Why RAM Matters More Than CPU for Most Laptop Users

The Counterintuitive Truth About Laptop Slowdowns

Ask most people why their laptop is slow and they'll blame the processor. It's old, they say. It's not powerful enough. They need a faster chip.

In many cases, they're wrong.

The CPU in a three-year-old laptop is rarely the bottleneck for everyday tasks. Modern processors — even mid-range ones from 2022 onwards — can handle email, documents, video calls, and web browsing without breaking a sweat. What actually causes the stuttering, the spinning beach balls, the agonising wait when switching between applications is almost always RAM running out.

RAM is where your laptop holds everything that's currently happening. Every open application, every browser tab, every background process — it all lives in RAM. When RAM fills up, the operating system starts improvising, and that improvisation is what makes laptops feel ancient.

Understanding this changes how you should buy a laptop and how you should evaluate one you already own.

What Actually Happens When RAM Runs Out

This is where things get interesting, and a little uncomfortable if you're running 8GB.

When your laptop exhausts its physical RAM, both Windows and macOS have a fallback: they move the least-recently-used data out of RAM and onto the storage drive, freeing up space for whatever you're trying to do now. In Windows, this is called the page file. On macOS, it's called swap. The concept is the same: your SSD pretends to be RAM.

The problem is that SSDs, fast as they are, are not RAM. A modern NVMe SSD reads data at 3,000–7,000 MB/s. Actual LPDDR5 RAM moves data at over 60,000 MB/s. That's roughly a 10–50x speed difference.

So when your laptop starts swapping heavily, you're replacing a fast motorway with a single-lane country road. You'll feel it immediately: apps take seconds instead of milliseconds to respond, switching tasks becomes painful, and the SSD activity light (if your laptop has one) stays on almost constantly because the system is frantically shuffling data back and forth.

In extreme cases — a badly RAM-starved machine trying to do too many things — the laptop can feel almost unusable despite having a perfectly good CPU that's barely being asked to do anything.

Why 8GB Is Increasingly Insufficient in 2026

Memory consumption in common software has grown substantially over the past five years. What worked on 8GB in 2020 is genuinely strained in 2026.

Browser tabs are the biggest culprits. A single Chrome or Edge tab uses anywhere from 100MB to 400MB depending on the site. A news site with autoplay video, background scripts, and advertising infrastructure can top 500MB on its own. If you keep 10 tabs open — which many people do routinely — you've used 1–4GB on your browser alone before you've opened anything else.

Video calls are heavy. Microsoft Teams, Zoom, and Google Meet each consume 500MB to 2GB depending on call size, whether you're sharing your screen, and how many participants have video on. Running a call while doing anything else puts immediate pressure on an 8GB system.

Office applications hold memory even when minimised. Microsoft Office apps are known for this — Outlook, Excel, and Word together can hold 800MB to 1.5GB while sitting quietly in the background.

Operating system overhead: Windows 11 uses approximately 2–3GB of RAM just for itself and background processes at idle. macOS is similar. That eats significantly into an 8GB budget before you've launched a single application.

Add these up and an 8GB laptop doing normal, modern work tasks is under pressure before you've opened anything unusual.

RAM and Virtual Memory: The SSD Tax You Don't Want to Pay

It's worth going slightly deeper on virtual memory, because understanding it changes how you think about the RAM question.

Virtual memory is a feature of every modern operating system. It creates the illusion of more RAM than you physically have by using part of your storage drive as overflow. The concept is clever. The execution, in practice, has a speed cost that compounds quickly.

When your system has to access data that's been moved to the page file or swap, it has to read it from storage, write other data out to make room, and bring the requested data back into RAM. This all happens in the background, but it creates latency. The more the system has to do this — and when RAM is exhausted it happens constantly — the worse it feels.

On a laptop with a cheap QLC NAND SSD rather than a fast TLC or MLC drive, this problem is worse. Budget laptops often cut costs here, which means they're both RAM-limited and storage-speed-limited simultaneously.

The practical takeaway: there is no software trick, no optimisation, and no setting that compensates for not having enough physical RAM. The SSD tax is real and unavoidable once you've run out.

Unified Memory: Why Apple Silicon Is Different

Apple's M-series chips (M3, M4, M4 Pro, and onwards) handle memory differently from traditional laptop designs, and this is worth understanding before you dismiss an 8GB MacBook.

In a conventional laptop, the CPU and GPU are separate components with separate memory pools. The CPU has its system RAM; a discrete GPU has its own VRAM. Transferring data between them takes time.

Apple Silicon uses a unified memory architecture: the CPU, GPU, and Neural Engine all share the same pool of high-bandwidth memory, mounted directly on the chip package. There's no bottleneck between the processor and memory, and no redundant copies of data.

This makes Apple's unified memory more efficient than an equivalent amount of traditional RAM. Apple's 8GB unified memory in an M4 MacBook Air genuinely performs closer to 12–16GB of traditional laptop RAM for typical tasks. Apple isn't making this up — the memory bandwidth (over 100 GB/s on M4) and the unified architecture mean the system manages memory more effectively.

That said, 8GB is still the minimum on Apple Silicon. Professionals doing video editing, running multiple heavy applications, or working with large datasets should still choose 16GB or 24GB on Apple hardware. The efficiency advantage doesn't make RAM infinite.

RAM Types: LPDDR5 vs DDR5 vs LPDDR4X

Not all RAM is equal in terms of speed, and the type your laptop uses affects performance — particularly for tasks that saturate memory bandwidth.

LPDDR5 and LPDDR5X are the current standard for ultrabooks and thin-and-light laptops. The "LP" stands for Low Power — these modules are optimised for efficiency and are physically integrated onto the motherboard. Fast speeds (up to 8533 MT/s on LPDDR5X), low power draw, not user-upgradeable.

DDR5 is used in larger laptops and workstations with socketed RAM slots. Slightly faster at peak speeds than LPDDR5 in some configurations, runs at higher voltage, and can be upgraded or expanded after purchase.

LPDDR4X is the previous generation, still found in some budget and mid-range laptops. Noticeably slower than LPDDR5 (around 4266 MT/s vs 6400+ MT/s), which matters for integrated graphics performance and memory-intensive tasks.

For most everyday work, the difference between LPDDR4X and LPDDR5 is noticeable but not dramatic. For integrated graphics gaming or creative work that taxes memory bandwidth, LPDDR5 is meaningfully better.

Soldered vs Upgradeable RAM: The Permanence Problem

This is perhaps the most practically important RAM consideration for laptop buyers.

A growing majority of modern laptops — including every MacBook, most Surface devices, the Dell XPS range, HP Spectre, ASUS ZenBook, and many others — solder the RAM chips directly to the motherboard. The benefit for manufacturers is a thinner chassis, better power efficiency, and lower failure rates (soldered joints are more reliable than socket connections). The cost to you is permanence: whatever RAM you buy with the laptop is the RAM you're stuck with.

This makes the RAM decision at purchase critically important. There's no upgrade path if you find 16GB isn't enough two years later. You either buy a new laptop or live with the limitation.

Some laptops still use socketed SO-DIMM slots. This is increasingly found on gaming laptops, workstation-class machines, and certain business laptops like ThinkPad T-series models. If upgradeability matters to you — if you want to start at 16GB and move to 32GB later — verify the specific model supports user RAM upgrades before buying.

How Much RAM Do You Actually Need?

Here's a practical breakdown by use case, without the fluff.

8GB: Email, basic web browsing with a few tabs, light document work on a Chromebook alternative. Increasingly marginal for Windows or macOS in 2026. Only acceptable if you genuinely never do anything else.

16GB: The comfortable baseline for most users. Handles video calls, 10–15 browser tabs, Office applications, Spotify or YouTube in the background, and general multitasking without complaint. Right for most students, home users, office workers, and light creative professionals.

32GB: For developers running containers or virtual machines, content creators editing 1080p or 4K video, professionals using heavy creative suites simultaneously, or power users who simply don't want to think about RAM for five years. Also appropriate for anyone who frequently has 20+ browser tabs open alongside several applications.

64GB and above: Machine learning engineers, architects and engineers running CAD or simulation software, professionals working with very large datasets, and post-production video editors working with uncompressed 4K or 6K footage. Overkill for the vast majority of laptop buyers.

RAM and Gaming: Why Dual-Channel Matters

Gaming on a laptop is partly a RAM story, and a nuanced one.

Modern games generally require 16GB as a practical minimum. Games themselves have grown in memory footprint, and running a game alongside Windows, Discord, a browser for guides, and background apps pushes 8GB systems to their limits consistently.

For laptops with a dedicated Nvidia or AMD graphics card, the game's textures and rendering data sit in VRAM (separate from system RAM), so system RAM requirements are mainly about running everything alongside the game.

For laptops relying on integrated graphics — Intel Arc, AMD RDNA integrated, or Apple Silicon — the GPU shares the system memory pool. This makes memory bandwidth critical for gaming performance, and dual-channel configuration becomes important. Dual-channel means two identical RAM sticks running in parallel, effectively doubling the memory bandwidth available to the integrated GPU. A laptop with dual-channel 16GB (two 8GB sticks) can produce meaningfully better frame rates in games than a single 16GB stick running in single-channel mode.

If you're buying a budget laptop and hoping to play games on integrated graphics, confirm that it uses dual-channel RAM.

RAM and Creative Work: Video Editing, Music Production, Photo Editing

Creative software is the most RAM-hungry category of everyday applications, and the requirements are worth understanding concretely.

Video editing: DaVinci Resolve and Adobe Premiere Pro both state 16GB as their minimum for HD editing, with 32GB recommended for 4K. In practice, timeline playback, colour grading, and rendering with effects running simultaneously consume RAM aggressively. On 16GB, you'll need to close other applications while editing. On 32GB, you can work more freely.

Music production: A DAW like Ableton Live, Logic Pro, or FL Studio itself isn't particularly RAM-hungry. The RAM consumption comes from sample libraries and virtual instruments. A large orchestral template in Kontakt or Spitfire Audio can consume 16–20GB of RAM on its own before you've added anything else. For serious music production with large sample libraries, 32GB is the practical starting point.

Photo editing: Adobe Lightroom is a RAM hog. Large catalogues with heavy local adjustments and lots of smart previews can consume 8–12GB in a long editing session. 16GB handles most photography workflows; 32GB gives you breathing room for very large RAW files (from 40+ megapixel cameras) and complex retouching work in Photoshop alongside Lightroom.

How to Check Your Current RAM Usage

Before buying anything, check how your current laptop actually uses RAM. This tells you whether you're genuinely constrained or just think you are.

On Windows: Open Task Manager (Ctrl+Shift+Esc), click the Performance tab, and select Memory. You'll see total RAM, what's in use right now, and what's in use at peak. Do this while you're doing your typical work — all your usual applications open, a video call running if that's part of your day. If the "In Use" figure regularly exceeds 75% of your total RAM, you're feeling RAM pressure.

On macOS: Open Activity Monitor (Applications > Utilities), click the Memory tab. The "Memory Pressure" graph at the bottom tells the story — green means comfortable, yellow means pressure, red means the system is actively swapping. Also check the "Swap Used" figure: if it's regularly above a few hundred MB, you're running short.

These checks take two minutes and tell you more about whether you need more RAM than any benchmark comparison.

The Upgrade Question: What Are Your Options?

If you've confirmed you're RAM-limited, your options depend on your laptop.

Socketed RAM (SO-DIMM): Check your laptop's service manual or iFixit page. If your laptop has SO-DIMM slots, you can buy compatible RAM from Crucial, Kingston, or similar brands and install it yourself. The upgrade is usually straightforward and cost-effective — often £40–£80 for an upgrade from 8GB to 16GB.

Soldered RAM — no upgrade path: For most modern thin-and-light laptops and all MacBooks, there's no hardware upgrade path. Your options are software optimisation (closing apps you're not using, disabling background processes) or accepting the limitation until you replace the machine.

Chromebooks: Chromebook RAM is always soldered and typically modest. ChromeOS manages RAM aggressively and efficiently, but the fundamentals of RAM limitations still apply.

The honest answer is that the best time to get the RAM right is before you buy. It's the one spec where the upgrade you skip at purchase becomes the frustration you live with for the life of the machine.