Question Why is my PC so laggy? Its 2 months old as of June, getting ping spikes when I game, but my network is perfectly fine.

[karoi]

https://www.userbenchmark.com/Software
Theres the benchmark, has every spec on there.
Note: I didn’t update windows/get any new updates. I didn’t download anything except MalwareBytes to see if it was a virus. The network is fine on all my other devices, its just weird on my PC.

 

[CountMike]

https://www.userbenchmark.com/Software
Theres the benchmark, has every spec on there.
Note: I didn’t update windows/get any new updates. I didn’t download anything except MalwareBytes to see if it was a virus. The network is fine on all my other devices, its just weird on my PC.

That link doesn't provide your results.

 

[karoi]

That link doesn't provide your results.

My bad, here https://www.userbenchmark.com/UserRun/68382249

 

[Eximo]

Hmm, not a great showing. But userbenchmark isn't an amazing tool.

Load up CPU-Z and get us pictures of the memory tabs to make sure that dual channel is working.

Ryzen 5500 is a cut down 5600G if I recall correctly. Only has 8x PCIe 3.0 to the GPU, so the bandwidth of the RTX4060 is a little low, but it is a light card, so that shouldn't be much of an issue.

Now, motherboard and power details won't come through. So if you could provide that information that may help.

Since this is a prebuilt, it may just not have the latest BIOS installed which can make improvements to memory performance/compatibility/etc.

 

[rgd1101]

brand and model of the psu?

 

[karoi]

Hmm, not a great showing. But userbenchmark isn't an amazing tool.

Load up CPU-Z and get us pictures of the memory tabs to make sure that dual channel is working.

Ryzen 5500 is a cut down 5600G if I recall correctly. Only has 8x PCIe 3.0 to the GPU, so the bandwidth of the RTX4060 is a little low, but it is a light card, so that shouldn't be much of an issue.

Now, motherboard and power details won't come through. So if you could provide that information that may help.

Since this is a prebuilt, it may just not have the latest BIOS installed which can make improvements to memory performance/compatibility/etc.

Well before this reply, I had just opened my PC, and for some reason, there was no more ping spikes or anything, it was like it never happened. I am honestly so confused 😭

 

[karoi]

Hmm, not a great showing. But userbenchmark isn't an amazing tool.

Load up CPU-Z and get us pictures of the memory tabs to make sure that dual channel is working.

Ryzen 5500 is a cut down 5600G if I recall correctly. Only has 8x PCIe 3.0 to the GPU, so the bandwidth of the RTX4060 is a little low, but it is a light card, so that shouldn't be much of an issue.

Now, motherboard and power details won't come through. So if you could provide that information that may help.

Since this is a prebuilt, it may just not have the latest BIOS installed which can make improvements to memory performance/compatibility/etc.

Nevermind, its back to spiking. I’ll run CPU-Z and send pictures of the memory tabs. And do you want me to provide you with the motherboard model (is that what you mean) and what do you mean by power details, I’m not too educated in PC parts 😢

 

[Eximo]

Yes, the motherboard model. This is often where companies like CyberPower really cheap out.

The power supply is the thing with all the power cables coming out of it and going to your motherboard and GPU. It should have a label detailing its model and complete specs printed on it. It may be necessary to remove it to see the details.

I doubt it is underpowered for such a light GPU. This is the other major place that CyberPower tries to save a buck. If it is a particularly low quality unit, it could be having an effect on your motherboard or GPU when you game. Not enough to crash the system, but enough to make it slow down to prevent that.

That or simple overheating during a long gaming session. You mentioned opening up the computer helped for a while. A lot of chassis these days come covered in tempered glass rather than having air vents, this is not ideal for computers.

 

[karoi]

Hmm, not a great showing. But userbenchmark isn't an amazing tool.

Load up CPU-Z and get us pictures of the memory tabs to make sure that dual channel is working.

Ryzen 5500 is a cut down 5600G if I recall correctly. Only has 8x PCIe 3.0 to the GPU, so the bandwidth of the RTX4060 is a little low, but it is a light card, so that shouldn't be much of an issue.

Now, motherboard and power details won't come through. So if you could provide that information that may help.

Since this is a prebuilt, it may just not have the latest BIOS installed which can make improvements to memory performance/compatibility/etc.

Is this what you mean by the memory tab?

 

[Bob.B]

My bad, here https://www.userbenchmark.com/UserRun/68382249

Run ubm like this.

Reboot.
Wait a few mins.
Run ubm with the browser closed.
Post a link to the results page.

 

[karoi]

Yes, the motherboard model. This is often where companies like CyberPower really cheap out.

The power supply is the thing with all the power cables coming out of it and going to your motherboard and GPU. It should have a label detailing its model and complete specs printed on it. It may be necessary to remove it to see the details.

I doubt it is underpowered for such a light GPU. This is the other major place that CyberPower tries to save a buck. If it is a particularly low quality unit, it could be having an effect on your motherboard or GPU when you game. Not enough to crash the system, but enough to make it slow down to prevent that.

That or simple overheating during a long gaming session. You mentioned opening up the computer helped for a while. A lot of chassis these days come covered in tempered glass rather than having air vents, this is not ideal for computers.

Also here’s this. I think it has the motherboard model in it.

 

[karoi]

Yes, the motherboard model. This is often where companies like CyberPower really cheap out.

The power supply is the thing with all the power cables coming out of it and going to your motherboard and GPU. It should have a label detailing its model and complete specs printed on it. It may be necessary to remove it to see the details.

I doubt it is underpowered for such a light GPU. This is the other major place that CyberPower tries to save a buck. If it is a particularly low quality unit, it could be having an effect on your motherboard or GPU when you game. Not enough to crash the system, but enough to make it slow down to prevent that.

That or simple overheating during a long gaming session. You mentioned opening up the computer helped for a while. A lot of chassis these days come covered in tempered glass rather than having air vents, this is not ideal for computers.

Also, how can I track the temp of cpu and gpu? What should the safe temperatures be and what temps should I not go beyond.

 

[helper800]

If it is a particularly low quality unit, it could be having an effect on your motherboard or GPU when you game. Not enough to crash the system, but enough to make it slow down to prevent that.

That is nearly impossible. A PSU will either provide power or not. If it stops providing the required power then you get an unexpected shutdown. It is exceedingly rare for such a case to play out as described.

I would rule out overheating components long before suggesting that the PSU is the issue here. Even a bad quality PSU is going to work in the binary way I described above.

 

[helper800]

Also, how can I track the temp of cpu and gpu? What should the safe temperatures be and what temps should I not go beyond.

Below 95 degrees C would be safe, but for performance you would want below 80 C. HWinfo is very overwhelming but it will show you the hottest temps reached during use. Just runs sensors and test. Check the information regarding hotspot temperature on the graphics card and CPU. You can post a screenshot of the CPU temps or the graphics card temps if you are overwhelmed with all the info.

 

[karoi]

Below 95 degrees C would be safe, but for performance you would want below 80 C. HWinfo is very overwhelming but it will show you the hottest temps reached during use. Just runs sensors and test. Check the information regarding hotspot temperature on the graphic card and CPU. you can post a screenshot of the CPU temps or the graphics card temps if you are overwhelmed with all the info.

Alright, thanks. Do you know anything about why my PC is doing this? It just randomly started last week, and I can’t seem to find a consistent fix for it.

 

[helper800]

Alright, thanks. Do you know anything about why my PC is doing this? It just randomly started last week, and I can’t seem to find a consistent fix for it.

It could be any number of things, but it is easier to run down the simplest explanations first unless there is something highly indicative of a more specific issue. For instance, if we rule out temperatures we can check your internet connection for packet loss which can cause lag. Can you describe the stuttering more specifically? How is your PC performing outside of the stuttering? Fine?

 

[Eximo]

That is nearly impossible. A PSU will either provide power or not. If it stops providing the required power then you get an unexpected shutdown. It is exceedingly rare for such a case to play out as described.

I would rule out overheating components long before suggesting that the PSU is the issue here. Even a bad quality PSU is going to work in the binary way I described above.

I mostly agree. Decent power supplies that behave properly should shut themselves down under overcurrent and undervoltage conditions.

Cheap PSUs may be more lax, or lack the protections entirely. Excessive voltage droop would be the main possibility there. GPUs are quite good about dealing with incoming power these days and may opt to back off rather than cause a system crash.

 

[helper800]

I mostly agree. Decent power supplies that behave properly should shut themselves down under overcurrent and undervoltage conditions.

Cheap PSUs may be more lax, or lack the protections entirely. Excessive voltage droop would be the main possibility there. GPUs are quite good about dealing with incoming power these days and may opt to back off rather than cause a system crash.

I disagree. As soon as the GPU or other components ask for power that cannot be delivered you get an unexpected shutdown. A component cannot function as soon as the power delivery becomes mismatched from the power required to complete its task. There is a buffer though, and that would be capacitors on the motherboard and the PSU itself. If any of those capacitors are drained completely by components using them before the PSU can fill them back up, you get a shutdown. Correct me if I am wrong, but my conviction is very high on this.

 

[karoi]

It could be any number of things, but it is easier to run down the simplest explanations first unless there is something highly indicative of a more specific issue. For instance, if we rule out temperatures we can check your internet connection for packet loss which can cause lag. Can you describe the stuttering more specifically? How is your PC performing outside of the stuttering? Fine?

Yeah it runs smoothly outside of games. My ping is constant, usually 70, then the screen freezes, lagging and going up to 1000, with it dropping back down to 70 after 15~ seconds.

 

[Eximo]

I disagree. As soon as the GPU or other components ask for power that cannot be delivered you get an unexpected shutdown. A component cannot function as soon as the power delivery becomes mismatched from the power required to complete its task. There is a buffer though, and that would be capacitors on the motherboard and the PSU itself. If any of those capacitors are drained completely by components using them before the PSU can fill them back up, you get a shutdown. Correct me if I am wrong, but my conviction is very high on this.

Yes, if capacitors drained to zero. But that isn't going to happen as long as the PSU is on.

Voltage and current. As your voltage goes down, current goes up to maintain a fixed power draw. High current draw at the PCB end of things will get noticed by the logic in the VRMs, so they will reduce power to maintain a certain current limit, thus protecting internal components. The upside, the system is now using less power. The downside, less power available for work, clock speeds go down.

Current limits on the PSU end can be exceeded, and we have seen examples where they will go beyond the point of failure before tripping. A cost reduced cheapo PSU may forgo overcurrent and undervoltage protections entirely.

Really old group regulated supplies also had an issue with 5V to 12V imbalance, which could lead to high 12V output, as much as 15 volts. That would usually either kill a system quickly or lead to immediate shutdown on the logic end, but the PSU would keep going. (I should add this is why using those old PSUs with really high 3.3V and 5V output is bad, because modern systems don't draw enough 5V and 3.3V to get the analog circuit to behave properly)

 

[Eximo]

Yeah it runs smoothly outside of games. My ping is constant, usually 70, then the screen freezes, lagging and going up to 1000, with it dropping back down to 70 after 15~ seconds.

Fire up Performace Monitor and configure it to show your network connection usage. This is a bit tedious to set up, but it is a good look at what is going on network wise.

Click the green + to add a new graph. Network Adapter and then pick the device you use for networking (ethernet, wifi, etc)

Deselect all but Bytes/Sec. Every time you get lag, take a look and see if it kept data going or not.

(Maybe someone knows an easier way)

 

[helper800]

Yes, if capacitors drained to zero. But that isn't going to happen as long as the PSU is on.

Voltage and current. As your voltage goes down, current goes up to maintain a fixed power draw. High current draw at the PCB end of things will get noticed by the logic in the VRMs, so they will reduce power to maintain a certain current limit, thus protecting internal components. The upside, the system is now using less power. The downside, less power available for work, clock speeds go down.

Current limits on the PSU end can be exceeded, and we have seen examples where they will go beyond the point of failure before tripping. A cost reduced cheapo PSU may forgo overcurrent and undervoltage protections entirely.

Really old group regulated supplies also had an issue with 5V to 12V imbalance, which could lead to high 12V output, as much as 15 volts. That would usually either kill a system quickly or lead to immediate shutdown on the logic end, but the PSU would keep going. (I should add this is why using those old PSUs with really high 3.3V and 5V output is bad, because modern systems don't draw enough 5V and 3.3V to get the analog circuit to behave properly)

Everything discussed above is true with exception of the VRMs, but none of those examples would lead to anything other than a shutdown or the effect of which would immediately cause a shutdown. My main sticking point with your original comment was the below highlighted:

If it is a particularly low quality unit, it could be having an effect on your motherboard or GPU when you game. Not enough to crash the system, but enough to make it slow down to prevent that.

ACPI controls what components get what power. If ACPI assigns a certain amount of power and that power value is not met, you get a shutdown. VRMs request power based on settings from the BIOS, like LLC among others, and usually maintain this power to the VRM. Voltage droop may affect performance but that is a consequence of either the quality/performance level of the VRMs themselves or power settings in the BIOS, and NOT because they did not receive the power they requested in the first place.

If you wish to continue this conversation I think we should take it to DMs where I will come back and correct any mistakes from prior comments here if I learn it to be my mistake.

 

[helper800]

Fire up Performace Monitor and configure it to show your network connection usage. This is a bit tedious to set up, but it is a good look at what is going on network wise.

Click the green + to add a new graph. Network Adapter and then pick the device you use for networking (ethernet, wifi, etc)

Deselect all but Bytes/Sec. Every time you get lag, take a look and see if it kept data going or not.

(Maybe someone knows an easier way)

There is a utility program that is a bit intense called pingplotter that can show a chart history of a connection based on website or IP address. It shows speeds, packet loss, ping, and jitter at every jump between you and the target test.

 

[helper800]

Yeah it runs smoothly outside of games. My ping is constant, usually 70, then the screen freezes, lagging and going up to 1000, with it dropping back down to 70 after 15~ seconds.

When your screen freezes can you alt tab and still use the PC or does the entire PC freeze in this instance?

 

[Eximo]

If you wish to continue this conversation I think we should take it to DMs where I will come back and correct any mistakes from prior comments here if I learn it to be my mistake.

Agreed.

I am under the impression that ACPI is the software implementation of power control, not at the hardware algorithm level.

VRMs get their commands from software but some of its inputs are hardware signals. Been a while since I looked at VRM chip design white paper, so could be different from what I remember. Never really dived into the PC power delivery side of things too much.