hence the term volatile memory.

Architecture and Principles

Main Types

• DRAM (Dynamic RAM): each cell consists of a transistor and a capacitor. The capacitor stores the electric charge representing the bit; because it leaks over time, cells must be refreshed periodically. This is the dominant technology in main memory for computers and phones.
• SRAM (Static RAM): each cell uses several transistors (typically six) in a flip-flop circuit. It requires no refresh, is faster but more expensive and occupies more area per bit. It is used in processor caches and high-speed embedded applications.

DRAM Cell Structure

A typical DRAM cell comprises an access transistor and a storage capacitor. The transistor acts as a switch connecting the capacitor to a bit line when the corresponding row (word line) is activated. The capacitor, with a capacitance on the order of tens of femtofarads, stores a charge interpreted as 0 or 1 according to a voltage threshold.

Applications

• General computing: main memory in desktops, laptops, and servers.
• Mobile devices: in smartphones and tablets, RAM determines how many applications can remain active in the background. A common value in the mid-to-high range is 8 GB, as seen in consumer device specifications (for example, the Google Pixel 10a is marketed with 8 GB of RAM and 256 GB of storage).
• Embedded and industrial: controllers, routers, medical equipment, and automotive systems use smaller-capacity RAM modules with specific reliability profiles.
• High-performance computing: cluster nodes and supercomputers employ hundreds or thousands of gigabytes of RAM to keep datasets in memory during intensive calculations.

Scope

RAM is present in virtually every electronic device that executes software: from microcontrollers with a few kilobytes of SRAM to data centres with petabytes of DRAM. Its capacity and speed are design parameters that condition the overall system performance.

Limitations

• Volatility: total data loss on power-off requires persistent state to be stored in non-volatile storage.
• Cost per bit: DRAM is significantly more expensive per gigabyte than NAND flash, limiting installed capacity in consumer devices.
• Refresh: DRAM consumes continuous power to maintain cell states, impacting battery life in portable devices.
• Latency and bandwidth: although access is «random», DRAM cell latency (typically tens of nanoseconds) is far higher than L1/L2 processor cache, creating bottlenecks if the memory hierarchy is not well managed.

Interpreting Its Presence in a Product

When a technical specification states a value such as 8 GB RAM, it is specifying the capacity of the installed working memory. This figure allows the user to estimate:

• The number of applications that can coexist actively without the system recycling memory.
• The capacity for multitasking and running demanding workloads (video editing, compilation, light virtualisation).

It is important not to confuse RAM with data storage (256 GB in the cited example): the former is volatile working memory; the latter is persistent space for files, applications, and the system. Both values complement each other and are typically listed together in a device's specifications.

In summary, RAM is the processor's «workbench»: the larger and faster it is, the more tasks it can handle simultaneously without resorting to slower storage.