Weird Issue with GPU Crashing.

[AurorasChaos]

Hi guys,

I recently (Late January) built a new system, based on the 7600X, a B650 PG Riptide board, 16GB of DDR5 4800MHz memory(Going to be upgraded at a later date), and the RTX 3070. Alongside an ASUS MAG 650W psu.

The issue I've been experiencing is really inconsistent where the GPU seems to crash under high load, but only sometimes. I've attempted to replace the psu, run overnight memory checks, ensure the windows installation has no errors, running disk checks. I don't believe it's temperature related since the GPU gets up to 81C, and it shouldn't throttle until a higher temperature limit.  The issue manifests itself as the screen just going black, and attempting to restart the GPU driver through the shortcut plays the audio but still leads to a black screen. The event viewer doesn't spit out any errors relating the the GPU, and no blue screen events are being triggered.

 

So I'm perplexed at what the issue is, and where the issue may originate from. I have noticed that disabling the iGPU in device manager seems to help, but due to its inconsistency it's hard to tell if this has actually resolved the issue.

Any help will be massively appreciated.

[ComradeBard]

How are the cpu temps?

Also what kind of psu do you have, and what is the replacment you tried? 

A low quality power supply could get overloaded

[AurorasChaos]

The cpu doesn't go above 70C, and the PSU is a MSI MAG A650BN, which is a 80+ 650W, but the same issue was present with a Corsair CMX650 80+ Gold PSU.

I want to avoid having to buy another PSU, if possible due to having 2 now.