PC Processors

[JamesCalnan1]

My GPU has crashed several times playing HALO Infinite. I have an AMD Radeon 5600 XT GPU and the CPU is an AMD Ryzen 9 3900X 12 Core. I'm not running anything in the background other than the AMD software to monitor the specs and temps while gaming. My GPUs Utilization while playing the game fluctuates between 96% to 99% while my CPU stays around 21%. I am trying to find a solution to prevent my game from crashing. Either fine-tune the GPU settings or tweak the CPU to take some of the load off the GPU while gaming. I am also not running in-game settings crazy high. I left the video settings at default. Any help will be appreciated.

[jakerules]

I'm going to second this post. I have been, for the past 3-4 days been dealing with random system lockup to where caps lock doesn't work, to driver crashes back to the desktop. I have tried uninstalling the game, running SFC and dism, this install of windows and drivers are pretty much clean as I reinstalled windows about a month ago, totally wiping my C drive. 

It seems to happen rarely on its own, but it still does. It almost always eventually happens when I have something open on my second screen, be it wallpaper engine, or some random application like teams, discord, chrome, or even file explorer. It happens with both the windows store version and steam.

I just posted about this as well. If you need my logs, reliability reports, whatever, hmu on that post and I will post them there.

[RPX100]

@JamesCalnan1 hey there. I am sorry to hear that you are having issues.

I believe you can get rid of your crashes by adjusting your GPU boost clocks and voltage.
Radeon Software's "automatic" performance profile will overclock your GPU by default.
-> This alone might already be the reason for your crashes.

And depending on silicon lottery your card might not be able to achieve the boost clocks.
You should find out the "gaming clock" that your manufacturer has set for your card,
and then switch to "manual performance tuning" (enable it) and change the max boost clock.

---
Also please note: 99% GPU utilization is more likely to trigger overclock related crashes.
So either turn down your boost clocks (for more stability) or enable vSync to reduce GPU usage.

Make sure the hot spot temperature stays below 90°C.
Temp delta between GPU core and hot spot should be within <15°C
Memory Temp should stay below 75°C

---
Yet another reason for crashes under load could be huge clock fluctuations.
These can happen due to energy saving features (deep sleep states).
You can fine tune / disable these "DS_" feature via MorePowerTool from igorsLab