• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Black Screen Hard Lock under Load

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.


Jan 18, 2009
Hello! I recently built a new PC for the first time in years. However I'm getting a rather annoying issue. Whenever I attempt to run something with a heavier load, I will get a complete system lock up with black screen. The GPU fans slow down as if the load is gone but the CPU fans and everything else act as if nothing has changed. I'm forced to hold the power button in order to reboot. Now: this only occurs in heavier load games/applications and it also seems to be they all use DX12 but that could be coincidence as in Red Dead Redemption 2, I get the same issue using DX12 or Vulkan.

Games it happens with: Red Dead Redemption 2 and Star Wars Squadrons within 5 minutes of starting a mission. 99% of the time. I tried Red Dead probably 50 times and did manage to get through the first mission before crashing, but typically it's less than 5 minutes from start to crash.
3D Mark Time Spy: Yesterday I managed to get through the entire benchmark but then attempted the stress test an it failed almost instantly. Usually though it fails within a minute of starting the benchmark.

Games that work: Fallout 4 Flight Sim 2020. Both I can play for hours with zero issues. Also Inventor and other applications which really don't use mush GPU power are completely fine.

What have I already tried?
- Complete removal and reinstallation of AMD drivers. Three times.
-Updated BIOS
-Re-seat the card and all power cables to PSU. (I do have 3 dedicated 8 pin cables for the card, no daisy chain connections)
-Windows, chipset drivers, and all games are up to date
-Different Display Port outputs and cables

I should also note that everything is out of the box, no overclocking or modifications on my end.

Motherboard: Asus Crosshair VIII Formula
CPU: AMD 5900X (Corsair H150i AIO)
GPU: Asus ROG Strix 6900XT (Factory AIO water cooler)
RAM: Corsair Dominator Platinum 64gb
PSU: Corsair RM850X
Windows 10 pro

Just trying to look for more options before I resort to sending in the card as I'd rather not be without my desktop for a month or more. I may buy another PSU to try narrowing it down to the card but I'd like to hear some opinions first. I can certainly provide pictures/video if needed.

If this is a load thing that's happening in a a stress test (read: non gaming) then id lean more towards the psu for now. I don't recall offhand what a 3dmark stress test does. Have you tried, and I hate to say it, furmark, to see if it happens on that?
How are your temps when the hard lock happens? I would agree with E_D and lean towards the PSU also. I just started having several semi hard locks. It happens when I stop gaming and shut the game off and get to desktop. Once on desktop I still get sound and mouse pointer moves, but I cannot close out from the game platform ie: Steam/Epic/Ubisoft.

I have to do ctrl/alt/del and sign out to be able to regain control.
Sound drivers updated? Move your RAM sticks around , if 4 try 2. At least install thema few times. New slots are tight , I have to push sticks in so hard on new builds I fear breaking the MOBO. After a few tries they seat better.
I have run fur mark actually, no issues. In fur mark as well as 3D mark, I never went over 61*C. I'm currently reinstalling windows and then I'll try taking out some RAM, I am utilizing all 4 slots. Since I am reinstalling windows, I will also get the latest sound drivers from Asus along with everything else (chipset, wifi, lan, etc)

My gut for no particular reason also suspects power so I'm considering picking up a 1000 or 1200 watt just to rule that out.
Well, the wattage amount isn't the issue here, but more of a chance of failure. 850W is plenty to run your listed system (may want to create a signature with your hardware so it travels with every post). I wouldn't go any higher, honestly. Surely not 1.2KW... just wasting cash at 1KW+ IMO.

But if you can run 3DMark or Furmark at the same time with a CPU stress test, then I doubt it's your PSU as that is about as heavy a load you can get on the system.

Speaking of hardware, when you create your sig, add a bit more detail... for example, the speed of you Dominator Plat RAM. Also, what speeds are you running it at? With all four slots populated, without some tweaking, I don't imagine it to run more than 3200-3600. And even then, you may need to add some voltage for it.
I have removed 2 sticks from the secondary slots and reinstalled all drivers etc. I will run a benchmark again.

BIOS says CPU core voltage = 1.449, CPU SOC Voltage = .968, DRAM Voltage = 1.192

However Bios says I'm at 2133MHz on the memory while the modules themselves are all rated 3600.
Which slots are they in, the two sticks, specifically? Just make sure they are in the correct ones (see manual).

It's normal for RAM to boot at the JEDEC settings (2133) for stability. 64GB should work there though.

Looking forward to hearing back on how two sticks do.......
Two sticks are in the two "primary" slots per manual and motherboard markings. I've also run every combination of my 4 sticks with identical results:

Running Time Spy (DX12) 10 times: first run crashed same as before. Second attempt completed, all subsequent crash at various points.
Running Fire Strike (DX11) on any of the levels including Ultra (4k) finishes perfectly fine.
Port Royal (Real Time Ray Tracing) also runs fine.
Fur Mark (2560x1440, msaa x8) runs just fine, temps don't exceed 62 and fans don't exceed 45%.

So is it safe to say this is DX12 dependent? Shouldn't a hardware issue (either card or power) be showing up on other stress tests as well?

Do any of the games you're having trouble with have an option to run in a different mode? Maybe try that?? I see rdr does the same thing in vulkan so...maybe it can't be pinned down.

Does windows event logs show anything?
I’ve found one event which references rd2.exe

Though the time stamp is minutes before the crash. There’s nothing at the exact time.


  • 5F38A7DC-6C0A-4562-AC0E-83E8D3D9CBD8.jpeg
    1 MB · Views: 3