Microstuttering in almost every game with RTX 3080

[Chris Redfield]

Hello everybody,

 

I have problem that lasts since I got my RTX 3080.

I experiencing constant microstuttering in almost every game despite having good PC.

For some games it happens every 7-10 seconds and for some like Resident Evil 3 Remake and Resident Evil Village it happens every 2-3 seconds constantly.

I tried following things:

 

Uninstall drivers with DDU

Turn off HPET

Set Nvidia CP to High Performance

Delete Nvidia Shade Catche

Scan PC from malwares

Install games on HDD and SSD

Turn fans to 90%

Clean C:Driver

Reinstall Windows after I bought SSD

 

I'm playing on 4K@60fps monitor and I tried various settings like turning down graphics,turning off and on V-sync,changing resolution and nothing helped except lowering resolution in Resident Evil Village from 4K to 1440p.

That seems like it solved problem regarding microstuttering from what I've seen (I've played for few minutes and I didn't noticed stuttering).

 

However,my PC should handle those games much better since it's above recommended settings and other played it without any problem.

Here are my PC specs:

Desktop:
OS:Windows 10 x64

Monitor:LG  27UD58-B

CPU:Intel i7 10700KF

CPU Cooler:Arctic Liquid Freezer II 360

MBO:MSI MAG Z490 Tomahawk

GPU:EVGA GeForce RTX 3080 XC3 ULTRA GAMING

HDD:1 TB WD Blue

SSD:256 GB ADATA SX8200 ASX8200PNP-256GT-C, PCIe NVMe, M.2

RAM:2X8 RAM Corsair Vengeance 3200mhz  (XMP enabled)

PSU:Seasonic FOCUS 750 Gold

Case:SilentiumPC Signum SG7V EVO TG ARGB

 

Max GPU temperature while gaming was 78c and memory junction temp was max 92c.

I have Windows and GPU drivers updated.

 

Could you please help me with this problem,it's happening constantly and I really don't know what else to do.

 

 

 

 

[DobertRownySr]

Install GPUZ and report back what your VRAM temps are.

[Chris Redfield]

4 minutes ago, DobertRownySr said:

Install GPUZ and report back what your VRAM temps are.

I have installed HwInfo64 that has new feature for measuring memory junction temperatures for 3080 and 3090.

Max temp for memory was 92c.

[DobertRownySr]

Just now, Chris Redfield said:

I have installed HwInfo64 that has new feature for measuring memory junction temperatures for 3080 and 3090.

Max temp for memory was 92c.

Is your bootdrive an SSD or HDD? Also, where are your Nvidia Drivers installed, on an SSD or HDD?

[Chris Redfield]

Just now, DobertRownySr said:

Is your bootdrive an SSD or HDD? Also, where are your Nvidia Drivers installed, on an SSD or HDD?

 My bootdrive is SSD that has both Windows and Nvidia Drivers installed on it.

[DobertRownySr]

Just now, Chris Redfield said:

 My bootdrive is SSD that has both Windows and Nvidia Drivers installed on it.

What model 3080 do you have? Also, what overlock are you running it at?

[DobertRownySr]

2 minutes ago, Chris Redfield said:

 My bootdrive is SSD that has both Windows and Nvidia Drivers installed on it.

 

1 minute ago, DobertRownySr said:

What model 3080 do you have? Also, what overlock are you running it at?

Sorry, I just realized you have the model in your original post. However, I'm still curious what OC you're running

[Chris Redfield]

Just now, DobertRownySr said:

What model 3080 do you have? Also, what overlock are you running it at?

I have EVGA GeForce RTX 3080 XC3 ULTRA GAMING and I usually have overclock on it for mining,but I turn it off every time I'm not mining.

 

[DobertRownySr]

1 minute ago, Chris Redfield said:

I have EVGA GeForce RTX 3080 XC3 ULTRA GAMING and I usually have overclock on it for mining,but I turn it off every time I'm not mining.

 

Have you tried undervolting? I have a 3090FE, the stock clock had some frame dips, but after I undervolted the card, it fixed the issue.

[Chris Redfield]

1 minute ago, DobertRownySr said:

Have you tried undervolting? I have a 3090FE, the stock clock had some frame dips, but after I undervolted the card, it fixed the issue.

I haven't tried undervolting since I don't have knowledge to do it,but did you had microstuttering or just occasional frame drops because of demanding games?

I should easily run this games with +85 FPS on 4k,so I don't think these are ''standard'' frame drops that happens when some game is very demanding and FPS hoovers around 60-ish and then drops in graphically demanding scenes.

To me it happens ever few seconds no matter how scene or game is demanding.

[DobertRownySr]

Just now, Chris Redfield said:

I haven't tried undervolting since I don't have knowledge to do it,but did you had microstuttering or just occasional frame drops because of demanding games?

I should easily run this games with +85 FPS on 4k,so I don't think these are ''standard'' frame drops that happens when some game is very demanding and FPS hoovers around 60-ish and then drops in graphically demanding scenes.

To me it happens ever few seconds no matter how scene or game is demanding.

For me, it was mainly just framedrops as opposed to stuttering, but it still might help. I will link a video on how to undervolt (it is very easy and has zero risk of doing any sort of damage)

 

Furthermore, you said you had a mining overclock. How hot did your memory and gpu get underload with that OC, and how long did you run the card when mining?

 

 

[Chris Redfield]

1 minute ago, DobertRownySr said:

For me, it was mainly just framedrops as opposed to stuttering, but it still might help. I will link a video on how to undervolt (it is very easy and has zero risk of doing any sort of damage)

 

Furthermore, you said you had a mining overclock. How hot did your memory and gpu get underload with that OC, and how long did you run the card when mining?

 

 

I had this settings on MSI Afterburner while mining:

-130 Core Clock

+1000 Memory Clock

69% Power Limit

88% Fan Speed

 

As for temperature of GPU while mining with these settings,I'm not sure I know it was in safe limits for mining as far as other people with RTX 3080 reported,so I always focused on memory temperature.

Memory temperature is around 85c average,depending on ambient temperature,but it never went above 94c and was almost always at 92c max.

I run the card for around 1 month constantly as the guys told me that it won't hurt GPU because it runs at constant frequency so it's less stressful for PC than gaming for around 6 hours per day and then turning PC off as long as I keep temperature low.

I started to mine on Feb. 22 until today and I turned off PC few times in that period.

 

 

[Chris Redfield]

This may seem like a dumb question,but would changing thermal paste help with this problem?

I read that guy had 60c on his RX 480 that's been previously used for mining and he didn't noticed any thermal throttling and he mange to get PC stable with undervolting,but once he change settings to stock,he had stuttering again,so he changed thermal paste and his stuttering stopped.

 

I know I have card that is only 6 months old and was used for mining since Feb 22. but is there a chance that paste degraded in such a short time?

[DobertRownySr]

17 minutes ago, Chris Redfield said:

This may seem like a dumb question,but would changing thermal paste help with this problem?

I read that guy had 60c on his RX 480 that's been previously used for mining and he didn't noticed any thermal throttling and he mange to get PC stable with undervolting,but once he change settings to stock,he had stuttering again,so he changed thermal paste and his stuttering stopped.

 

I know I have card that is only 6 months old and was used for mining since Feb 22. but is there a chance that paste degraded in such a short time?

Unlikely, your temps are well within safe limits.I mean, you're more than welcome to try, but I wouldn't bet on it helping.

[Chris Redfield]

1 minute ago, DobertRownySr said:

Unlikely, your temps are well within safe limits.I mean, you're more than welcome to try, but I wouldn't bet on it helping.

Yeah,I agree completely,I only thought of that solution because that guy had same issue with stuttering like I do and despite having ''good'' temperatures with his GPU he manage to solve his problem.

[DobertRownySr]

Just now, Chris Redfield said:

Yeah,I agree completely,I only thought of that solution because that guy had same issue with stuttering like I do and despite having ''good'' temperatures with his GPU he manage to solve his problem.

Yeah, try it. It might help. You might as well replace the thermal pads too if you do it.

[Chris Redfield]

Just now, DobertRownySr said:

Yeah, try it. It might help. You might as well replace the thermal pads too if you do it.

OK,I'll give it a shot,I was planning to buy thermal pads too,since EVGA allows mods like that without void of warranty.

Would you recommend Arctic MX-5  paste for my GPU?

Also,is original thermal paste that comes with GPUs high quality or is it one of  the ''cheaper'' ones that is used for mass production?

[DobertRownySr]

6 minutes ago, Chris Redfield said:

OK,I'll give it a shot,I was planning to buy thermal pads too,since EVGA allows mods like that without void of warranty.

Would you recommend Arctic MX-5  paste for my GPU?

Also,is original thermal paste that comes with GPUs high quality or is it one of  the ''cheaper'' ones that is used for mass production?

I just replaced my 3090 thermal pads (because the VRAM was hitting 110C, dropped to 90C after replacing the pads). While I did that, I also repasted. 

For the thermal paste I bought Thermal Grizzly Hydronaut

For Thermal Pads, I used: Thermalright Extreme Odyssey Thermal Pad with a thickness of 1.5mm (Because that's what the FE version used for thickness)

Arctic MX-5 is also good, but I haven't used it on a 3000 series gpu, so you would need to look up thermal benchmarks for an honest opinion.

I think the original thermal paste is pretty high quality. Because my GPU core temps haven't changed very much after repasting. In fact, they might have gone up 1 or 2 degrees because the new pads are cooling the memory a lot better.

[Chris Redfield]

18 minutes ago, DobertRownySr said:

I just replaced my 3090 thermal pads (because the VRAM was hitting 110C, dropped to 90C after replacing the pads). While I did that, I also repasted. 

For the thermal paste I bought Thermal Grizzly Hydronaut

For Thermal Pads, I used: Thermalright Extreme Odyssey Thermal Pad with a thickness of 1.5mm (Because that's what the FE version used for thickness)

Arctic MX-5 is also good, but I haven't used it on a 3000 series gpu, so you would need to look up thermal benchmarks for an honest opinion.

I think the original thermal paste is pretty high quality. Because my GPU core temps haven't changed very much after repasting. In fact, they might have gone up 1 or 2 degrees because the new pads are cooling the memory a lot better.

How did you manage to reach such a high VRAM temperatures as of 110c?

Did you used it for mining or for gaming?

[OCNewbee]

The common causes for microstuttering are:

1. ram

2. swap file 

3. Undervolted gpu

4. Thermal throttling

 

[DobertRownySr]

1 hour ago, Chris Redfield said:

How did you manage to reach such a high VRAM temperatures as of 110c?

Did you used it for mining or for gaming?

Gaming, the FE's stock thermal pads are genuinely garbage.

[Chris Redfield]

16 minutes ago, OCNewbee said:

The common causes for microstuttering are:

1. ram

2. swap file 

3. Undervolted gpu

4. Thermal throttling

 

1.Would memtest64 or memtes86 show if I have faulty RAM, and would it help if I swap RAM sticks in different sockets?

2.I'm not sure what you mean by swap file and how to solve it?

3.I didn't undervolted my GPU yet

4.My temperatures seems fine from what I saw.I wouldn't say I'm having thermal throttling.

[OCNewbee]

8 hours ago, Chris Redfield said:

 

1.Would memtest64 or memtes86 show if I have faulty RAM, and would it help if I swap RAM sticks in different sockets?

its improbable that you have a faulty ram. That would give your a BSOD. Try different sockets.

 

8 hours ago, Chris Redfield said:

2.I'm not sure what you mean by swap file and how to solve it?

Swap files are page files. Page files should be set to auto or double the size of your total ram capacity. For example, 32gb would necessitate minimum of 64GB. This 64GB page file should be located on your SSD. Also, your games should be on SSD as well.

https://mcci.com/support/guides/how-to-change-the-windows-pagefile-size/

 

8 hours ago, Chris Redfield said:

3.I didn't undervolted my GPU yet

Try overvolting the GPU. Many times, stutters happen due to insufficient power. In the same vein, your PSU might be the problem. Check whether your PSU is good enough. Check for total Watt and whether it has a minimum of gold rating or not.

 

8 hours ago, Chris Redfield said:

4.My temperatures seems fine from what I saw.I wouldn't say I'm having thermal throttling.

What is your CPU when running p95 small fft for 30minutes? If your temp is above 85 C, this is a good indicator of a thermal throttling.

[Chris Redfield]

Just a little update.

I found a solution for stuttering and now game run perfectly smooth.

It seems like I needed to set Vertical Sync in Nvidia Control Panel to ''On''.

I also enabled CPU Boost in BIOS and viola stuttering was completely gone.

[OCNewbee]

On 5/9/2021 at 5:07 AM, Chris Redfield said:

Just a little update.

I found a solution for stuttering and now game run perfectly smooth.

It seems like I needed to set Vertical Sync in Nvidia Control Panel to ''On''.

I also enabled CPU Boost in BIOS and viola stuttering was completely gone.

congrats.

I was gonna share this with you

https://forums.tomshardware.com/threads/stutter-freezing-only-whilst-playing-1-game.3601359/

 

It seems that for some games NVcache is messed up. This causes 2~3 seconds of pauses. Anyways good thing that everything is sorted. Happy gaming.