When deciding on the RAM for your server, you will see suggestions based on what the server is intended for, which help set the bottom threshold. However, going bigger is always recommended: the more RAM, the better performance. Let’s take a closer look at why this is the case.

The servers need plenty of RAM because fast and efficient performance requires more capacity the larger your database is. Larger RAM is needed for a larger number of users and dynamic content. The OS size and expected growth rate should be counted as well.

1. A Faster Server Needs Faster RAM

RAM (random access memory) is a volatile memory structure that can only hold information temporarily, unlike HDD (hard disk drive). Its key function is optimizing processing operations. In the case of a server, that includes all the apps that run on it and all the stored and retrieved data.

RAM size determines how much data it can hold simultaneously. When RAM runs out of space, it must remove older data and return to the HDD to acquire new files.

The smaller the memory capacity, the more often your RAM will be forced to delete data and access it from the HDD whenever you need it again.

Larger capacity, on the other hand, can significantly decrease the number of required cycles. The more data can be stored in RAM, the less it will have to reach out to the HDD, allowing faster speed.

That’s the general idea of how RAM speed works. You may sometimes see it being specified as frequency and measured in MHz. Since this parameter is highly important for RAM’s overall performance and efficiency, you will easily find it in the product description.

However, frequency is not exactly the RAM speed but one factor contributing to it. Put shortly, it measures how many reading-writing cycles your RAM can perform per second. These cycles, also called clock cycles, determine how much data can be stored and how fast it can be read and processed.

Another factor is latency, or CAS latency (Column Access Strobe latency). It is typically specified in numbers that show how many clock cycles it takes for the RAM to process and execute a command received from the CPU (central processing numbers). Here, smaller numbers mean faster speed.

In reality, RAM speed is a little more complicated than it seems. It is determined by capacity, frequency, and CAS latency, which give different results in different combinations. However, sufficient RAM capacity remains the basis of smooth and fast RAM performance.

2. Larger Databases Need Larger RAM Capacity

RAM size is tied to the size of a database. The larger it is, the more capacity is required to process its information effectively. This factor is probably the most crucial when settling on the optimum size.

What the server will be used for also plays a role here. Recommended RAM size is usually stated in the percentage of the database size, yet other factors can weigh in, causing a shift in numbers.

To give you an idea, here are the general recommendations for SQL servers to use as a reference:

Database Size	RAM Size
10-59 GB Database	74%
60-224 GB Database	23%
225-600 GB Database	13%
600 or more GB Database	6%

You can see a pattern here that may surprise you: as the databases grow, the percentage of data size that determines the required RAM capacity drops. Indeed, when we look at larger databases, we see little change in memory size.

Part of the reason is that other factors contribute to the minimum RAM besides the size of the database, including the space reserved for the OS. As a database’s size expands, the required capacity only grows so much.

3. More Users Increase the Load on the RAM

The number of users is another factor that can call for bigger RAM. The more people are expected to use the server at a given time, the greater will be the load on your RAM. To ensure it can be safely handled, you must provide enough for processing user requests.

However, this factor won’t play much of a role for small businesses and servers with local purposes. If you expect up to 15 people to be engaged simultaneously and can safely say you won’t need more capacity to cover that, the number of users won’t impact the RAM size.

On the other hand, if your scope is larger, you must count that in when determining the RAM size you need. A rule of thumb is that 1-2 GB should be added to the recommended size for every 5 additional users.

This rule works when we talk about tens or hundreds of users. If the project is intended for an even bigger audience, you may need significantly more, but the proportion of GB of RAM per user decreases. You will likely require more than one server for hundreds of thousands of users.

Unlike in previous instances, the correlation here is straightforward: the more people are expected to put a load on the server, the more RAM is needed, which can result in significant growth of the required capacity.

4. Dynamic Content Needs More RAM

The types of content a server deals with also influence the required RAM size. More precisely, if a lot of dynamic content is involved, the minimum capacity can be much larger than needed to load static content.

Let’s look at these two content types and determine their key differences to see how each impacts the required RAM.

As the name suggests, static content requires no changes before sending it to a user. All the server has to do is find the file and deliver it upon request without any additional operations involved. A website article, a YouTube video, or a picture are just a few examples of static content.

Put shortly, static content is stored in its final form and displayed similarly. It’s simple in processing, so it loads faster and isn’t demanding in terms of RAM capacity.

Dynamic content, on the other hand, is more complicated. Rather than being secured in one form and displayed to each user similarly, it changes with each request depending on its conditions. Messengers, ads, and personalized recommendations represent this content type.

Because of its higher level of complexity, dynamic content puts a heavier load on the server. Instead of just pulling the necessary file, the server needs to access various files and generate a personalized response.

The two content types often co-exist on the same page. For instance, if you’re viewing a YouTube video, you’re accessing static content. However, the constantly changing comment section below is dynamic.

The more dynamic content is displayed, the more RAM capacity the server requires for decent performance. Because dynamic content is becoming increasingly widespread, the RAM size requirements for most projects are also higher.

You will mostly deal with static content for simple home servers, so this won’t be an issue.

5. You Have to Leave Space for the OS

Don’t forget that a part of the required RAM size is reserved for the system files. However, how much space is needed for the OS depends on the overall RAM size.

The general recommendations for the OS RAM capacity are as follows:

1 GB of default space.
1 additional GB for every 4 GB up to 16 GB of RAM.
1 additional GB for every 8 GB above 16 GB of RAM.

Here’s an example to help you understand how these should be calculated. Let’s say you have 32 GB RAM. Here’s how you determine the size required for the OS:

We’re starting with the default capacity of 1 GB.
For the first 16 GB of RAM, we’re throwing in 4 more GB for the OS, one for each 4 GB of RAM.
Now, we have 16 more GB to cover above the initial 16 GB. We need 1 GB for the system per 8 GB of RAM, which makes a total of 2 GB added.
Finally, we get 1 + 4 + 2 = 7 GB of RAM reserved for the OS.

These calculations determine how much space the OS will take approximately, allowing you to see how much is left for the data. Most software, however, will provide this information to you, often specifying the minimum requirements and recommended optimum RAM capacity for the OS.

Sometimes, the software will also provide the value for the maximum RAM size it can handle.

I’ve mentioned that, generally, the more RAM, the better, and it’s beneficial for the performance of your server to go beyond the recommended minimum.

However, some software can pose limitations to how far beyond you can actually go. If so, it will be specified in its description and requirements, so study them carefully beforehand.

6. Servers Need to Account for Growth Possibilities

Finally, it’s always best to plan ahead and consider your growth expectations. Sometimes, the RAM size recommendations can seem excessive. However, it’s advised to include more capacity than you need to ensure the server can support your growth rate.

The required RAM size should be calculated individually based on the anticipated database and user growth. To get an accurate estimate, consult IT specialists and request an assessment of your prospects and the suggested optimum RAM capacity suitable for the potential growth rate.

If this factor is overlooked, multiple issues are likely to arise in the future. Upgrading the memory size requires taking the server down for maintenance, which can interfere with your work. In addition, the unavoidable upgrades will bring extra financial costs.