So a couple of months ago I assembled a custom PC for my wife, focused on 3D Modeling and Rendering. The specs are as follows:
- Intel Core i9 14900KF
- Asus ProArt Z790-Creator WiFi Motherboard
- Asus ProArt GeForce 4070Ti GPU
- Asus ROG Thor 1200W Platinum II 80 Plus PSU
- 2x Samsung Serie 990 PRO 2TB (configured for Raid 1 from BIOS)
- 4x 16GB Kingston Fury DDR5 (using XMP configured to 5600 MHz in BIOS, 64GB total RAM)
- Lian Li Lancool 216RC-W Case
- Lian Li GA2T36INW Galahad II Trinity 360 AIO
The PC is using genuine Windows 11 Pro and all hardware and software is up-to-date. A couple of days ago I flashed the motherboard's BIOS to the most recent version and I thought that might fix it, but today it happened again. All drivers are installed.
The workflow my wife uses is normally having either Autodesk Maya and Adobe Substance Painter, or Unity and Adobe Substance Painter or some other Adobe app open. They are resource-consuming, but the PC has more than enough horsepower for that. All software is genuine and up-to-date.
What happens is that at a random moment while no specific event is occurring, the screen just goes blank and all the case fans go full throttle. Nothing happens unless I force turn it off holding the power button. As for temperature, the CPU rarely goes above 60 and is normally in the 40-50°C range. Also, the case has abundant ventilation and the CPU is cooled with a 360 AIO.
I have tried having Task Manager open and the usage rarely comes to 50% of either RAM or CPU. I also think I overkilled it a little bit with the PSU as it has a voltage usage indicator and I think I've never even gotten to 50% usage, mostly in the vicinity of 300-400W usage.
I am kind of lost here as I am not sure what might be causing this. The other day I was just playing some Black Mesa with only Black Mesa and Steam open and it happened the same way, so my initial theory that it might be 3D Modeling app-related was immediately canceled.
Any idea as to what might be happening or what should I check?
23/May/2024 Update
So I followed the advice in the Tom's Hardware and what I did was put:
- long duration power limit to 280W
- short duration power limit to 300W
- CPU current limit to 350.00A
I put a little higher values, because the article was for a Core i7 and it shouldn't be an issue. There's also a config I don't remember where the value was Intel's FailSafe.
Also, a couple of days ago some Intel named updates came through Windows update, but it didn't say anything specific.
So, it still keeps happening, but there's an important difference I've noticed that makes me think the problem might be elsewhere.
So, before, when this happened and I was listening to something or a game, the audio got frozen too and I had to force-restart. But today, my wife was using the computer and having a google meet, she noticed that even though the screen went blank and fans went crazy, she was still able to hear and speak with the other people and also was able to save her work without having visuals.
So this lets me believe, could this be a GPU issue then? The GPU is updated with the latest GeForce studio drivers. My problem is I don't have another GPU at the moment so I have no way of testing this.
24/May/2024 Update
As suggested in the comments, I disabled Resizable BAR and it behaved well for a while, but I tried the 3D Mark test and it happened again. Afterwards I went back to the voltages and increased:
- long duration to 400W
- short duration to 450W
I then proceeded to open a couple of graphic demanding apps and games, ran the same 3D Mark test 2-3 times and the computer remained stable. So far so good, let's see if it will keep up.
