Getting a lot of crashes after getting a new gaming PC, and it seems to be the GPU overloading in situations where it shouldn'

aeades247

Honorable
Apr 24, 2016
10
0
10,510
As the title describes, I'm getting a lot of crashes which appears to be the GPU being overloaded in situations where I think it should be fine.

Specs:

-Win 10 OS

- Z370 AORUS Gaming 7 motherboard

-Intel i7 8700k OC @4.4ghz

-Nvidia GTX 1070 (MSI afterburner tells the GPU clock is 1088mhz (+100) and the Memory clock is 4303mhz (+300), with a 120% power limit and temp limit of 92 degrees. I haven't touched these settings at all because I'm not at all familiar with them, and these were the stock settings post assembly)

-16gb (2x8gb) Corsair Vengeance LPX DDR4 RAM

-3tb Toshiba hard disk drive

-Corsair RM650x power supply

CPU and GPU are in an open loop water cooling system. Temperatures never exceed 50 degrees even post crash or long sessions of high usage,

I have two monitors, both 24'', one at 144hz (main) and one at 60hz (secondary), both different brands.

Issues occur on a few games I have played so far, examples include Farcry 5, PUBG, Rainbow 6 siege. It can run these on ultra settings with 100+ fps. These games are quite graphically intensive so crashing while I try and run them at max isn't unheard of.

But my issue is the same happens with FAR less graphically intensive games crashing in exactly the same way, namely Fortnite and TF2. When I look in task manager its always showing the GPU graph running at max or spiking to max just prior to crashing. Fortnite won't even let me unlock past 60fps, if its any higher it will crash in the menu given enough time, or without fail if going into a game. The difference I see is that my GPU runs at about 60% usage when locked to 60fps, and 100% any higher, however I read that having 100% GPU usage isn't a bad thing and should be happening?

I'm still wet behind the ears when it gets down to it, I've been PC gaming for 5-6 years now, most complicated thing I've ever done is fit a new GPU or a heatsink to a CPU. Up until now it was on a significantly less powerful rig. However, my old PC never crashed even when I pushed it past its limits, so I never had issues like this. It could run Fortnite at 120 fps steadily. My question to you is why is my new one crashing even with significantly less challenging workload comparatively, and in some cases, literally?

Thank you for reading my question, this issue has been irritating me ever since I got the thing. Any queries I will try to answer to the best of my ability. This is the most useful information I could think of and it is true to my knowledge, but if you can think of anything else that would be helpful I will provide that too.
 
Solution
email the mb vendor most time a b bios is a beta bios. it may not let you go backwards. also see from the mb vendor if there any new bios not posted.

R0GG

Distinguished
Did you try gaming with power limit at 100% instead, and with a temperature target a little low at 85 degrees (not too low to avoid card throttling).
You also could adjust nvidia control panel to run the card in more power efficient way like in adaptive mode, you'll loose some performance but you'll gain in stability.
Running a card at 100% is not a cause for card crashing, it usually overheating that makes card crash, visible reported temperature is not accounting for card internals temp like vrms and memory etc, I would suggest adjusting card fan curve to increase more rapidly RPM (more abrupt curve) and for improving air flow in the immediate surrounding of the card like placing extra 120mm with configurable speeds above and below card by the drive cage, which you can increase speeds while gaming.
 

aeades247

Honorable
Apr 24, 2016
10
0
10,510


Thanks for your reply, first thing I did when I got it was checked all the connections were secure. I checked my BIOS version and it was 'F6b' which strangely isn't listed, should I still download F5?.
 

aeades247

Honorable
Apr 24, 2016
10
0
10,510


Case already has 7 fans, 3 on the cooling radiator and 4 pointed directly into the case, so I cant imagine airflow would be a huge issue, additionally the GPU has a water block, I'll give it a shot though because I can't find a solution