[SOLVED] My RTX 3080 FPS suddenly went down

[brendz1993]

Hi guys. I hope someone could help me with my issue here. This may be a little long but I need to detail out everything.

At the time of this writing, there's a very noticeable fps drop on my games. I mostly play Monster Hunter World and Horizon Zero Dawn and my FPS usually runs for 110-150 at ultra settings (1440p). I also noticed that the power draw went down to 270-280w in Afterburner (previously at 320w)

Here's what I've been looking at as the possible reasons to what happened in my GPU

This past 3 weeks, I've been using my rig to mine ETH for 12 hours everyday (I mine while working from home. Also I work on my laptop so my gaming rig is not doing anything except mining). Core clock was reduced to -400mhz, power limit to 72%, memory clock +1000mhz
My rig is custom watercooled in an ambient temp of 25-26 degrees celsius. I think its worth mentioning that I played a few hours last week when I'm not in the mood to mine and everything is same as before.


Second possible reason is maybe the nvidia driver update? I updated my driver last Tuesday to the latest one (version 466.11) but I did not play any games until now. That's when I noticed the significant drop in my fps and power draw. I checked my Nvidia settings and it is set to Prefer Max Performance

What do you guys think? Is it the driver update or I'm now experiencing the drawbacks of gpu mining?

Below is my specs btw

i7 10700k
Zotac RTX 3080
16gb T-Force Delta 3200mhz
Lexxar 1TB nvme ssd
Corsair RM1000x PSU
Lian Li O11 Dynamic XL


Thanks in advance!

 

[white.a.drew]

Hi guys. I hope someone could help me with my issue here. This may be a little long but I need to detail out everything.

At the time of this writing, there's a very noticeable fps drop on my games. I mostly play Monster Hunter World and Horizon Zero Dawn and my FPS usually runs for 110-150 at ultra settings (1440p). I also noticed that the power draw went down to 270-280w in Afterburner (previously at 320w)

Here's what I've been looking at as the possible reasons to what happened in my GPU

This past 3 weeks, I've been using my rig to mine ETH for 12 hours everyday (I mine while working from home. Also I work on my laptop so my gaming rig is not doing anything except mining). Core clock was reduced to -400mhz, power limit to 72%, memory clock +1000mhz
My rig is custom watercooled in an ambient temp of 25-26 degrees celsius. I think its worth mentioning that I played a few hours last week when I'm not in the mood to mine and everything is same as before.


Second possible reason is maybe the nvidia driver update? I updated my driver last Tuesday to the latest one (version 466.11) but I did not play any games until now. That's when I noticed the significant drop in my fps and power draw. I checked my Nvidia settings and it is set to Prefer Max Performance

What do you guys think? Is it the driver update or I'm now experiencing the drawbacks of gpu mining?

Below is my specs btw

i7 10700k
Zotac RTX 3080
16gb T-Force Delta 3200mhz
Lexxar 1TB nvme ssd
Corsair RM1000x PSU
Lian Li O11 Dynamic XL


Thanks in advance!

I'm willing to bet it's the update. Due to nvidia trying instill strick no mining claws kinda "NOT REALLY" To prevent people from mining on the cards. It's propably detecting the mining softwate on the pc and throttling the gpu. It may not be that it could just be a faulty update that sunk the preformance of the cards that will propbably be resolved in the next update. But on of the things nvidia is trying to do is lock the max usage of GAMING cards installed on minning systems to prevent people from buying them all to mine on

 

[sizzling]

Have you been keeping an eye on VRAM temps? What are they now when gaming?

 

[brendz1993]

I'm willing to bet it's the update. Due to nvidia trying instill strick no mining claws kinda "NOT REALLY" To prevent people from mining on the cards. It's propably detecting the mining softwate on the pc and throttling the gpu. It may not be that it could just be a faulty update that sunk the preformance of the cards that will propbably be resolved in the next update. But on of the things nvidia is trying to do is lock the max usage of GAMING cards installed on minning systems to prevent people from buying them all to mine on

Thanks for the insight! That's probable. I'll try to uninstall my mining software and check if my fps goes back to normal. Appreciated!

 

[brendz1993]

Have you been keeping an eye on VRAM temps? What are they now when gaming?

Unfortunately there's no way for me to check. Even HWinfo has no temp monitoring for the VRAM. But according to other forums I read with the same card as mine, they tested and measured vram temps and it reached 95degrees. I'm assuminh that mine would be better than that since I have a watercooling setup and i have my gpu in a vertical mount. The 3 fans at the bottom of my case directly blow air to the backplate which could help to lower the vram temp itself.

 

[white.a.drew]

Unfortunately there's no way for me to check. Even HWinfo has no temp monitoring for the VRAM. But according to other forums I read with the same card as mine, they tested and measured vram temps and it reached 95degrees. I'm assuminh that mine would be better than that since I have a watercooling setup and i have my gpu in a vertical mount. The 3 fans at the bottom of my case directly blow air to the backplate which could help to lower the vram temp itself.

That and you underclockes the ram you said didn't you? So the vram should be fine In theory due to you did as you should as you overclock core MHz, yes I still agree could potentially still be the issue.... Not likely though I would still presume something else at fault at this point don't give up just yet... You are mining on a personal gaming rig which generally requires you to disable your firewalls... Have check to see if someone isn't hacking your system waiting for the right moment to steal your coins

 

[sizzling]

Unfortunately there's no way for me to check. Even HWinfo has no temp monitoring for the VRAM. But according to other forums I read with the same card as mine, they tested and measured vram temps and it reached 95degrees. I'm assuminh that mine would be better than that since I have a watercooling setup and i have my gpu in a vertical mount. The 3 fans at the bottom of my case directly blow air to the backplate which could help to lower the vram temp itself.

I’d try resolving why you cannot see the VRAM temperature. My 3080 I had to cripple it mining to keep it at a safe 90c or below, 95c is the max GDDR6X is rated for and I would not want to keep it their long. As for airflow over the back of the card I tried rigging fans a few different ways and it made no real difference. I have read of people adding additional heat sinks to the back of the card to help. My thought here is if excessive heat from mining has prematurely aged the thermal pads on the VRAM causing them to overheat even while gaming. If you haven’t kept an eye on vram temperature I think there a possibility of this regardless of the gpu being water cooled. Just a theory.

 

[white.a.drew]

I’d try resolving why you cannot see the VRAM temperature. My 3080 I had to cripple it mining to keep it at a safe 90c or below, 95c is the max GDDR6X is rated for and I would not want to keep it their long. As for airflow over the back of the card I tried rigging fans a few different ways and it made no real difference. I have read of people adding additional heat sinks to the back of the card to help. My thought here is if excessive heat from mining has prematurely aged the thermal pads on the VRAM causing them to overheat even while gaming. If you haven’t kept an eye on vram temperature I think there a possibility of this regardless of the gpu being water cooled. Just a theory.

Thats a fair point

 

[Afro_ninja199]

I'm willing to bet it's the update. Due to nvidia trying instill strick no mining claws kinda "NOT REALLY" To prevent people from mining on the cards. It's propably detecting the mining softwate on the pc and throttling the gpu. It may not be that it could just be a faulty update that sunk the preformance of the cards that will propbably be resolved in the next update. But on of the things nvidia is trying to do is lock the max usage of GAMING cards installed on minning systems to prevent people from buying them all to mine on

if it was the update andthey were throttling the GPU, wouldn't you be able to see that on HWiNFo etc?

 

[carocuore]

Thanks for the insight! That's probable. I'll try to uninstall my mining software and check if my fps goes back to normal. Appreciated!

You should uninstall the new drivers and reinstall the previous and see what happens, not uninstall the mining program

if it was the update andthey were throttling the GPU, wouldn't you be able to see that on HWiNFo etc?

No because the throttling is done at driver level (driver reports to monitoring soft), firmware if they've decided to sneak in an update with the drivers.

 

[brendz1993]

You should uninstall the new drivers and reinstall the previous and see what happens, not uninstall the mining program


No because the throttling is done at driver level (driver reports to monitoring soft), firmware if they've decided to sneak in an update with the drivers.

Noted. Will check if everything goes back once I downgrade the driver

 

[brendz1993]

That and you underclockes the ram you said didn't you? So the vram should be fine In theory due to you did as you should as you overclock core MHz, yes I still agree could potentially still be the issue.... Not likely though I would still presume something else at fault at this point don't give up just yet... You are mining on a personal gaming rig which generally requires you to disable your firewalls... Have check to see if someone isn't hacking your system waiting for the right moment to steal your coins

That's odd. I did not disable my AV and my firewall is up and running. I'm still getting 95-96mh/s without problem. Or maybe I missed something?

 

[white.a.drew]

That's odd. I did not disable my AV and my firewall is up and running. I'm still getting 95-96mh/s without problem. Or maybe I missed something?

Run a few anti virus programs. Hopfully they can catch something for you. This is generally the most common problem with people minning. Hidden malware.

 

[brendz1993]

Run a few anti virus programs. Hopfully they can catch something for you. This is generally the most common problem with people minning. Hidden malware.

So here's an update:

I used DDU to downgrade my nvidia driver to v465. Nothing changed. Power draw is still around 270-280w and FPS is still in 75-85fps. What's odd though is when I ran 3DMark Time Spy stress test, my gpu's power draw went up to 320w which is the TDP for an RTX 3080 and my FPS hovered around 120-130fps. The test ran for 1 hour and no drop in power draw and fps whatsoever. But when I play a game, power draw and fps goes down. I think I'm missing something here but I can't figure it out.

 

[sizzling]

So here's an update:

I used DDU to downgrade my nvidia driver to v465. Nothing changed. Power draw is still around 270-280w and FPS is still in 75-85fps. What's odd though is when I ran 3DMark Time Spy stress test, my gpu's power draw went up to 320w which is the TDP for an RTX 3080 and my FPS hovered around 120-130fps. The test ran for 1 hour and no drop in power draw and fps whatsoever. But when I play a game, power draw and fps goes down. I think I'm missing something here but I can't figure it out.

Ok that is strange and points to the problem being something else. Time Spy gpu test doesn’t stress the other components that much like a game would.

Can you try running userbenchmark and sharing the public link to the results. Not sure if it will show us what is possibly underperforming but is worth a try.

 

[brendz1993]

Ok that is strange and points to the problem being something else. Time Spy gpu test doesn’t stress the other components that much like a game would.

Can you try running userbenchmark and sharing the public link to the results. Not sure if it will show us what is possibly underperforming but is worth a try.

I'll be happy to do so. I'll post the link as soon as possible.

 

[brendz1993]

Here's the userbenchmark run: https://www.userbenchmark.com/UserRun/42147596

I updated the link. Found a quick fix for the Sphere GPU test error.

 

[white.a.drew]

Here's the userbenchmark run: https://www.userbenchmark.com/UserRun/42147596

I updated the link. Found a quick fix for the Sphere GPU test error.

Umm are you oc'ing?

 

[brendz1993]

Umm are you oc'ing?

CPU is OC'd @4.9ghz
GPU runs on stock when gaming. I just let the gpu boost do its job

 

[white.a.drew]

CPU is OC'd @4.9ghz
GPU runs on stock when gaming. I just let the gpu boost do its job

Okay so something is off onthe oc , look at user bench mark it's toping at 4.8 not 4.9 trying backing your oc down to 4.6 for a moment see if this makes any difference in performance. I'm wondering if it's not the GPU taking the abuse but the CPU

 

[brendz1993]

Okay so something is off onthe oc , look at user bench mark it's toping at 4.8 not 4.9 trying backing your oc down to 4.6 for a moment see if this makes any difference in performance. I'm wondering if it's not the GPU taking the abuse but the CPU

Did your suggestion and nothing changed. CPU utilization remains at around 35-50% regardless of the OC. I also flashed the GPU vBIOS and enabled Resizable BAR. I can't figure out why 3DMark can push my card to 100% power limit, but playing an actual game caps me at 80% only. I'm starting to think now that there might be something wrong with my card's power delivery caused by mining?

 

[brendz1993]

To close this thread, I would like to inform that I was able to resolve this issue already. Seemed like my gpu's vrm was overheating. I removed my waterblock and there's some oily residue on the vrms from the thermal pads. I changed it and everything's fine now. I can now play for hours without the power draw going down below 99%. Appreciate all the inputs you provided. I hope this helps to others that will encounter this issue in the future.