System Freezes During Gaming – Tried Everything, Still No Solution

xpto97 · May 17

Hi all,

I'm reaching out for help after two months of intensive troubleshooting. My PC is experiencing random system freezes, mostly during gaming, occasionally during light tasks.

The screen locks completely (frozen image), and the system becomes unresponsive, requiring a hard power-off via the case button.

System Specs

Case: NZXT H9 Elite (bought in 2023)
GPU: MSI GeForce RTX 4080 Gaming X Trio (bought in 2023)
CPU: AMD Ryzen 9 5900X (bought in 2022)
Motherboard: ASUS ROG Strix X570-F Gaming (ATX, bought in 2020)
RAM: 64GB (4×16GB) G.Skill Trident Z Neo RGB DDR4-3600MHz (bought in 2022)
PSU: Corsair RM1000X 80+ Gold, fully modular (bought in 2023, using OEM Corsair cables)
12VHPWR Cable: Premium Individually Sleeved 12+4pin PCIe Gen 5 12V-2x6 600W cable, Type 4, BLACK
Cooling: NZXT Kraken Elite 360 RGB AIO (bought in 2024)
Monitor Setup: Initially 3 monitors, now testing with just 1

Symptoms

Complete system freeze (static image), requiring hard reset.
Happens during gaming, occasionally during lighter loads.
No BSOD, no auto-reboot, no logs, just a hard lock.

What I’ve Tried So Far

Online solutions with similar problem.
Swapped RAM config: tested with 2 sticks, 1 stick, and different slots.
Reseated GPU several times, removed GPU support bracket.
Replaced and reconnected all PSU cables (OEM Corsair, new 12VHPWR cable)
Removed case side panels to test airflow.
Reduced monitor count to 1.
Disconnected all unnecessary USB devices.
BIOS updated and reset to default settings.
NVIDIA driver updated.
Tried all NZXT CAM cooling profiles (Silent, Normal, Performance)
Tested PSU multirail mode (though RM1000X is single-rail only)
Different power outlet.
MemTest86 all 4 RAM sticks (all passed), although it was 3 months ago.

Thermal Diagnostics

I bought a thermal gun to inspect temps across the system:

12VHPWR Cable: ~50 - 55°C under load.
VRMs: ~50 - 60°C
PSU cables & motherboard: All within normal thermal range.
CPU & GPU: ~70 - 75°C

I know a thermal gun isn't the most precise tool, but it offers a decent overview and was affordable. Nothing seems thermally abnormal...

However, the case panels where off in order to check the temperatures, with the panels on it could be different.

Stress Test Results

FurMark, 3DMark and OCCT 3D Adaptive: Pass without issues!
OCCT AVX2 Power Test:
- Fails (system freeze) after 20 - 60 minutes with case closed and airflow on Silent or Normal.
- Passes 12-hour test when side panels are removed and NZXT CAM is set to Performance mode.
- CPU/GPU temps remain steady at ~70–75°C under full load in both cases.

What seems to be working (Temporarily):
Oddly, the only way I can prevent crashes is by removing the case's back side panel and using the Performance cooling profile in NZXT CAM for better airflow but the temperatures seem fine!?

This suggests something airflow or heat-related, though temps appear normal. Could be VRM heat saturation or a flaw in thermal dissipation inside the NZXT H9 Elite?

I'm really lost and I don't know what else I can do, it's driving me crazy...

Spoiler

HWiNFO_20250501.CSV HWiNFO_20250502.CSV HWiNFO_20250502_2.CSV

xpto97 · May 17

Update

I observed that my top case fans were configured as intake, directing heated air towards the VRMs via the radiator. I have reconfigured them as exhaust fans, ~~but have yet to conduct thorough testing. I will provide updates on this thread regarding the efficacy of this modification in resolving the issue.~~

Solution

Top case fans (mounted with an AIO radiator) were configured as intake, pushing hot air from the radiator back into the case, right over the CPU socket and VRMs. This caused heat buildup in that area, especially with side panels on, despite all reported temps (CPU, GPU, VRMs) being within "safe" range. It was enough to cause freezes under sustained load like OCCT AVX2 or gaming.