Published: Sat Jul 13 2024
Random access memory, often referred to as just ‘memory’, allows data to be accessed by the processor (CPU) as fast as possible, maximising performance. You can think of memory as the middle-man between your computer’s CPU and its storage devices. RAM is fast – faster than even a top-of-the-range solid state drive. When you open a program, it’s not loaded straight off the storage device. This would be too slow. Instead, essential data for running the program is temporarily transferred over to memory. This ensures that the CPU has access to the data it needs at a moment’s notice, rather than going to and from a relatively slow storage device. Without memory, even the simplest task on your computer would be unbearably slow. Consequently, the more memory a computer has, the more efficiently it can handle multiple tasks and resource-intensive programs.
RAM is what’s known as ‘volatile’ memory, meaning it requires power to keep hold of the data that’s stored on it. Unlike long-term storage devices, which retain data even when they’re powered off, any data stored in memory will be wiped clean when you shut down your computer. So while RAM may be ultra-fast, it’s only used for short-term, temporary storage. Hard drives and solid state drives, on the other hand, will retain your data on a long-term, permanent basis.
Memory is not the same as storage, though we can see why it’s easy to conflate the two. While they may sound interchangeable, memory and storage perform very different roles in a computer. In short, memory (RAM) is fast but volatile, whereas storage (HDD, SSD) is non-volatile but comparatively slow. To reiterate, when you open a program, its data is transferred from storage to memory. Then when you close a program, its data is saved back to storage for safe keeping. Since it’s volatile, RAM is used only as a temporary workspace for the processor.
Everything on your computer requires RAM to function. It doesn’t matter if it’s a memory-hungry triple-A game or a small background process sitting in your system tray – it all consumes RAM. You need to make sure you’ve got enough RAM in your computer. Otherwise, a computer will resort to substituting storage for memory in a process called ‘paging’ or ‘swapping’. When this happens, it’s not a good time. As storage is much slower than memory, running off this virtual memory will degrade performance and result in slowdowns, freezes, and sometimes outright crashes.
The operating system itself requires memory to function properly. Common OS-related tasks like managing resources, processes, and caching frequently used data all rely upon there being available RAM. Insufficient RAM can lead to the rest of the computer bogging down as the operating system struggles to allocate resources efficiently.
Each program consumes a portion of memory to store its instructions and data for the CPU to execute on. So by equipping a PC with more RAM, you allow it to run more programs simultaneously. The more memory a computer has, the more data it can keep readily available. This reduces the need to swap data in and out of slower storage devices, which would otherwise decrease efficiency and performance. As a result, having a large amount of memory is important for those who like to multi-task between multiple programs.
While the amount of RAM needed varies from user to user depending on their use cases (as detailed thoroughly in our blog How Much RAM Do You Really Need), it’s recommended to have a bit more than the bare-minimum recommended amount. If, for example, a new game requires 8GB of RAM, it’s generally a good idea to have 16GB of RAM to account for everything else running in the background of a system. This could include voice-over-IP software like Discord, a browser with many tabs open, and the operating system itself. Having this extra memory headroom will help ensure a smoother, stutter-free computing experience.
CPU themselves feature a small amount of ultra-high-speed memory called cache. Cache serves the same purpose as RAM: to give the CPU near-instant access to data. However, there’s only so much cache that can be squeezed onto a CPU due to their size. For example, even with its impressive 3D-stacked cache technology, the AMD Ryzen 7 7800X3D still only has 96MB of L3 cache. This is only enough to hold the most essential of data. RAM is like an extension of this cache for all your other data, even if it is a bit slower since it has to travel across the motherboard.
The relationship between RAM frequency and RAM latency is crucial for understanding memory performance. RAM frequency describes that rate at which data can be transferred to and from a memory module. On most sites, including ebuyer, it’s measured in MHz (megahertz). At higher frequencies, more data can be moved in a given amount of time, leading to faster performance. RAM latency, on the other hand, describes the time delay between when a command is issued to RAM and when the data actually become available to use. It’s measured in clock cycles and is usually represented by a CL (CAS Latency) value: CL14, CL16, and so on. Lower CL values mean lower response times and, in turn, better performance.
Ideally, you’d want both high-frequency and low-latency RAM for optimal performance. However, it’s common for there to be a trade-off between the two. Sometimes, higher frequency memory modules may have slightly higher latency, and vice versa. So having extremely high frequency RAM won’t necessarily translate to better performance if its latencies are too high. Even if the on-paper transfer rate is fast, the delay in accessing the data can negate any performance advantage. Conversely, RAM with lower latency can compensate for lower frequency by providing much quicker access to data. In some cases, PCs with lower-frequency, lower-latency RAM may outperform those with higher-frequency, higher-latency RAM. So while both frequency and latency are important, it’s crucial to strike a balance between the two. When shopping for RAM, don’t just look at frequencies at the top of the product description – scroll down further to see its latencies, too.
When shopping for memory modules, you will come across DDR4 and DDR5 RAM. DDR stands for Double Data Rate, a technology used in memory modules to improve performance. As the name implies, DDR allows data to be transferred twice per clock cycle, effectively doubling data transfer rates compared to older, now-obsolete SDR (Single Data Rate) RAM. There are several generations of DDR RAM, each bringing improvements in speed, bandwidth, and efficiency.
Importantly, each generation of DDR RAM is not designed to be backward compatible with the previous generation. You cannot, for example, install a DDR5 RAM module in a motherboard with a DDR4 RAM slot. Upgrading to a newer generation of DDR RAM brings with it a noticeable improvement in performance, especially in memory-hungry programs and bandwidth-starved workloads.
There are two different form-factors of memory modules. DIMM (Dual In-Line Memory Module) is the standard form-factor used in desktops, servers, and high-performance workstations. It’s a full-sized form-factor featuring a longer, rectangular shape with pins on either side of the module. Since it’s larger than SO-DIMM, DIMM is typically used in roomier, stationary systems. SO-DIMM (Small Outline Dual In-Line Memory Module) is used in laptops, mini PCs, all-in-one PCs, and other compact systems where space is at a premium. Compared to DIMM, SO-DIMM is smaller, stubbier, and only has pins on one side of the module. While both DIMM and SO-DIMM modules serve the same purpose in a computer, their different sizes affect compatibility. It’s important to be able to distinguish between DIMM RAM And SO-DIMM RAM so you can ensure compatibility with your system’s motherboard. Over at ebuyer, we break RAM down into three different categories: Laptop RAM (SO-DIMM), Desktop RAM (DIMM), and Server RAM (DIMM again).
Written by:Falak Kumar