Question Why is my system underperforming compared to the similar setups in Youtube?

[Crows_]

Hey everyone,

I have a strange question. I have

4080 Super Gainward Super PantherOC,
7800 X3D,
G.Skill 32GB(2x16) Flare X5 AMD EXPO 6000mhz CL30
However, I keep getting minus 40-80 fps compared to those on youtube benchmarks.

For Example;

View: https://www.youtube.com/watch?v=tHYOzORhdLc&t=1439s


He gets 280 fps, I get 190-210 fps.

Or this;

View: https://www.youtube.com/watch?v=koR0nf7poEY&t=357s


He has a worse CPU than me, yet yields more FPS.

What could be the issue here?

 

[COLGeek]

Your first issue is using YT videos to compare to your system. Many (most) of these videos are fakes and should not be used for comparison.

There are verifiable, legit means to benchmark and compare, but even then you can't directly compare two systems unless you can verify setup/configs.

 

[JohnBonhamsGhost]

you can't trust any "Youtube" performance videos.
you have to consider that these videos are made to attract viewers and that is it.

you'd be better of using benchmarking software that offers comparisons based on hardware,
but even then you have no idea how many of the top contenders have modified their systems to enrich their scores.

 

[Crows_]

Why did I fall for this. Of course they are... are there any trust worthy place that I can compare benchmarks?

 

[Colif]

I feel we need a sticky saying "Youtube videos may not show reality, don't use them for comparisons"

Systems made to test games might only exist to make video. Also overclocking etc.

 

[COLGeek]

Why did I fall for this. Of course they are... are there any trust worthy place that I can compare benchmarks?

From a very similar thread, just yesterday:

Test with this tool and then compare to the leaderboard. Pay attention to clock speeds.

UNIGINE Benchmarks

Performance benchmarks by Unigine

benchmark.unigine.com

 

[Crows_]

I feel we need a sticky saying "Youtube videos may not show reality, don't use them for comparisons"

Systems made to test games might only exist to make video. Also overclocking etc.

I think you should. I was discussing this with my friend just after I posted this thread. Youtube is basically a pool of attention. You are right.


I will check out that link!

 

[Colif]

I would if people actually looked at stickies... but time shows that most don't.

Forums probably wouldn't suggest when you post anyway. I doubt many people look at the suggestions it makes.

We might get the front page to write an article about it instead. More people might see that.

 

[helper800]

I think you should. I was discussing this with my friend just after I posted this thread. Youtube is basically a pool of attention. You are right.


I will check out that link!

I would also make sure that Resizable BAR is enabled in BIOS. Windows game mode option to off for a test. Make sure that you are on the latest Nvidia Driver. Make sure that you have the latest Chipset drivers from AMD for the Motherboard/CPU. Disable any/all background programs or services. I recommend doing a clean boot and then testing performance with UNIGINE to make sure that there are no 3rd party background services running.

 

[helper800]

I would if people actually looked at stickies... but time shows that most don't.

Forums probably wouldn't suggest when you post anyway. I doubt many people look at the suggestions it makes.

We might get the front page to write an article about it instead. More people might see that.

Personally, I link to our stickies all the time, so even if people ignore them and post anyways, its very easy to direct them to read the stickied post!

 

[Colif]

i know some read them but they hardly highlighted. Easy to miss them at top of every forum.

 

[helper800]

i know some read them but they hardly highlighted. Easy to miss them at top of every forum.

Do you think there should be a forced reading of certain stickies before you can post your first questions depending of the subforum you want to post in? For instance, before you can post your first question in the Systems section of the forums you are directed to read either the NO POST , Community Rules, or How to ask for new build / upgrade advice stickies?

 

[Colif]

We can't force people to do that. Its not a function of the software.

I know other forums may be able to.

We wandering off topic - this better for the forum feedback area

 

[Crows_]

Thanks for sharing that software. I can see that my GPU is performing as expected. I did the same with my CPU as well with Cinebench, but I think I cannot compare it with other people, is there are a similar software for CPU?

 

[Zerk2012]

Thanks for sharing that software. I can see that my GPU is performing as expected. I did the same with my CPU as well with Cinebench, but I think I cannot compare it with other people, is there are a similar software for CPU?

For multi core it should of been around 1115

https://www.cpu-monkey.com/en/benchmark-amd_ryzen_7_7800x3d-cinebench_2024_multi_core

 

[Crows_]

For multi core it should of been around 1115

https://www.cpu-monkey.com/en/benchmark-amd_ryzen_7_7800x3d-cinebench_2024_multi_core

it is 1076, is that a huge difference?

 

[helper800]

it is 1076, is that a huge difference?

No, its within margin. If you did a clean boot first, likely you would score higher because there would be no background services running for anything other than Windows.

 

[Zerk2012]

it is 1076, is that a huge difference?

About right it will very depending on the motherboards BIOS settings and memory.