Random fps spikes after new RTX 2080 install

[Nooting]

Hello, I recently had to RMA a gigabyte RTX 2080 due to a bad fan after 4 months of use. My new one came in 2 days ago and I used DDU to uninstall all old drivers and install the recent one. After everything was cleaned and installed I hopped into a match of Call Of Duty Black Ops 4 and about after 2 matches my fps spiked down to 40-50 fps for about 20 seconds before returning to my fps limit of 144. I also noticed my gpu temps were 84°C after only 10 minutes of playing. I set a custom fan curve in MSI afterburner and managed to maintain a somewhat stable 74-76° but the problem still persists. A few people said high gpu temps wouldn't cause fps spikes but considering I didn't have this problem with my old RTX 2080 I don't know what else it could be. I tested apex legends as well and while the fps didn't drop to the extreme like bo4 it had little skips. Bo4 is way more graphically demanding than apex though on maxed out settings. 

Specs:

Gigabyte RTX 2080

Water cooled Intel i7 7900k

16 GB ram

ASRock Z390 Phantom Gaming motherboard

Newly installed corsair 750watt 80+ Gold Certified PSU

 

Could this be an ssd problem?

 

[potato_man]

what's your GPU utilization? it may be a problem with your drive if it's overheating (3d-nand modules are very temperature sensitive); you could use aida64 to monitor your SSD temperatures.

[SpookyCitrus]

Your new GPU reaching 84c in a matter of 10 minutes is a serious issue, might want to contact the RMA department and see about that.  Also as @potato_man said you can check the  temps on your SSD aswell. I would suggest downloading Hardware Monitor and run some stress tests to see how hot everything is getting. It's very possible this is thermal related.

[jones177]

7 hours ago, SpookyCitrus said:

Your new GPU reaching 84c in a matter of 10 minutes is a serious issue, might want to contact the RMA department and see about that.  Also as @potato_man said you can check the  temps on your SSD aswell. I would suggest downloading Hardware Monitor and run some stress tests to see how hot everything is getting. It's very possible this is thermal related.

RTXs do that if you run them past 100% power limit.

 If the 2080 is at 100% it should not get past the 70s.

 

The RTXs are power limited but the limits are high for air. If I bought another one it would be water cooled. 

 

My RTX 2080 ti EVGA XC will run at 85c at 130% power limit and not drop clocks. At 100% it runs at 73c. This is a good thing because it means that it can be put on water and run full out. If it was limited to what the fans can handle there would be no point.  I keep my  XC at 130% to handle the spikes in 3D open world games. 

 

Even the 2080 ti FTW3 Ultra gets to hot at 124% power limit. Not the GPU but the ram. This is an issue in benches and not in games since I can only get to 124% with no AA at 1080p.

 

 

 

8 hours ago, Nooting said:

Hello, I recently had to RMA a gigabyte RTX 2080 due to a bad fan after 4 months of use. My new one came in 2 days ago and I used DDU to uninstall all old drivers and install the recent one. After everything was cleaned and installed I hopped into a match of Call Of Duty Black Ops 4 and about after 2 matches my fps spiked down to 40-50 fps for about 20 seconds before returning to my fps limit of 144. I also noticed my gpu temps were 84°C after only 10 minutes of playing. I set a custom fan curve in MSI afterburner and managed to maintain a somewhat stable 74-76° but the problem still persists. A few people said high gpu temps wouldn't cause fps spikes but considering I didn't have this problem with my old RTX 2080 I don't know what else it could be. I tested apex legends as well and while the fps didn't drop to the extreme like bo4 it had little skips. Bo4 is way more graphically demanding than apex though on maxed out settings. 

Specs:

Gigabyte RTX 2080

Water cooled Intel i7 7900k

16 GB ram

ASRock Z390 Phantom Gaming motherboard

Newly installed corsair 750watt 80+ Gold Certified PSU

 

Could this be an ssd problem?

 

 

The only way to test a RTX card is to run it stock and let the card determine the clocks and fan speed. I would also recommend using the Gigabyte software instead of MSI Afterburner. The Gigabyte software will tell you how much power the card is using in real time.

 

Here is my EVGA XC running at 100% at 74c and 130% at 84c. There is a slight overclock on the card and that is to get it up to FTW3 Ultra stock speeds.

The pics are in 4k so you can zoom in. I have ran my EVGA 2080 ti XC for hours at 85c waiting for it to downclock. It never has.

 

  

 

 

[Nooting]

2 hours ago, jones177 said:

RTXs do that if you run them past 100% power limit.

 If the 2080 is at 100% it should not get past the 70s.

 

The RTXs are power limited but the limits are high for air. If I bought another one it would be water cooled. 

 

My RTX 2080 ti EVGA XC will run at 85c at 130% power limit and not drop clocks. At 100% it runs at 73c. This is a good thing because it means that it can be put on water and run full out. If it was limited to what the fans can handle there would be no point.  I keep my  XC at 130% to handle the spikes in 3D open world games. 

 

Even the 2080 ti FTW3 Ultra gets to hot at 124% power limit. Not the GPU but the ram. This is an issue in benches and not in games since I can only get to 124% with no AA at 1080p.

 

 

 

 

The only way to test a RTX card is to run it stock and let the card determine the clocks and fan speed. I would also recommend using the Gigabyte software instead of MSI Afterburner. The Gigabyte software will tell you how much power the card is using in real time.

 

Here is my EVGA XC running at 100% at 74c and 130% at 84c. There is a slight overclock on the card and that is to get it up to FTW3 Ultra stock speeds.

The pics are in 4k so you can zoom in. I have ran my EVGA 2080 ti XC for hours at 85c waiting for it to downclock. It never has.

 

  

 

 

Thanks for your reply, so I should download the gigabyte software and hit the reset button or whatever it is that uses default settings and see what happens?

[Nooting]

11 hours ago, potato_man said:

what's your GPU utilization? it may be a problem with your drive if it's overheating (3d-nand modules are very temperature sensitive); you could use aida64 to monitor your SSD temperatures.

my gpu in bo4 is 80-90%. Sometimes I noticed it matching my cpu utilization at 60% sometimes 70%

[SpookyCitrus]

44 minutes ago, Nooting said:

Thanks for your reply, so I should download the gigabyte software and hit the reset button or whatever it is that uses default settings and see what happens?

Usually overclocks and power limits and settings are saved in the software you use, did you previously overclock your other card? I would remove any software you have for overclocking the GPU off of the computer and then see. Also usually with RMA the card you recieve is another unit that the RMA department recieved and "fixed" sometimes they send a card that is worse than the one you RMA'd. But the units they send back are usually refurbished units, and if your card isn't over voltage and is running at stock settings then I would definitely contact them about it. Because for stock settings on any card shooting up to 84c in that short of time is bad. Another thing that could contribute to the temps is your case and how you have your cooling set up.

[Nooting]

1 hour ago, SpookyCitrus said:

Usually overclocks and power limits and settings are saved in the software you use, did you previously overclock your other card? I would remove any software you have for overclocking the GPU off of the computer and then see. Also usually with RMA the card you recieve is another unit that the RMA department recieved and "fixed" sometimes they send a card that is worse than the one you RMA'd. But the units they send back are usually refurbished units, and if your card isn't over voltage and is running at stock settings then I would definitely contact them about it. Because for stock settings on any card shooting up to 84c in that short of time is bad. Another thing that could contribute to the temps is your case and how you have your cooling set up.

I think my cooling is fine because my other card was pretty cool in the same exact set up. And ive heard from multiple people that gigabyte is notorious for sending broken refurbished cards, even without my side panel on the card was getting up to 77-78°. I've requested a full refund and bought an MSI Gaming GeForce RTX 2080 8GB GDRR6 256-bit VR Ready Graphics Card (RTX 2080 GAMING X TRIO). I'll keep everyone posted to see if the problem is still there.

[SpookyCitrus]

26 minutes ago, Nooting said:

I think my cooling is fine because my other card was pretty cool in the same exact set up. And ive heard from multiple people that gigabyte is notorious for sending broken refurbished cards, even without my side panel on the card was getting up to 77-78°. I've requested a full refund and bought an MSI Gaming GeForce RTX 2080 8GB GDRR6 256-bit VR Ready Graphics Card (RTX 2080 GAMING X TRIO). I'll keep everyone posted to see if the problem is still there.

Alright hope it works out for you. Update us once you get the new card.

[Nooting]

Update: The first MSI RTX 2080 i bought at best buy was also defective straight out of the box. Both my monitors said no signal. I returned it and had them confirm that it was indeed defective and couldn't get it to work. Go figure 3 gpus in a row!! my luck with computer parts are trash! Thankfully they had one more left in stock so I snagged it and installed it in my pc and managed to get both monitors up and running. I opened up my MSI OSD and launched Bo4 and after one match my temps didn't go past 60°c! So buyers beware of the GIGABYTE WINDFORCE 8G Geforce RTX 2080. Thank you to everyone who told me it was more than likely another dud.