We have the answers to your questions! - Don't miss our next open house about the data universe!

RAM or “Random Access Memory”: What is it?

 m de lecture

One of the specifications that determines "how strong" a new computer, smartphone, tablet or even video games console is going to be is its RAM memory. This type of memory is therefore one of the fundamental pillars of computing, and if you don't know exactly what it is and what it's for, we're going to try and explain it to you today.

What is RAM?

RAM is your device’s main memory and temporarily stores data from the programs you are currently using. Its acronym stands for Random Access Memory, and it is a type of memory that you can find in any device, from desktop computers to mobile phones.

RAM memory has a particular characteristic that sets it apart from other types of storage. Data is only stored temporarily. This means that every time you restart or switch off your computer, it is normal for the data stored in RAM to be lost.

Bearing in mind that RAM is used by the computer or mobile device to manage data from running applications, the amount you have directly affects the performance of your device. The more RAM you have, the more applications you can manage at the same time, which is why it’s so important, because if there isn’t enough, the computer starts to slow down.

What you usually call RAM memory when you talk about it as a physical component are cards that are connected directly to your computer’s motherboard. These cards have different integrated memory modules that are connected to each other, and in devices like computers, there are slots to hold several of these cards.

In this way, if you want your PC to have 32GB of RAM, you can have two cards, so that if one fails, you can continue to use the other until you have a replacement. This applies to desktops and laptops, as devices such as mobile phones generally do not allow you to change their components.

What types of RAM are there?

There are two types of RAM memory. DDR (Double Data Rate) memory is characterised by its ability to perform two operations per clock cycle, unlike SDR (Single Data Rate) memory, which performs only one read or write operation. To make this possible, DDR chips are activated twice for each clock cycle, either by level (high or low) or by edge (high or low).

This mode of operation is the same in the DDR3 and DDR4 standards, as well as in the future DDR5 or its predecessors. However, the more modern standards you use, the faster it will work. This means that DDR4, which has a data rate of 3.2 GB/s and a maximum transfer rate of 25.6 GB/s, is not the same as DDR5, which has between 6.4 GB/s and 51.2 GB/s.

As a result, this new generation of RAM offers better performance than previous generations, as well as greater scalability and energy efficiency. So it may be worth upgrading if you want more performance. However, there’s a catch, and that’s that the standard of RAM you can use is defined by your device’s motherboard.

For example, if you have an older computer that had DDR3 RAM, it probably has RAM slots dedicated to that standard. Each has a different connector, so if you want to use more advanced RAM, you’ll need a different motherboard.

What is RAM used for?

Your computer or mobile phone does not carry out all actions using the hard disk alone, because if it did, it would take too long to execute them.

For this reason, a much faster type of memory is used to carry out these more immediate tasks, and is responsible for storing the processor instructions or data that applications constantly need.

These instructions remain there until the computer is switched off or until they are replaced by new ones.

To help you understand, let’s imagine you’re working at your desk. The drawers are your hard drive, where you store everything you can use. But you also need to put the tools and files you use on the desktop itself, and that’s what RAM would do in this case. So you don’t have to keep opening and closing drawers, you have everything in front of you.

What's the difference between RAM and ROM?

There are some key differences between ROM (read-only memory) chips and RAM (random access memory) chips. The difference lies in their use, storage capacity and function, as well as the physical size of the ROM and RAM chips.

A good example of ROM is the computer’s BIOS, which is a PROM chip that stores the programmes needed to start the computer’s initial boot process. Using non-volatile storage media is the only way for computers and other devices to start this process. ROM chips are also used in game system cartridges such as the original Nintendo, Gameboy, Sega Genesis and others.

ROM and RAM storage capacity

A ROM chip is a non-volatile storage medium, which means that it does not require a constant power source to retain the information stored on it. A RAM chip, on the other hand, is volatile, meaning that it loses all the information it contains when the power is cut off. Basically, ROM is used for permanent storage and RAM is used for temporary storage.

A ROM chip stores MB (megabytes) of data, typically 4 to 8 MB per chip, while a RAM chip can store GB (gigabytes) of data, from 1 to 256 GB per chip.

What is the minimum RAM size required for Data Science?

The more RAM you have, the more data it can handle, which means faster processing. You can use your machine to run other things like model trains with extra RAM. In terms of processing speed, 16GB and above of RAM is useful when cloud computing isn’t an option, because waiting for your algorithms to finish is a real headache.

By contrast, 32GB is harder to come by, although it does have a bigger advantage. Although it’s clearly faster and superior, most laptops don’t include it as an essential feature.

What’s more, most laptops don’t have 32GB of storage, but if yours does at the till, 32GB will greatly simplify the job of in-memory data science, especially when large data platform files aren’t available or accessible.

How much RAM do I need to process Data?

When we talk about Big Data, we mean massive data sets. As the name suggests, Big Data requires a lot of hardware to process large quantities of diverse data.

The data sizes we’re talking about require up to 1TB of RAM, where a data warehouse is needed to meet the processing demands.

As mentioned earlier, RAM is similar to your desktop in that it allows you to work on a variety of items, the bigger the desktop the more documents, files and tasks you can have at once without having to sort through them.

At least 16GB of RAM will be able to handle your big data needs on the computer, but what you really need to consider is at least 64GB of memory to handle serious large data block problems.

Hopefully this has been an informative read, feel free to check out our other blog posts.


DataScientest News

Sign up for our Newsletter to receive our guides, tutorials, events, and the latest news directly in your inbox.

You are not available?

Leave us your e-mail, so that we can send you your new articles when they are published!
icon newsletter


Get monthly insider insights from experts directly in your mailbox