Question Computer stuttering while gaming

[Aditya Kumar]

Whenever I play Apex Legends (it's the only game I play) it makes the whole PC stutter every 3-4 mins (for 2-3 secs everything freezes). Although stuttering also happens when I am not gaming and just browsing but it's 2-3 times per day. This issue only started last month. Also I got my GPU replaced last month, so it can't be GPU. I checked temperatures during gaming and it's all under 70 Celsius on GPU, CPU & Mobo. PC is little less then 2 years old but PSU is 12 years old.

CPU : AMD Ryzen 5 3600x
Motherboard : Asus TUF Gaming X570 Plus
RAM : 2 X 8GB A-Data XPG 3200Mhz CL16
SSD : Silicon Power 1TB NVMe M.2 Gen 3
GPU : Zotac GTX 1660 Twin Fan
PSU : Corsair 550W (Bought in 2009)

Where should I start ? Coz it could be motherboard or CPU or PSU or RAM or GPU, it's hard to figure out as I don't have 2nd desktop to switch parts & find the faulty hardware.
I think it's most likely PSU but I just don't want to buy new PSU for $100 then later realize it was something else which was faulty.

 

[bigboxes]

I would buy a new power supply and be done with it. Think of the CPU as the brain of your computer. The power supply is the heart. $100 is not much for a stable computer.

Have you tested your RAM? Also, what are your temps for the SSD? When the controller gets hot it throttles. All sorts of things for you to check. I'd get a new power supply regardless.

 

[VirtualLarry]

I've never really heard that PSU problems cause "stutters". Generally, PSU problems, result in blue-screens or power-offs / crashes. Sometimes, though, slightly off CPU calculations can result from PSU issues as well, but that usually results in appcrashes. TBH, that's usually what bad RAM causes too.

I'm going to go out on a limb here, and say that you should consider doing a BIOS update, and possibly a BIOS reset to defaults, both before and after the BIOS flash.

Edit: I would also suspect storage as the causes of the pauses or stutters. Adata isn't know for the highest-quality drives, if you know what I mean. I've personally had some issues with Adata NVMe SSDs, a few years ago. I don't know what they are like these days, but LTT did a YT video where he identified FIVE different hardware configurations used for the SAME product model code. If that tells you anything.

 

[Steltek]

Also, have you tried running the game as administrator? Right click on the shortcut you use to start it up, hit Properties, and under Compatibility check the box to run as administrator. Then, launch the program.

EDIT:
I initially missed the fact that it is also stuttering on occasion when browsing. That could indicate a video driver issue as most modern web browsers use hardware acceleration. Were it me, I'd try to downgrade the video driver by installing a slightly older video driver version to see if it is a driver bug. Are you running the older Zotac OEM driver, or the newest Nvidia driver?

Larry's suggestion about the BIOS upgrade/reset is also a good one.

I presume you've checked for firmware updates for your NVMe drive to make sure it is running the most recent one?

 

[KarlRos]

1. Try updating your Graphics Driver.
2. Try running the game at lower graphics settings
3. If nothing works, upgrade your Graphics Card to run heavy titles.

 

[Aditya Kumar]

I would buy a new power supply and be done with it. Think of the CPU as the brain of your computer. The power supply is the heart. $100 is not much for a stable computer.

Have you tested your RAM? Also, what are your temps for the SSD? When the controller gets hot it throttles. All sorts of things for you to check. I'd get a new power supply regardless.

Memtest on RAM passed with 0 error. SSD temps max 45 Celsius. Every part in my pcc s under warranty except PSU.

I've never really heard that PSU problems cause "stutters". Generally, PSU problems, result in blue-screens or power-offs / crashes. Sometimes, though, slightly off CPU calculations can result from PSU issues as well, but that usually results in appcrashes. TBH, that's usually what bad RAM causes too.

I'm going to go out on a limb here, and say that you should consider doing a BIOS update, and possibly a BIOS reset to defaults, both before and after the BIOS flash.

Edit: I would also suspect storage as the causes of the pauses or stutters. Adata isn't know for the highest-quality drives, if you know what I mean. I've personally had some issues with Adata NVMe SSDs, a few years ago. I don't know what they are like these days, but LTT did a YT video where he identified FIVE different hardware configurations used for the SAME product model code. If that tells you anything.

I have silicon power SSD not A-data. RAM passed memtest. I did BIOS thing u said also. Should I RMA the SSD ?

Also, have you tried running the game as administrator? Right click on the shortcut you use to start it up, hit Properties, and under Compatibility check the box to run as administrator. Then, launch the program.

EDIT:
I initially missed the fact that it is also stuttering on occasion when browsing. That could indicate a video driver issue as most modern web browsers use hardware acceleration. Were it me, I'd try to downgrade the video driver by installing a slightly older video driver version to see if it is a driver bug. Are you running the older Zotac OEM driver, or the newest Nvidia driver?

Larry's suggestion about the BIOS upgrade/reset is also a good one.

I presume you've checked for firmware updates for your NVMe drive to make sure it is running the most recent one?

Newest Nvidia Driver from there website. Can't find drivers for silicon power nvme. There is one software for its for Sata SSD I think.

 

[aigomorla]

sounds like possibly thermal throttling how its repeatable in gaming and when not, bascially it seems like it freezes when s/he does something.

3600x was a wicked hot cpu and the wraith spire cooler is better off recycled for new beer cans.

Oh the nVME can also throttle too.

 

[bigboxes]

Memtest on RAM passed with 0 error. SSD temps max 45 Celsius. Every part in my pcc s under warranty except PSU.




I have silicon power SSD not A-data. RAM passed memtest. I did BIOS thing u said also. Should I RMA the SSD ?





Newest Nvidia Driver from there website. Can't find drivers for silicon power nvme. There is one software for its for Sata SSD I think.

When I first built my current build back in April, I experienced exactly what you were describing. I first turned off all power management and set everything to never turn off after I set it for performance. Then, I optimized all the drives. Windows 10 does a lot better job at that than Win 7. But I still went through the list and did check everything. I also updated my SSD's (Samsung 980 Pro) firmware through the Samsung Magician software. My Sabrent NVMe drive didn't have any options. Instead, it just uses the Windows driver. So, I'm hoping the optimization helped because the stuttering went away.

Also, you said it stutters even when not gaming. Just more often when gaming. Try an older version of the Nvidia driver. Still having problems? Could still be your video card. I would just use HWiNFO64 and see what the temps on your different parts when this event happens. High temps will make it throttle. Video card can throttle. CPU can throttle. NVMe can throttle. I installed a Thermalright heatsink to the Sabrent and it dropped temps ten degrees. TEN degrees. It's usually 10-15 degrees cooler than the 980 Pro that lies beneath the stock heatsink (for my motherboard).

 

[DAPUNISHER]

I think it's most likely PSU but I just don't want to buy new PSU for $100 then later realize it was something else which was faulty.

Pro tip: That PSU is going to fail. It's not a matter of if, but when, with all components. But a '09 PSU is going to be much sooner than the rest of the system, in all probability. And it'll potentially take out other components when it does. If it turns out the issue is another component, it will still be among the wisest $100 expenditures you could make.

Use DDU in safe mode, nuke the drivers, install latest. Still no joy? Start looking at what software is running. And do full malware scans.

Also, what GPU did you have before?

 

[blckgrffn]

If you are in the US, there is a EVGA Supernova 650W for ~$55 on Amazon right now. I mean, it's worth a shot and worst case you have a new PSU that is supposed to be pretty high quality.

 

[Aditya Kumar]

sounds like possibly thermal throttling how its repeatable in gaming and when not, bascially it seems like it freezes when s/he does something.

3600x was a wicked hot cpu and the wraith spire cooler is better off recycled for new beer cans.

Oh the nVME can also throttle too.

All temps are super good. I checked every temp sensor reading max temps.

When I first built my current build back in April, I experienced exactly what you were describing. I first turned off all power management and set everything to never turn off after I set it for performance. Then, I optimized all the drives. Windows 10 does a lot better job at that than Win 7. But I still went through the list and did check everything. I also updated my SSD's (Samsung 980 Pro) firmware through the Samsung Magician software. My Sabrent NVMe drive didn't have any options. Instead, it just uses the Windows driver. So, I'm hoping the optimization helped because the stuttering went away.

Also, you said it stutters even when not gaming. Just more often when gaming. Try an older version of the Nvidia driver. Still having problems? Could still be your video card. I would just use HWiNFO64 and see what the temps on your different parts when this event happens. High temps will make it throttle. Video card can throttle. CPU can throttle. NVMe can throttle. I installed a Thermalright heatsink to the Sabrent and it dropped temps ten degrees. TEN degrees. It's usually 10-15 degrees cooler than the 980 Pro that lies beneath the stock heatsink (for my motherboard).

Voltages are all good on all rails . GPU goes to 100% usage which is normal , CPU don't go more then 60-70% and SSD around 70% max read speed & 11% write speed max . I reinstalled windows too and used older nvidia drivers but same issue.. Attaching graphs and logs please check. Frame time spikes up every now and then for no reason from 8-10ms to 70ms or something.

Pro tip: That PSU is going to fail. It's not a matter of if, but when, with all components. But a '09 PSU is going to be much sooner than the rest of the system, in all probability. And it'll potentially take out other components when it does. If it turns out the issue is another component, it will still be among the wisest $100 expenditures you could make.

Use DDU in safe mode, nuke the drivers, install latest. Still no joy? Start looking at what software is running. And do full malware scans.

Also, what GPU did you have before?

Deleted all partitions on SSD, reinstalled full windows with older NVidia drivers but no joy. I had GTX 1660 before since I build this PC 2 years ago, just 1 month ago I got it RMA'd as one of the fan was dead. But this problem started 1 week ago, so the RMA'd GPU is less likely to be faulty.

I think I will buy PSU and if problem is still there then RMA's all PC parts 1 by 1 as all are under warranty except 2009 PSU. First I will change motherboard, then CPU, then RAM, then SSD. lol and if still the issue doesn't go away I will jump from the roof lol.

If you are in the US, there is a EVGA Supernova 650W for ~$55 on Amazon right now. I mean, it's worth a shot and worst case you have a new PSU that is supposed to be pretty high quality.

Not in USA, I will do some research of what's available here.

LOGS : https://drive.google.com/drive/folders/1JOle3JYrQT7QnFACfr6UjDfz7v0Cm4_c

 

[blckgrffn]

I support the replace one thing at a time until there is no longer a problem strategy when the issue is extremely non-obvious.

Good luck!

 

[aigomorla]

so the RMA'd GPU is less likely to be faulty

uhhh... not unless that RMA came sealed and new.
Ive seen several vendors refurbished QC pass stuff which should of been thrown in the incinerator without even thinking twice.

Sometimes vendors will send you junk, just so they can get another 2-4 weeks trying to source the replacement.
Especially with the GPU shortages, and how many RMA's they must of gotten from mined cards.

Not in USA, I will do some research of what's available here.

what country? Maybe we can help you.

 

[Aditya Kumar]

uhhh... not unless that RMA came sealed and new.
Ive seen several vendors refurbished QC pass stuff which should of been thrown in the incinerator without even thinking twice.

Sometimes vendors will send you junk, just so they can get another 2-4 weeks trying to source the replacement.
Especially with the GPU shortages, and how many RMA's they must of gotten from mined cards.


what country? Maybe we can help you.

RMA'd GPU was not sealed but was brand new. I live in India. If after buying new PSU it still doesn't work should I RMA all PC part in 1 go? instead of doing 1 by 1 and wasting months to figure out what was causing the issue.

 

[Aditya Kumar]

One thing also I forgot to mention there was BSOD issue which fixed by itself maybe it is connected to this issue, you can check here :

Question - BSOD whenever I start gaming

Whenever I launch any game and play for 2-3 minutes it crashes the whole PC. This issue only started last month. Also I got my GPU replaced last week, so it can't be GPU. I checked thermals during gaming and it's normal on GPU, CPU & Mobo. PC is little less then 2 years old. CPU : AMD Ryzen 5...

forums.anandtech.com

 

[solidsnake1298]

So I looked at your hardware monitoring log. You aren't hitting your GPU power cap, temps are fine, GPU memory usage is within the 6GB VRAM, it doesn't appear you are OC'd. But your GPU usage is all over the place. In a GPU limited situation, GPU usage should be 98%+ all the time. In a CPU limited situation the GPU usage will be all over the place, like in your situation (see chart I made from your hardware log).

One thing I don't see in any of your posts is what version of Windows you are running. Other uses have spoken about Windows 7 and 10, but you never clarified which you are running. I checked your other threads and you also don't mention what version of Windows you are running. I think it is important that we know because Microsoft has made significant improvements to their CPU scheduler to account for Ryzens CCX and CCD designs in later versions of Windows 10. In your case, the 3600X has one CCD with two 3 core CCX's. I used to use this program called Process Lasso to automatically keep games on the same CCX to improve frame time consistency, which is the issue you are having. But in later revisions of Windows 10, starting with 1903 I think, Microsoft made changes to the scheduler to better keep related processes on the same CCX and CCD, when possible. After 1903 I no longer needed to use Process Lasso. If you are running Windows 7 or an early revision of Windows 10, that may be the cause of your problems.


(Blue = GPU usage, Orange = Overal CPU usage, Gray = Frame Time)

 

[Aditya Kumar]

So I looked at your hardware monitoring log. You aren't hitting your GPU power cap, temps are fine, GPU memory usage is within the 6GB VRAM, it doesn't appear you are OC'd. But your GPU usage is all over the place. In a GPU limited situation, GPU usage should be 98%+ all the time. In a CPU limited situation the GPU usage will be all over the place, like in your situation (see chart I made from your hardware log).

One thing I don't see in any of your posts is what version of Windows you are running. Other uses have spoken about Windows 7 and 10, but you never clarified which you are running. I checked your other threads and you also don't mention what version of Windows you are running. I think it is important that we know because Microsoft has made significant improvements to their CPU scheduler to account for Ryzens CCX and CCD designs in later versions of Windows 10. In your case, the 3600X has one CCD with two 3 core CCX's. I used to use this program called Process Lasso to automatically keep games on the same CCX to improve frame time consistency, which is the issue you are having. But in later revisions of Windows 10, starting with 1903 I think, Microsoft made changes to the scheduler to better keep related processes on the same CCX and CCD, when possible. After 1903 I no longer needed to use Process Lasso. If you are running Windows 7 or an early revision of Windows 10, that may be the cause of your problems.


(Blue = GPU usage, Orange = Overal CPU usage, Gray = Frame Time)
View attachment 47444

I am on latest build. Frame time is all over the place. Attaching hwinfo log while gaming.

hwinfo log.CSV

drive.google.com

 

[solidsnake1298]

Your initial post said you have 2x8GB of RAM, but your HWInfo log says you only have 8GB with over 80% used. If your memory is running in single channel that is 100% the cause if your frame time issues. Check your RAM and make sure they are inserted all the way.

 

[Aditya Kumar]

Your initial post said you have 2x8GB of RAM, but your HWInfo log says you only have 8GB with over 80% used. If your memory is running in single channel that is 100% the cause if your frame time issues. Check your RAM and make sure they are inserted all the way.

View attachment 47448

I am just testing with 1 RAM at a time, to rule out bad ram Issue. I just took 1 RAM out for few mins and trying gaming. Then I will switch to other RAM and try if stutter stops.

If you want I can put both ram in and give u new logs.

 

[solidsnake1298]

The log you posted yesterday, the one I made a chart of above, was that with 1 stick or 2? If that log was run with 1 stick, can you post a similar log with 2 sticks?

 

[Aditya Kumar]

The log you posted yesterday, the one I made a chart of above, was that with 1 stick or 2? If that log was run with 1 stick, can you post a similar log with 2 sticks?

That log I posted yesterday was from msi it was with 2 sticks.

I will again make new logs from hwinfo & msi afterburner today with 2 sticks.

 

[solidsnake1298]

For anyone curious, I charted some key values in the second hwinfo log that OP posted. The chart is on the second sheet and must be in Google Sheets, proper, and not the preview.

hwinfo log.xlsx

drive.google.com

 

[Aditya Kumar]

For anyone curious, I charted some key values in the second hwinfo log that OP posted. The chart is on the second sheet and must be in Google Sheets, proper, and not the preview.

hwinfo log.xlsx

drive.google.com

Thanks for the help

Updated logs today with 16GB RAM please check :

Stutter Problem 2 – Google Drive

drive.google.com

Will you be able to make graph for it as well?




Here the screenshot of hwinfo after 1.5 hours of gaming (Right click and open it new tab to zoom in) :

 

[solidsnake1298]

Graphed. Still doesn't look great, even with dual channel. Your power usage is well under the max of your PSU. If there was a problem with your PSU I could see it causing stability problems. But stuttering? I'm not sure I buy that.

Temps are fine, VRAM usage is fine, CPU usage is fine, main RAM usage is fine (but a little high). While GPU usage isn't consistently high, the spikes in frame time don't correspond to dips in GPU or spikes in CPU usage. Both for this set of data and the last set I graphed. There is no 1-to-1 correlation to GPU/CPU usage and frame time spikes.

What do you have going on in the background in terms of programs? I know for a fact that gaming alone won't use 70-75% of your RAM if that is the only thing you are doing. Especially on a GTX1660 you probably aren't running on more than high settings, 1080P. I have HWInfo64 constantly running on my second monitor and know that while gaming I am using ~55% of my 16GB of RAM.

So, to me, that sounds like something in the background is interrupting your game process.

hwinfo log 2

Usage

docs.google.com

 

[Aditya Kumar]

Graphed. Still doesn't look great, even with dual channel. Your power usage is well under the max of your PSU. If there was a problem with your PSU I could see it causing stability problems. But stuttering? I'm not sure I buy that.

Temps are fine, VRAM usage is fine, CPU usage is fine, main RAM usage is fine (but a little high). While GPU usage isn't consistently high, the spikes in frame time don't correspond to dips in GPU or spikes in CPU usage. Both for this set of data and the last set I graphed. There is no 1-to-1 correlation to GPU/CPU usage and frame time spikes.

What do you have going on in the background in terms of programs? I know for a fact that gaming alone won't use 70-75% of your RAM if that is the only thing you are doing. Especially on a GTX1660 you probably aren't running on more than high settings, 1080P. I have HWInfo64 constantly running on my second monitor and know that while gaming I am using ~55% of my 16GB of RAM.

So, to me, that sounds like something in the background is interrupting your game process.

hwinfo log 2

Usage

docs.google.com

RAM never went above 60% i just checked HardwareMonitoring.hml

Sometimes I open chrome when I am waiting for in match queue, my chrome has 50 tabs open. But I close chrome when match starts. No memory hogging task running in background when I am in game. I reinstalled Win10 yesterday, got rid of all bloatwares/auto-updaters.

I think these issues could be related :

One thing also I forgot to mention there was BSOD issue which fixed by itself maybe it is connected to this issue, you can check here :

Question - BSOD whenever I start gaming

Whenever I launch any game and play for 2-3 minutes it crashes the whole PC. This issue only started last month. Also I got my GPU replaced last week, so it can't be GPU. I checked thermals during gaming and it's normal on GPU, CPU & Mobo. PC is little less then 2 years old. CPU : AMD Ryzen 5...

forums.anandtech.com