My ASUS TUF Gaming F15 FX507ZE has started crashing during games, usually after 5–15 minutes. The laptop has an Intel Core i7-12700H, RTX 3050 Ti Laptop GPU with 4GB VRAM, 24GB DDR5-4800 RAM configured as 1x16GB and 1x8GB, Intel Iris Xe graphics, a 144Hz display, BIOS 316, Windows 11, and the original ASUS charger.
The crashes happen in multiple games, including Forza Horizon 5 and Genshin Impact. Event Viewer reports nvlddmkm Event ID 153 with "Error occurred on GPUID: 100." Reliability Monitor also shows LiveKernelEvent 141 and 117. The bucket IDs reference nvlddmkm.sys and "Ampere_UserOC," although the GPU is running at completely stock settings and has not been manually overclocked.
I have already tried a clean Windows reset, DDU driver removal, several NVIDIA driver versions, the current Game Ready driver, BIOS and chipset updates, ASUS System Control Interface updates, Armoury Crate's Standard and Ultimate GPU modes, maximum-performance settings in NVIDIA Control Panel, removing background applications and peripherals, disabling NVIDIA audio drivers, SFC, DISM, different power configurations, and a different wall outlet.
A 10-minute OCCT 3D Adaptive test completed without errors. The GPU reached roughly 80–81°C and about 95W, so it does not seem to be a simple overheating problem. However, games still crash, sometimes during shader or game optimization on the first launch and sometimes after running normally for around 10 minutes.
The keyboard and fans were repaired or replaced shortly before these crashes became a regular issue, so I am also concerned about a possible physical problem involving the motherboard, GPU power delivery, PCIe connection, or another component. Interestingly, forcing maximum performance previously made the laptop last longer, but it did not prevent the eventual crash.
Could this combination of nvlddmkm Event 153, GPUID 100, and LiveKernelEvent 141/117 indicate a hardware fault even though OCCT passes? Or is it more likely to be related to NVIDIA drivers, ASUS firmware, Armoury Crate, power-state changes, VRAM instability, or the mismatched memory configuration?
1 Answer
The repair timing makes a physical issue possible, and a short synthetic test does not necessarily reproduce the constantly changing workload of a real game. Games repeatedly load and unload assets, change clocks and power states, and exercise different parts of the GPU and memory subsystem. Try a longer or different GPU workload such as 3DMark, Superposition, or FurMark while monitoring temperatures and power behavior with HWiNFO.
It is also worth uninstalling Armoury Crate with ASUS’s official removal tool and testing with G-Helper or with as little ASUS control software as possible. For troubleshooting only, you can try a small GPU core and memory reduction in MSI Afterburner, followed by power-limit tests around 95%, 90%, and 85%. If lowering clocks or power makes the crashes disappear, that would point toward marginal GPU, VRAM, or power-delivery stability rather than a normal driver installation problem.
Your 1x16GB plus 1x8GB memory setup is another variable. Remove the 8GB module temporarily and run the 16GB stick at standard JEDEC settings, then retest. Also check HWiNFO for CPU/GPU temperatures, hotspot readings, throttling, and power fluctuations.
I tested the Armoury Crate removal, G-Helper, reduced clocks, lower power limits, and the memory configuration separately, but unfortunately the crashes still occurred. Thanks for the suggestions.