What is the performance impact of upgrading to faster RAM?

Introduction

For some PC and laptop models, we offer RAM upgrades with different speeds. In this article, we explain their impact in various scenarios.

What is RAM speed?

RAM speed indicates how quickly the memory can transfer data. Data is exchanged between the processor (CPU) and RAM. The speed is officially specified in megatransfers per second (MT/s). In some cases, you may also see MHz (megahertz) used instead. Although this is not technically correct, it is commonly used as a synonym.

Examples: DDR5-5600 delivers 5600 MT/s (million data transfers per second). DDR5-6400 delivers 6400 MT/s, which is about 14% more transfers per second than DDR5-5600.

Integrated GPUs (iGPUs, found inside CPUs) and scenarios in which large amounts of data are frequently exchanged between the processor and RAM, without other bottlenecks, benefit from higher RAM speeds. We will explain which specific scenarios these are later in the article.

The performance gain also depends on latency (access delay). Modules with higher clock speeds often also have slightly higher latency values (for example CL46 vs. CL52), so that effective speed and access times influence each other. The actual difference is therefore often smaller than the pure MT/s figures suggest.

Price and energy consumption

Higher-clocked RAM is usually more expensive per gigabyte. If your budget is limited, you could, for example, invest in a larger SSD instead of faster RAM. This decision should be carefully considered.

In addition, power consumption increases slightly with clock speed. In laptops, this can have a minor effect on battery life, usually only in the single-digit percentage range.

The higher power consumption under load also generates slightly more heat, both in the RAM itself and in the CPU’s memory controller. In principle, this additional heat is not a problem: our PCs and laptops are always designed, equipped and validated to cool their maximum configuration without any performance loss. Nevertheless, the question of total system energy efficiency should be kept in mind.

Office and web

Apart from special cases, you will hardly notice any difference in everyday applications. Virtually all office software, web browsers or multimedia applications run just as fast with DDR5-5600 as with DDR5-6400, as long as there is sufficient RAM capacity.

When in doubt, RAM capacity, measured in gigabytes (GB), is always more important than RAM speed. This is especially true for multitasking (many open tabs, frequent switching between programs). Having too little RAM will lead to temporary swapping to the SSD, which causes sluggish system response.

As of 2025, 16 GB is considered the absolute minimum. When in doubt, 32 GB is a much better choice, and power users benefit from even larger capacities.

Our recommendation: Upgrading RAM speed alone for office or web use is hardly worthwhile, if at all. Capacity is the decisive factor. Faster RAM only becomes interesting in the more advanced scenarios outlined below.

Gaming without a dedicated graphics card (iGPU Only)

If a laptop does not have a dedicated GPU, faster RAM can significantly increase the gaming performance of the integrated graphics unit (iGPU/APU).

iGPUs do not have their own memory but use system RAM as video memory. The RAM bandwidth is therefore also the bandwidth of the video memory.

iGPUs have many processing cores or render pipelines waiting to be fed with data – faster RAM ensures that these cores spend less time idle.

In practice, the performance of integrated graphics units scales almost linearly with memory clock speed. Depending on the game, the difference between DDR5-5600 and DDR5-6400 can mean an average FPS increase of 10 to 15%.

However, taking iGPU performance into consideration only makes sense if the iGPU itself is powerful enough for gaming. Desktop-derived CPUs such as the Intel Core Ultra 9 275HX or AMD Ryzen 9 9955HX have very small iGPU units – they are not suited for gaming, regardless of RAM speed.

iGPU gaming only becomes relevant when CPUs have relatively large integrated graphics units, such as the AMD Ryzen AI 9 HX 370 with Radeon 890M or the Intel Core Ultra 7 255H with Intel Arc 140T in the XMG EVO series. In the XMG EVO 15 (M25) with Intel, we actually offer RAM speed upgrades thanks to Intel’s platform support for CSO-DIMM.

Our recommendation: The upgrade (if offered in the configurator) is worthwhile if the system is fundamentally suitable for iGPU gaming and intended for that purpose.

Gaming with a dedicated graphics card (dGPU)

In systems with a dedicated graphics card (for example NVIDIA GeForce RTX series), the dGPU has its own video memory (GDDR6, GDDR6X, GDDR7, etc.). Therefore, its performance does not depend on system RAM speed.

However, faster RAM can still be beneficial in certain situations, especially in CPU-limited games and scenarios such as:

• Gaming at low resolutions.
• Targeting very high FPS in eSports titles.
• Strategy games, simulations, or open-world games with many simultaneously calculated objects and NPCs.

Faster RAM can also improve the so-called "1% low" frame rates – the lowest 1% of measured FPS responsible for perceived stutters or lags.

(The same applies to CPUs with particularly large caches, such as the AMD Ryzen 9 9955HX3D with 3D V-Cache. Although this CPU only supports DDR5-5600, the large cache significantly reduces dependence on memory speed.)

If a game is GPU-limited, however, RAM speed is almost irrelevant. This includes:

• Gaming at very high resolutions (such as 4K or ultrawide).
• Gaming with maximum graphical details, including ray tracing – more typical of single-player titles.
• Use of VR headsets, which also operate at very high resolutions.

However, memory bandwidth utilisation in gaming is difficult to predict and depends on many factors. CPU bottlenecks (for example short-term load peaks that affect 1% low FPS) can also occur unexpectedly in certain games or poorly optimised titles.

Our recommendation: To prepare your gaming PC or laptop for all possible scenarios, we recommend upgrading to a higher memory speed, if this is available in the configurator.

Productivity workloads

Professional and productivity workloads include 3D rendering, video encoding, software development, scientific calculations, AI and similar tasks. Virtually all productivity applications have one thing in common: it never hurts to have more RAM. The relevance of higher RAM speed, however, depends on the specific application.

3D rendering via software

In tile-based rendering (Blender, Cinebench, etc.), a scene is divided into tiles, each processed by one of the CPU cores. The amount of data per core usually fits into the CPU’s cache. The calculation is therefore limited by core performance, and rendering times hardly benefit from faster memory.

Content creation

Professional-level content creation can benefit from faster RAM. PugetSystems found a significant difference between very slow and very fast RAM in Adobe Premiere Pro on Intel systems in certain export and playback tests. On AMD systems, however, this effect was barely noticeable. Minor differences were also seen in After Effects, Photoshop and Lightroom (source).

Software development and compiling

In very large projects (for example Android AOSP Build, Unreal Engine Compile), where huge numbers of files are processed in parallel and much intermediate data is stored in memory, the throughput of your memory can have an impact. However, many other factors also play a role (storage speed, number of CPU cores, parallelisation, RAM capacity), so RAM speed alone is not decisive.

Scientific calculations / Simulation / CAD

Some specialised applications benefit from faster memory when processing large data sets that do not fit entirely into the CPU cache. These include:

• Finite element simulations
• Database in-memory analyses
• SPICE simulations
• MATLAB matrix calculations
• Professional CAD workstation software

These scenarios benefit from faster memory because very large amounts of data are continuously exchanged between the CPU and RAM at high bandwidth.

AI applications

Modern laptop CPUs include dedicated AI accelerators, also known as neural processing units (NPUs). These include Intel AI Boost and AMD Ryzen AI Engine. Like iGPUs, these NPUs utilize the system's main memory. One might assume that faster RAM is important here, but in practice, performance is usually limited by the relatively modest processing power of these small AI accelerators and by limited software support.

For serious AI workloads (large LLMs, image or video generation, etc.), powerful dedicated GPUs (or cloud servers) are used. In such cases, system RAM speed is irrelevant, as GPUs have their own high-bandwidth memory.

Short AI tasks that run purely on the CPU or NPU may benefit slightly from faster RAM, but the effect is minor and currently not a convincing reason for a RAM speed upgrade.

Synthetic benchmarks

For completeness, it should be noted that synthetic memory benchmarks or memory-intensive micro-benchmarks (for example in the HPC sector) show greater differences. The 7-Zip compression benchmark, for instance, can scale with memory speed, and tools such as the AIDA64 Memory Benchmark show dramatic MB/s uplifts. These figures may look impressive but often do not reflect real workloads. In real-world applications, as described above, we usually see a much smaller percentage impact or none at all.

Conclusion

Above all, for productivity workloads, having sufficient RAM capacity is crucial. If you reach the capacity limit, system performance drops dramatically. By contrast, higher RAM speed can only yield a few additional percentage points of performance, even in the best case.

Our recommendation: RAM capacity is more important than RAM speed. However, if your workloads have at least a "medium" impact according to the table above, upgrading the RAM speed can be worthwhile – provided it does not come at the expense of a CPU, GPU or SSD upgrade.

Overview table

The following table summarises the impact that upgrades to faster memory can have in each use case.

Our recommendation: For impacts ranging from "medium" to "large", we strongly recommend upgrading to faster RAM (if such upgrades are available in the configurator).

Note on dual channel

Even more important than memory speed is dual-channel operation – the simultaneous use of two identical RAM modules. As a rule, all our PCs and laptops are configured with dual-channel memory, no matter which capacity level you pick. We do not offer single-channel configurations.

Exception: if you downgrade to 1×8 GB in the configurator, you will initially receive a single-channel configuration. This option is intended for customers who plan to install their own RAM later. If you choose this route, you should ensure that you upgrade to dual-channel operation afterwards.

Contact

We hope these explanations help you make an informed decision. If you have any questions about the optimal configuration, our team will be happy to advise you.

Read more

External articles and performance comparisons:

• Impact of DDR5 Speed on Content Creation Performance (2023 update) [pugetsystems.com]
• DDR5 vs. DDR4 Gaming Performance [techspot.com]
• Does RAM Speed REALLY Matter?? (DDR5 Edition) [Linus Tech Tips (YouTube)]
• Does RAM MHz Matter? Guide for Gamers & Professionals [tlmintl.com]

Related FAQ articles:

• Why is it important to pick dual channel memory when putting together a PC or laptop?
• How can I set a program or game to run on either the iGPU or the dGPU?
• Is it worth it to pick a faster SSD with PCIe 4.0 (or even PCIe 5.0) instead of PCIe 3.0?
• What kind of hardware do I need to get ready for AI?
• Which XMG or SCHENKER laptops are recommended for my specific use-case?