PC freezing under load - causes and fixes

Idle, fine. Open a game or a render and it locks up solid. No blue screen, no error, just a frozen image and a hard reset. This is one of the trickier faults to diagnose because several different components can cause identical symptoms.

1. Check your temperatures first

This is always the starting point. Download HWiNFO64 or MSI Afterburner and watch your CPU and GPU temperatures under load. General danger zones:

• CPU above 90-95°C - thermal throttling or shutdown
• GPU above 95°C - thermal throttling, artifacts, or hard freeze
• VRM temps above 110°C - a less visible but equally serious problem

If temperatures are spiking before the freeze, you've found your culprit. The fix is usually dried-out thermal paste (replace it), a blocked heatsink (clean it), or a failing fan that isn't spinning up under load.

2. Test your RAM

Bad or mismatched RAM is a classic cause of load-triggered freezes. Under idle conditions your system might only be using 4-6GB - push it with a game and suddenly it's pulling from a stick with a fault.

Run MemTest86 overnight. It boots from a USB drive and tests memory independently of Windows. Any errors at all mean the RAM needs replacing. If you have two sticks, try each one alone to isolate which is faulty.

Also check that your XMP/EXPO profile is enabled in the BIOS. Running RAM out of spec can cause instability that only shows up under load.

3. Check the GPU

A GPU that's starting to fail will often hold together at desktop but fall apart the moment it's pushed. Signs it's GPU-related:

• Freeze happens specifically when the GPU hits high load
• You see artifacts (flickering, coloured blocks, screen corruption) just before the freeze
• The system is stable using integrated graphics

Run FurMark or 3DMark and watch for artifacts or a freeze within minutes. If it freezes faster at higher GPU load, the card is the likely cause. This can be a dying VRAM chip, a failing power delivery stage on the card, or solder joint failures on older GPUs.

4. Check PSU stability

A PSU that can't sustain clean power under load is a very common cause of hard freezes with no error message. The system draws significantly more power under load, and if the PSU is undersized, degraded, or has a failing capacitor, voltage can sag enough to cause an immediate shutdown or freeze.

If your PSU is more than 4-5 years old and you've upgraded your GPU since buying it, this is worth taking seriously. There's no reliable way to test a PSU under load without specialist equipment - a multimeter at idle will show fine voltages even on a failing unit.

5. Rule out storage

A failing SSD or HDD can cause freezes when the system tries to stream data - common in open world games or any software doing heavy disk reads. Check CrystalDiskInfo for your drive's health status. Any warnings or reallocated sectors on an HDD are a sign it's on the way out.

Also check for driver and software issues. A corrupted GPU driver can cause load-triggered freezes that look identical to hardware faults. Do a clean driver reinstall using DDU (Display Driver Uninstaller) in safe mode before assuming the hardware is at fault.

When to stop and send it in

If you've ruled out temperatures, RAM, software, and storage, and the freeze is still happening - the most likely remaining culprits are a failing GPU, a degraded PSU, or in some cases a VRM fault on the motherboard that only shows up under sustained CPU or GPU load.

These all require bench testing under controlled load to isolate properly. We can stress-test each component individually, measure PSU rail voltages under load, and identify whether a GPU fault is repairable at component level or needs replacing.