2990wx Shutting down

[Ekladiuos]

Hi, Guys,
 I am recently facing a problem with my desktop, here is the build before:

- AMD Threadripper 2990WX 3 GHz 32-Core Processor
- Enermax Liqtech TR4 II 280 80.71 CFM Liquid CPU Cooler
- Asus ROG ZENITH EXTREME ALPHA EATX sTR4 Motherboard
- Corsair Vengeance LPX 64 GB (4 x 16 GB) DDR4-3600 CL18 Memory
- Samsung 970 Evo 1 TB M.2-2280 NVME Solid State Drive
- Asus GeForce RTX 2080 SUPER 8 GB STRIX GAMING Advanced Video Card
- Fractal Design Define XL R2 ATX Full Tower Case
- Corsair 1600 W 80+ Titanium Certified Fully-Modular ATX Power Supply

Before it was freezing and I have to force to restart from the button on the case since few updates ago stopped to happen but now it's shutting down suddenly without any load.

Your help is much-appreciated guys if you can help to solve this. I am so fed up with it I am already thinking to sell it

Update

after shutting down suddenly, when I turn it on again it makes the beep sound on the motherboard when it resets.

[PROH]

Check your Enermax cooler. It's known to fail after some time. Replace it with a Noctua air cooler.

[Ekladiuos]

It's Kinda new actually the workstation is few months old.

[PROH]

Might be enough, but check the CPU temps.

[maru]

There were similar cases in the past (typical symptoms: freezing / crashing which required unplugging the PC from power) and the solution was to either:
- Fully update Windows, all drivers, including the chipset driver, and especially update BIOS
- Replace the motherboard (it was too weak for this CPU - this is mostly about VRMs)

Examples:
https://forum.corona-renderer.com/index.php?topic=24900.0
https://forum.corona-renderer.com/index.php?topic=27699.0

[PROH]

@Maru - in one of the cases you're linking, the solution was to use the exact motherboard OP has. So I doubt it's the motherboard. Unless it's faulty.

[Ekladiuos]

There were similar cases in the past (typical symptoms: freezing / crashing which required unplugging the PC from power) and the solution was to either:
- Fully update Windows, all drivers, including the chipset driver, and especially update BIOS
- Replace the motherboard (it was too weak for this CPU - this is mostly about VRMs)

Examples:
https://forum.corona-renderer.com/index.php?topic=24900.0
https://forum.corona-renderer.com/index.php?topic=27699.0

I already checked the drivers, and I guess the motherboard model has one of the best vrm temp with 2900wx, maybe it;s defected i donno actually.
is there a way to check?

[Juraj]

You have the best motherboard in terms of power cascade, but it's Asus so... the quality control is what it is :- ). Unfortunately. Do you have the latest bios? Chipset drivers? (it has to be installed in that order, you need the former for the latter, esp. for stable 1903/2004 Windows 10 builds).

Enermax Liqtech can be faulty even from brand new but you can only tell by monitoring temperatures. With 2990WX, and Liqtech, they should barely if ever touch above 70.

What about the memory? Is it running XMP? Definitely don't try to run it faster than 2933/CL16. Definitely not 3200-3600. Even if it works for a while.
Is that PC unstable even with default (2133) frequency?

[Ekladiuos]

You have the best motherboard in terms of power cascade, but it's Asus so... the quality control is what it is :- ). Unfortunately. Do you have the latest bios? Chipset drivers? (it has to be installed in that order, you need the former for the latter, esp. for stable 1903/2004 Windows 10 builds).

Enermax Liqtech can be faulty even from brand new but you can only tell by monitoring temperatures. With 2990WX, and Liqtech, they should barely if ever touch above 70.

What about memory? Is it running XMP? Definitely don't try to run it faster than 2933/CL16. Definitely not 3200-3600. Even if it works for a while.
Is that PC unstable even with default (2133) frequency?

I will double-check the drivers, maybe I will install them again in the same order just in case.
regarding the memory, I was running it on 2933, I was not able to go higher than this, it doesn't boot. However yesterday I reset everything in the bios and it shut down again. it happened couple of times before but after the latest update of windows 10 it shuts down like every two hours without any load on the PC.

[Juraj]

Hmm, I still have one rendering node on 2990WX/Zenith Alpha running Windows 1903. It's fully stable but it has cheap memory that runs only 2800 at 128GB config but that's besides.

One questions: When it "shuts" down, can you just reset it with normal (or on-board) power button? Or do you have to turn power on PSU on/off? I remember having issues with this and it related to memory as well (hence why mine only runs on 2800).

[Ekladiuos]

Hmm well before it was freezing and the mouse was moving but nothing is functioning and i had to press the restart button on top of the case, i couldn't do it from the windows.

Now when it shuts down i just press the power button normally  from the case, it lights and  turns off and it lights again by itself making the reset sound of the board.

Regarding the reseting that i did yesterday, it was from the bios setting.I just set the default settings.

[Ekladiuos]

Update

Yesterday I started to render for 10 min and monitored the temperatures. Nothing was crazy and it didn't turn off. and after I left it for a while and nothing happened. I dunno but it's not consistent at all.
I started to doubt the MEMORY RAMS but how to make sure or to adjust it?

[Juraj]

You can check memory stability with MemTest. There are few, you can run it from bios or Windows. It's best to let it run for hours.

If it finds mistakes, it doesn't mean the memory is bad, just that it's not good at related working speed.

[Ekladiuos]

The result of the Ram Memory test. I really cannot think about other reasons. What else can cause a sudden shutdown?

[sebastian___]

Maybe you already did this, but what if you would take the desktop side door off, to make sure your computer is as cool as possible, even though in your last screen - your temp are not high at all. And then render something for a while.
 Maybe there's some weird obscure component on the motherboard or something which is not cooled enough.
 And maybe just to make sure, place there a room fan - maybe one with a leg if you have that - and point it to blow in the computer. This should ensure best temps for every component in there.