Why Does My PC Keep Crashing or Freezing While Gaming?

If your PC runs perfectly for browsing and email but crashes, freezes or black-screens the moment a game gets busy, the load is the clue. Gaming pushes your graphics card, processor, memory and power supply far harder than anything else you do, so a weakness that stays hidden during normal use gets exposed the second the action ramps up. The four usual suspects are heat, power, unstable memory and graphics drivers, and this guide shows you how to tell them apart.

This is the gaming-specific companion to our general walkthrough on how to fix random PC crashes or restarts. If your machine also crashes at the desktop, on the login screen or while idle, start there first, because that points to a broader fault. If it is rock solid until you launch a game, you are in the right place. And if you would rather hand the whole thing over, our gaming PC repair service in Manchester can diagnose it for you and replace whatever turns out to be at fault.

First, what is your PC actually doing when it crashes?

Crash is a catch-all word, but the precise behaviour tells you a lot before you touch a single setting. Watch carefully next time and match it to this list, because each pattern leans towards a different cause.

• Hard power-off (the machine cuts dead and reboots): the classic signature of heat protection kicking in or a power supply that cannot deliver a sudden spike. There is usually no blue screen, it just goes.
• Full freeze or lock-up (image sticks, sound loops or buzzes): often memory instability, an overclock that is not solid, or a driver that has hung. You normally have to hold the power button.
• Black screen, then recovery (screen goes dark for a second or two, then the desktop returns): this is the graphics driver crashing and restarting itself. Windows may pop up "Display driver stopped responding and has recovered."
• Crash to desktop (the game closes but Windows keeps running): usually a software, driver or game-file problem rather than failing hardware, though unstable memory can do it too.
• Blue screen with a stop code (BSOD): Windows caught something serious. The stop code (for example VIDEO_TDR_FAILURE or a memory management error) narrows it down considerably.

Note the symptom down every time it happens, along with what you were doing. A pattern (only in one game, only after 30 minutes, only in busy scenes) is worth more than any single crash.

Why gaming specifically pushes a PC over the edge

Ordinary tasks use your components in short, gentle bursts. Gaming is the opposite: it drives the graphics card and processor to sustained high load, keeps the memory busy constantly, and streams data off your drive without a break. That does three things a spreadsheet never will.

• It generates real, sustained heat. Temperatures climb for the whole session rather than spiking for a second, so anything marginal (dried thermal paste, clogged fans, poor airflow) gets found out.
• It demands real power, in violent bursts. A modern graphics card can jump from a trickle to its full draw in milliseconds when a busy scene loads, and those transient spikes stress the power supply far more than a steady load.
• It exercises every part at once. If your memory profile is a touch too aggressive or your overclock is not quite stable, gaming is the workload that finally trips it, because it is the only time all the parts are flat out together.

That is why the fix is rarely about the game itself. The game is just the stress test that reveals a hardware or configuration weakness that was always there.

Heat and thermal shutdown while gaming

Overheating is the most common reason a gaming PC cuts out under load. When a component reaches its safety limit, it protects itself: first by throttling (dropping speed, which shows as sudden stutter and frame drops), and then, if the temperature keeps climbing, by shutting the whole system off instantly. That hard power-off with no warning is textbook thermal protection.

Rough guide numbers for what is normal versus worrying while gaming:

• Graphics card: comfortable in the 65 to 85°C range under load. NVIDIA cards generally throttle around 90°C, AMD cards a little higher, and AMD also reports a "junction" or hotspot figure that can safely reach around 110°C. If your card sits pinned at its limit and then black-screens or shuts down, heat is a prime suspect.
• Processor: below roughly 80°C under load is healthy. Modern chips will happily boost into the high 80s, but a hard shutdown as it touches its limit means the cooler is not coping.

The usual culprits are dust choking the fans and heatsinks, thermal paste that has dried out over years of use, a fan that has failed or slowed, or simply poor case airflow. Cleaning and re-pasting can drop temperatures by 10 to 20°C and cure the crashes outright. We go into the causes and the airflow fixes in detail in why does my gaming PC overheat. A quick test: if the crashes stop or delay noticeably when you run with the side panel off and a desk fan blowing in, heat is very likely your problem.

Power: when the PSU simply cannot cope

The power supply (PSU) is the most under-appreciated cause of gaming-only crashes. It can run your PC flawlessly at idle and then fail the instant a game demands everything at once. The tell is a clean, instant power-off or reboot with no blue screen, most often right as a scene gets busy or a game first loads.

Two things go wrong here:

• Not enough headroom. A PSU that is barely big enough on paper has nothing left for those millisecond power spikes a modern graphics card throws out. When the spike exceeds what it can safely deliver, its protection circuit cuts power rather than risk damage, and your PC dies on the spot.
• Age and wear. PSUs weaken over years of use. A unit that coped fine when the PC was new may no longer hold voltage steady under a heavy gaming load, so it crashes only when pushed.

Cheap or generic units are the worst offenders, because their real, stable output is often well below the number on the box. If you recently fitted a more powerful graphics card and the crashes started then, the PSU is the first thing to look at. We cover the warning signs and how to be sure in is my PSU failing: the signs to watch for. Because power delivery is hard to test without swapping in a known-good unit, this is one where a workshop test saves a lot of guesswork.

Memory: XMP and EXPO instability that only shows under load

Here is a cause that catches out a huge number of self-built and prebuilt gaming PCs. Your RAM has an advertised speed, but out of the box it often runs slower until you switch on its speed profile in the BIOS, called XMP on Intel platforms and EXPO on AMD. That profile is essentially a factory overclock, and it is not guaranteed to be stable on every combination of processor, motherboard and memory.

When a memory profile is slightly too aggressive for your particular hardware, the PC can seem perfectly fine day to day and then crash, freeze or blue-screen under the full load of a game. Because gaming is the workload that hammers the memory hardest, an unstable profile often shows up nowhere else. Common signatures are random freezes, crashes to desktop, or stop codes mentioning memory management.

How to test it:

• Turn the profile off. In the BIOS, disable XMP or EXPO so the memory runs at its safe default speed, then game as normal. If the crashes vanish, you have found your cause, and the fix is a more conservative memory setting rather than new parts.
• Test the memory properly. Run a bootable tool like MemTest86 for several full passes, or a Windows-based memory test, and aim for zero errors. A single error means the current settings are not stable. Even one is too many.

If the memory is only stable at its default speed, that is usually fine: the real-world gaming difference is small, and stability beats a few frames every time. If it errors even at default, the module itself may be faulty and worth replacing.

Graphics driver crashes: black screens, TDR and crash to desktop

If your screen goes black for a second or two and then recovers, or you get a blue screen quoting VIDEO_TDR_FAILURE, you are looking at a graphics driver crash. Windows has a built-in safety net called Timeout Detection and Recovery (TDR): if the graphics card stops responding for about two seconds, Windows resets the driver instead of letting the whole PC hang. When that works you see "Display driver stopped responding and has recovered"; when the reset fails you get the VIDEO_TDR_FAILURE (bug check 0x116) blue screen documented by Microsoft.

Driver-related crashes have several common triggers:

• A bad or half-installed driver. The single most common cause. Updates go wrong, or leftovers from an old version conflict with a new one.
• Overclocking or aggressive factory profiles. A graphics card pushed too hard, or an unstable memory profile, often surfaces as a driver timeout.
• Heat or power under the surface. A card that is starved of power or overheating can look like a driver fault, so rule those out too.

The reliable fix is a clean driver reinstall: fully remove the existing graphics driver and install the current version from the maker's website (NVIDIA, AMD or Intel) rather than layering one on top of another. If crashes persist across a genuinely clean driver install, the problem is more likely hardware, power or heat. AMD systems have their own quirks around drivers and chipset software, which is why we run a dedicated AMD computer repair service in Manchester.

Storage, VRAM and the assets your game is streaming

Games constantly stream textures and level data off your drive and into your graphics card's memory (VRAM). Two related problems can cause freezes and stutters that feel like crashes.

• A failing or overloaded drive. A drive that is on its way out, nearly full (below about 10 percent free space cripples an SSD), or thermal throttling can stall the game while it waits for data. The giveaway is stutter and brief freezes as you move into new areas, rather than during explosions or heavy combat. Check the drive's health with a tool like CrystalDiskInfo and look for reallocated sectors or warnings. A dying drive can also corrupt game files, which then crash the game to desktop.
• Running out of VRAM. If you play at very high resolutions or texture settings on a card with limited memory, the game can exhaust its VRAM and stutter, hitch or crash. Lowering texture quality is a quick test: if the crashes ease, VRAM pressure was the issue.

It is also worth verifying the game's own files (Steam and other launchers have a "verify integrity" option), because a single corrupt file can cause repeatable crashes in exactly one spot, which no amount of hardware testing will explain.

Overclocks, undervolts and "it was stable last week"

If you (or whoever built the machine) tuned the processor, memory or graphics card beyond stock, that is a prime suspect for load-only crashes. An overclock that passed a quick test can still fail under a long, demanding gaming session as heat builds and the workload varies. Undervolting a graphics card to run cooler and quieter can also cause crashes if taken a step too far, and it very often shows up only in games.

The diagnostic move is simple: set everything back to stock. Clear any processor and graphics overclock, disable XMP or EXPO, and remove any undervolt. Then play. If the machine is now stable, you have proven the tuning was the cause, and you can reapply it more conservatively or leave it at stock. This one change resolves a surprising number of "my PC keeps crashing while gaming" cases at no cost.

A decision-tree: is it heat, power, RAM or drivers?

Rather than changing everything at once, use the symptom to point you at the most likely cause first, then confirm it. This table maps what you see to where to look.

What you see	Most likely cause	First thing to check
Instant power-off and reboot, no blue screen, often in busy scenes	Power supply, or heat	Watch temperatures while playing; test with a known-good PSU
Sudden stutter and frame drops, then a shutdown if it continues	Overheating and thermal throttling	Card or CPU temperature at the moment it happens
Screen goes black for a second or two, then recovers	Graphics driver (TDR)	Clean-reinstall the graphics driver
Full freeze or lock-up, sound loops or buzzes	Unstable memory or overclock	Disable XMP or EXPO; clear overclocks
Blue screen with a memory management stop code	RAM or memory profile	Run MemTest86; set memory to default speed
Game closes to desktop but Windows stays up	Driver, game files or software	Verify game files; clean-reinstall driver
Stutter and freezes only when loading new areas	Storage or VRAM	Check drive health; lower texture settings

The golden rule underneath all of this: change one thing at a time. If you swap the PSU, reinstall drivers and disable XMP all at once and the crashing stops, you still do not know which one fixed it, and the next upgrade could bring it straight back.

The monitoring tools that catch the culprit

You do not have to diagnose blind. A couple of free tools let you watch exactly what your PC is doing at the instant it crashes, which turns guesswork into evidence.

Tool	Best for	What it tells you
HWiNFO	Detailed sensor logging	Every temperature, voltage, clock and fan, and it can log to a file so you can see what happened in the seconds before a crash
MSI Afterburner (with RivaTuner)	In-game overlay	Live on-screen temperatures, clocks, framerate and frame times while you actually play
CrystalDiskInfo	Drive health	SSD or hard drive condition, temperature and warning signs of failure
MemTest86	Memory stability	Whether your RAM (and its XMP or EXPO profile) is genuinely stable
Windows Reliability Monitor and Event Viewer	After-the-fact clues	A timeline of crashes and the stop codes or events behind them

Two things to look for in the logs. First, does a temperature hit its limit right before the crash? That is heat. Second, does the whole log just stop dead with everything looking normal? A clean cut-off with healthy temperatures points hard at power. On the Windows side, an Event ID 41 (Kernel-Power) only tells you the PC restarted without shutting down cleanly, so treat it as confirmation of a hard crash, not the cause itself.

A step-by-step plan to isolate and fix it

Put it all together and work through it in this order. Stop as soon as the crashing stops, because that is your answer.

• 1. Record the symptom and the pattern. Note exactly what happens, in which games, and after how long. Half the diagnosis is here.
• 2. Set everything to stock. Disable XMP or EXPO and clear any processor or graphics overclock or undervolt. Free, fast, and rules out the most common configuration causes.
• 3. Update Windows and clean-reinstall the graphics driver. Fully remove the old driver, then install the latest from the maker's site. This alone fixes a large share of black-screen and crash-to-desktop cases.
• 4. Watch the temperatures. Log with HWiNFO or run an overlay and play until it crashes. If a temperature is at its ceiling when it goes, clean the dust, check the fans, and consider fresh thermal paste.
• 5. Rule out storage and game files. Check drive health and verify the game's files.
• 6. Test the memory. Run MemTest86 for several passes. Any error means the settings or a module need attention.
• 7. Suspect the power supply. If temperatures are fine, memory is clean and drivers are fresh but it still cuts out instantly under load, the PSU is the leading suspect, especially after a graphics card upgrade or on an older or budget unit.

If you would rather not open the case, order matters, or you simply want it sorted properly, that is exactly the kind of methodical diagnosis we do every day.

How Manchester PC can help

Gaming crashes are one of the most satisfying faults to fix, because there is almost always a single part or setting at the bottom of it once you test in the right order. We diagnose the machine properly rather than throwing parts at it: we watch it under real load, check temperatures, memory and drive health, and test power delivery with known-good equipment. From there we replace or upgrade whatever turns out to be at fault, whether that is a stronger power supply, fresh cooling and thermal paste, new memory, a healthy SSD or a driver and Windows sort-out. To be straight with you, we do not carry out board-level or soldering repairs; where a fault is on the board itself we replace the faulty part or the whole board, or advise you on the best route.

We offer free local collection and return right across Manchester and the surrounding areas, so your PC can be picked up, diagnosed and brought back working without you leaving the house. Call us on 0161 820 1992 for a chat about what it is doing, or request a tech support quote and we will tell you the likely cause and a fair price to put it right.