Question Why is my GPU underperforming

[shadi.obaia]

I recently purchased a 2060 super and for some reason, I can't run games as well as I thought I would be. For example, GTA 5 runs at 35-45 FPS on normal settings when I thought it would be able to run high-ultra at 60. I have no idea why my frames are bad.

UserBenchmarks: Game 62%, Desk 42%, Work 42%
CPU: AMD Ryzen 5 2400G - 49.3%
GPU: Nvidia RTX 2060S (Super) - 84.4%
SSD: Adata XPG SX6000 NVMe PCIe M.2 128GB - 87%
HDD: Seagate Barracuda Green 2TB - 64.6%
RAM: Adata DDR4 3600 2x8GB - 80%
MBD: MSI B450-A PRO (MS-7B86)

edit: after doing the benchmark, I can see that my CPU is heavily underperforming. No idea why.
https://www.userbenchmark.com/UserRun/25050308

 

[shadi.obaia]

Open your case and take a look. Since you are asking for help, we can not guess your components, some of the work has to be done by you!

Sorry for the wait. As mentioned before, my knowledge of computers is limited so it took me a while to find out the PSU. The case I have hides the labels of the power supply so it took me a bit to find out how to take it out. The side of the power supply says "EVGA 500w"; here's a link to what it looks like https://www.evga.com/products/product.aspx?pn=100-W1-0500-KR. Thank you for your patience.

 

[Oussebon]

Just to say, instead of userbenchmark which is at best dubious (frankly it's misleading garbage), use the free Timespy benchmark to assess CPU and GPU performance.

Timespy is part of 3dmark and you can get it free on steam
3Dmark: https://store.steampowered.com/app/223850/3DMark/
Timespy addon: https://store.steampowered.com/app/496100/3DMark_Time_Spy_benchmark/

The results are far more reliable, and easier to compare to more exact measurements of what you should be getting.

 

[shadi.obaia]

Just to say, instead of userbenchmark which is at best dubious (frankly it's misleading garbage), use the free Timespy benchmark to assess CPU and GPU performance.

Timespy is part of 3dmark and you can get it free on steam
3Dmark: https://store.steampowered.com/app/223850/3DMark/
Timespy addon: https://store.steampowered.com/app/496100/3DMark_Time_Spy_benchmark/

The results are far more reliable, and easier to compare to more exact measurements of what you should be getting.

It says it costs money. I would if it was free though.

 

[Oussebon]

It's free - use the Download Demo button for the free version, which is all you need.

The paid-for version lets you customise tests, skip the intro video etc, but for reference against other users you only need the free version which is labelled as a demo

 

[kerberos_20]

install amd chipset drivers https://www.amd.com/en/support
then go to power plans and select ryzen balanced plan
https://www.tenforums.com/tutorials/2843-change-power-plan-settings-windows-10-a.html
run userbechmark again

 

[shadi.obaia]

install amd chipset drivers https://www.amd.com/en/support
then go to power plans and select ryzen balanced plan
https://www.tenforums.com/tutorials/2843-change-power-plan-settings-windows-10-a.html
run userbechmark again

The AMD software says it can't detect any hardware.

 

[kerberos_20]

The AMD software says it can't detect any hardware.

chipset -> amd socket am4 ->b450 -> submit
then click on windows 10 64bit
AMD Chipset Drivers revision number 2.01.15.2138 u click download

 

[shadi.obaia]

chipset -> amd socket am4 ->b450 -> submit
then click on windows 10 64bit
AMD Chipset Drivers revision number 2.01.15.2138 u click download

It seems to be running a bit better than it was before. Still low numbers for the CPU but it did go up. thanks! https://www.userbenchmark.com/UserRun/25059809

 

[kerberos_20]

It seems to be running a bit better than it was before. Still low numbers for the CPU but it did go up. thanks! https://www.userbenchmark.com/UserRun/25059809

background apps reduces performance, 16% cpu usage while benchmarking

anyway try gaming

 

[shadi.obaia]

background apps reduces performance, 16% cpu usage while benchmarking

anyway try gaming

Gta 5 Very high @ 20fps

 

[Oussebon]

Seriously though, stop using UBM to measure performance as it really doesn't tell you anything reliable Also gaming performance can vary a lot depending on what you do in the game, and not everyone has X game to compare performance to yours.

Try Timespy as that will give a much, much clearer indication.

 

[Phaaze88]

Seriously though, stop using UBM to measure performance as it really doesn't tell you anything reliable

I agree that it shouldn't be used to measure performance, but I disagree that it doesn't tell users anything reliable.
It does have it's uses in helping to troubleshoot hardware and software issues, but people need to ignore the ratings - 'below expectations, above', etc.

Timespy isn't quite as clear as you say it is, compared to UBM, at least.
-it excludes how much space is being used on storage, nor does it reflect drive performance; it just gives a name and the max capacity
-it doesn't flag for high background usage
-it doesn't show how much memory is being reserved
-it only runs 1080p. There can be some rather conflicting combinations of hardware out there, and knowing if they're actually running outside of 1080p helps

 

[Oussebon]

What Timespy does do is give clear and accurate CPU and GPU performance measurements, that can be compared to what the hardware ought to be providing, unlike UBM.

That needs to be established first before any of the other issues / before working out what then is causing any performance drop.

Which don't need to be - and shouldn't be - diagnosed via UBM.

There isn't a role for UBM.

 

[shadi.obaia]

What Timespy does do is give clear and accurate CPU and GPU performance measurements, that can be compared to what the hardware ought to be providing, unlike UBM.

That needs to be established first before any of the other issues / before working out what then is causing any performance drop.

Which don't need to be - and shouldn't be - diagnosed via UBM.

There isn't a role for UBM.

Here is what I got from the 3Dmark demo
https://www.3dmark.com/3dm/44337570?