• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

FPS dropping everytime GPU watts drop?

yumerifter

New Member
Joined
Jun 15, 2023
Messages
4 (0.01/day)
Hi, I hope this is in the right category,
I don't know what's wrong with my computer
Everytime I play a game (any game, could be even a pixel game) I get fps drops but for a slight period time, I'd say around 30-40 seconds. Then it goes back to normal for a few minutes, then I get fps drops again and the loop continues.

I took a pair of screenshots with Bendy and the Ink Machine as an example since I had it open, but like I mentioned this happens with every game so just take this game as an example, I could try running any other if needed but it gives me the same problem

As you can see here I have 120 fps here and 129 watts
1bendy.png


But then after a bit, the watts are 45 and I get 25 fps (at the moment of the screenshot.. because I've been at lower than that)

2bendy.png



Then it goes back to 120 fps and with 129+ watts.

Currently in those screenshots my GPU is undervolted + with medium graphics in-game settings even though I can run the highest to see if that would fix it, but it didn't, this problem has been present even before I tried undervolting it.

I've tried:
-overclocking
-changing the NVIDIA settings to highest performance instead of optimal
-changing the Windows power settings to balanced to highest performance
-Switching the HDMI cable in the monitor from both ends (lol)
-changing the positions of my frontal fans to have 3 on the front (I used to have 2 on the front and one above my CPU)
-running everything on default NVIDIA settings again, etc... and the problem persists...

Here is another screenshot from TechPowerUp GPU-Z
gdfgdfggh.gif



I've had this computer for only 1 year. All of my specs (and other pc parts just in case) are:

OS: Windows 11
CPU: Intel Core i5-12400F
GPU: Nvidia GeForce RTX 2060 6GB GDDR6
RAM: 16GB (XPG SPECTRIX D50 8GB x2)
SSD: 480 GB (M.2 Nvme WD GREEN SN350)
SSD 2: 1tb (WD Green 2.5)
MB: Biostar H610MH
PSU: GAME FACTOR 650W / 80+ Bronze / PSG650

And here's my UserBenchmark

https://www.userbenchmark.com/UserRun/62122805


Could anyone please tell me what could be wrong or what I could try to fix this? :confused: Thank you
 
Last edited:
I see two issues there:
1. 11.3V ? of that is correct, then it might be an issue with the PSU. Do you have a spare one to test with ?
2. The CPU is at 91ºC while gaming, that is a lot, on those spikes i guess it goes over 100 ? Check to see if it is mounted properly. if that dose not void your warranty and it was pre-build.
 
Last edited:
I see two issues there:
1. 11.3V ? of that is correct, then it might be an issue with the PSU. Do you have a spare one to test with ?
2. The CPU is at 91ºC while gaming, that is a lot, on those spikes i guess it goes over 100 ? Check to see if it is mounted properly. if that dose not void your warranty and it was pre-build.

^^ This. 11.4 depending on how accurate gpu-z is report is very close to shutdowns.

You could all so click the boxes or the arrows in GPU-Z and show the max numbers too.

Get a quality PSU to start with as they are not worth cheeping out on. What case are you using ?.
 
ATX Spec 12V is 5%(+-) 5V and 3 are the same, the psu is toast. The cpu is overheating

Max 2060 temperature is 88 degrees before throttling. You need to get a hwinfo64 and run it in background with sensors and have a gaming session to see what max temps and volts are
 
Remember software voltage rail sensors are notoriously untrustworthy though.
 
Ty for the replies everyone

I see two issues there:
1. 11.3V ? of that is correct, then it might be an issue with the PSU. Do you have a spare one to test with ?
2. The CPU is at 91ºC while gaming, that is a lot, on those spikes i guess it goes over 100 ? Check to see if it is mounted properly. if that dose not void your warranty and it was pre-build.
1. Sadly no I don't have another PSU to test with, do you think I should invest in another one of higher W?
2. Yes it has hit 100ºc, haven't seen it go past that. My warranty expired like a month ago so I think I'm fine with tampering with it now. I noticed it also has intel's default CPU cooler...

^^ This. 11.4 depending on how accurate gpu-z is report is very close to shutdowns.

You could all so click the boxes or the arrows in GPU-Z and show the max numbers too.

Get a quality PSU to start with as they are not worth cheeping out on. What case are you using ?.

This is what it gave on max
max.gif


The case is rgb munfrost shine, looks like this (not my pic)
1686908212903.png


ATX Spec 12V is 5%(+-) 5V and 3 are the same, the psu is toast. The cpu is overheating

Max 2060 temperature is 88 degrees before throttling. You need to get a hwinfo64 and run it in background with sensors and have a gaming session to see what max temps and volts are

Sorry I don't understand much about PSUs, is that bad?

I reset all settings on NVIDIA control panel again, reset MSI Afterburner to stop the undervolting to run the tests on how I'd usually play + ran hwinfo64 this is the result from the sensors

1. Ran it in the background playing Bendy for ~50 minutes
2. First 5 minutes were fine. Then frames dropped heavily again for 1 whole minute, after that 1min the remainig minutes were stable fps
4. I closed the game to save the report but I found strange that I didnt lag.

Report from hwinfo64:
https://www.dropbox.com/s/b4i1k0zczmeciu5/1.CSV

5. Opened the game and hwinfo64 again (for only ~1 minute):

ss1.png


It's weird to me that I didn't lag while taking that report since in the evening I hosted and played a game through Parsec, and the problem happened from the very beginning. Played for 3.6h. The game was fine, then lagged, then fine, and so on till the point it became unplayable (it wouldn't go back to being stable).

Before that, I didn't even use my computer to do anything else other than to use Chrome. The local weather at the time I was playing was probably around 35ºc at 9 PM. At the moment I started playing Bendy for hwinfo64 was around 2-3:30 AM when the local weather was 26ºc. Now I'm starting to think that besides my CPU burning it might be the weather partaking in this since I don't have AC in my place and my whole thing is on fire :eek:
 
Pretty much everything in your system seems to be overheating, what's the ambient temps in the room & I guess cooling in the rest of the case?
 
Pretty much everything in your system seems to be overheating, what's the ambient temps in the room & I guess cooling in the rest of the case?

Pretty hot ngl. I'd say around 30-35ºc on average. For cooling on the case I have the default fan that came with the case inside + 3 additional fans on the front since that case doesn't allow liquid cooling
 
Not sure how good the boxed cooler is that came with the CPU, but if it's always been like that then you need to try to reseat it once & see if that changes temps or get a better cooler. Depending on where you are even a decent $20 one should do for that chip.
 
Not sure how good the boxed cooler is that came with the CPU, but if it's always been like that then you need to try to reseat it once & see if that changes temps or get a better cooler. Depending on where you are even a decent $20 one should do for that chip.

Thank you for your help, I will definitely change the CPU cooler
 
Pretty much everything in your system seems to be overheating, what's the ambient temps in the room & I guess cooling in the rest of the case?

Your on about this one ?.

il.png



Case cannot breath, i would take the front of the case off and put 2? fans spaces at the front as intake and 1-2 fans at the top but would have to check the best direction as they might just pull the air from the front.

Might be better of buying a much better case.
 
Your case doesn't have proper airflow, buy a case with a mesh front, which results in good airflow and lower hardware temps.

Example:

Screenshot 2023-06-16 135101.png
 
Even then 100c is at least 10c too much, I had a similar case with 2700 & stock AMD cooler alongside 3080 & temps never went above 80 IIRC. That CPU shouldn't be touching 100c even in some of the worst cases out there.
 
Back
Top