100% GPU usage in games - help!

Soldato
Joined
18 Dec 2008
Posts
6,512
Location
Liverpool
So I'm having performance issues over the past few weeks and I really shouldn't be...

Okay so when playing games such as Fallout 4 or AC: Odyssey I'm averaging around 70 - 80fps on my current settings at first. Then at some point, may be 10 minutes in or 2hrs my 1080ti will suddenly hit 100% usage and the game will turn into a slide show at around 20fps or lower.

I've got minimal tasks running in the background, have made no recent changes to the system. I have DDU'd my drivers and it makes no difference if I run the card at stock settings or with my custom profile which is/was stable (110+ Core & 500+ Memory). Temperature on the card never goes above 80c even at full load.

I've a 9900k at 5Ghz which isn't breaking a sweat, highest cpu temp so far has been 74c. I don't know what's going on and could really use some suggestions.
 
Last edited:
Not sure if it's related, but on RE2R, when filters/textures were too high, the GPU memory wouldn't be enough, or something. Remember the colour going to red. But 11GB is plenty.
I would check the temperature of the M.2 drive, if you're using one. Long shot, but was my issue with the Vega 56 Pulse. Hot air blowing straight over the M.2.
 
Not sure if it's related, but on RE2R, when filters/textures were too high, the GPU memory wouldn't be enough, or something. Remember the colour going to red. But 11GB is plenty.
I would check the temperature of the M.2 drive, if you're using one. Long shot, but was my issue with the Vega 56 Pulse. Hot air blowing straight over the M.2.

Jay had an issue with ssd temps and explains it.

 
Untitled.png


Is my current profile with seemingly more scope to push it a bit further... Will check out the m.2 issue but why would it be happening a different times? If the gpu is putting out a consistent temp while in use.
 
60C is the first or the second line from HWiNFO, regarding the NVME? First line memory, second controller. Controller runs hotter than memory. If too high, thermal throttling.
 
Seems like thermal throttling could be an issue, hows the airflow in your pc?
 
I have 3 high speed 140mm Silent wings 3 fans up front, 3 high speed silent wings 3's on my silent loop 360mm AIO exhausting out the top and a further silent wing 3 in the rear for exhaust as well. Case is a P600s with just the mesh and fans are set to a pretty aggressive curve. Temps really shouldn't be a problem.

@MrPotato So looking at the sensor page on hwifno the maximum temp it got to on drive temp 2 (I'm assuming this is the controller) was 60c. As far as I understand it they tend to have problems around the 85c mark.

I have the z390 Aorus Master motherboard, the nvme is under the included heat spreader.
 
So I may be on to something after picking the brains of a friend earlier... GPU Memory clock maximum 6000Mhz minimum 375Mhz according to HWinfo. Now I'm not sure if this is happening when the frames drop or not yet but it's something to keep an eye on through afterburner in real time. Hoping it's something simple.
 
So I may be on to something after picking the brains of a friend earlier... GPU Memory clock maximum 6000Mhz minimum 375Mhz according to HWinfo. Now I'm not sure if this is happening when the frames drop or not yet but it's something to keep an eye on through afterburner in real time. Hoping it's something simple.

I've had it before where the core clock drops which then causes the stuttering, but not on this card. It was an AMD R290 issue and had to force the clock speeds while gaming
 
I've had it before where the core clock drops which then causes the stuttering, but not on this card. It was an AMD R290 issue and had to force the clock speeds while gaming

Is it something duff on the card then? Not sure how to force it to keep the memory clock up...
 
Is it something duff on the card then? Not sure how to force it to keep the memory clock up...

In my example, it would fluctuate causing stuttering, frame drops and performance issues. I can't remember the piece of software that I used to manage it, but it worked like a treat. You'd create a profile based on a game (*.exe) and then force it to keep the clocks up. It was almost magical, worked like a charm
 
So my problem continues...

I suspected it may be anti virus software causing issues as I'd seen a massive dip on gpu memory clock, so I uninstalled bit defender and thought this had solved it. However it's doing it again.

The only thing I can find in hwinfo this time is that CPU core 1 & 3 have hit a minimum of 800Mhz at some point, the rest remaining around the 5Ghz mark with little fluctuation. No temperature spikes or anything. I wouldn't mind but I've even lowered my setting again and have gone from around 90fps to 22fps in games with no changes made, game has been running for about 4 hrs though this time before it happened again.

@orbitalwalsh sorry to drag you into this one but have you got any ideas mate?
 
So my problem continues...

I suspected it may be anti virus software causing issues as I'd seen a massive dip on gpu memory clock, so I uninstalled bit defender and thought this had solved it. However it's doing it again.

The only thing I can find in hwinfo this time is that CPU core 1 & 3 have hit a minimum of 800Mhz at some point, the rest remaining around the 5Ghz mark with little fluctuation. No temperature spikes or anything. I wouldn't mind but I've even lowered my setting again and have gone from around 90fps to 22fps in games with no changes made, game has been running for about 4 hrs though this time before it happened again.

@orbitalwalsh sorry to drag you into this one but have you got any ideas mate?

dip in memory clock... only thing i can think of is fault memory controller/ ram chip on the GPU card .. they can fail, rare - but if you have i recording or VRAM speed decreasing then increasing should help fight for card RMA testing
 
Back
Top Bottom