[SOLVED] 3080 only 4% faster than 16Gbps 2080Ti?

[Gintama69]

I normally use Unigine Superposition 3D benchmark to compare the overall GPU performance different system. (@ 1080p medium, i.e. default settings)
Link: https://benchmark.unigine.com/superposition

I know that stock high-end 2080Ti cards such as Asus Strix or MSI Gaming X score ~26.5k on Superposition with gg gaming CPU(9900K or faster) @ stock settings, i.e. nothing overclocked.

I get ~28K on my PC(10900KF CPU, 32GB G.Skill 4000C15 RAM, 16Gbps MSI RTX2080Ti Gaming Z) with nothing overclocked.
I thought 10900K and 3080 would get ~31K on Superposition at stock settings, but my friend only got ~29K on his new PC(10900K, MSI RTX3080 Ventus 3X OC) which is only ~4% faster than my PC.

Can someone please post their Superposition score with 3080 GPU(with no overclock) so I can compare the result? Thanks

10900KF and 16Gbps MSI RTX2080Ti Gaming Z: 27934 -

https://imgur.com/AgHEEGJ

View: https://imgur.com/AgHEEGJ

10900K and MSI RTX3080 Ventus 3X OC: 29153 -

https://imgur.com/a/InwNBBx

View: https://imgur.com/a/InwNBBx

Difference: 1219 or 4.36%

 

[XxDarkMario20xX]

Everyone going to be getting other scores dose not matter if you have same CPU and GPU aslong as your games runs smoothly and fast thats all what matters

 

[geofelt]

3080 is designed for gaming @4k resolution.

at 1080p the gpu can loaf and the fps is determined mostly by the cpu.

 

[Gintama69]

I only play at 1440p resolution so 1080p or 4K performance is irrelevant for obvious reasons.

But considering 3080 is only ~4% faster than 16Gbps 2080Ti(which is one I have atm) @ 1080p, I don't think it's gonna be much of an upgrade even at 1440p.. maybe 9-12% performance increase won't be hardly noticeable on most games.

 

[hotaru.hino]

A handful of reviews of the 3080 that included 2080 Ti benchmarks show me that it gets anywhere from 10% to 30% better performance at 1440p, but hovering more towards the 25%-30% range. Either way, I don't think it's worth it to buy the next generation if you're already at the top tier.

 

[johnsoner13]

Why would anyone in their right mind upgrade from a 2080 ti to a 3080 to play at 1080p medium settings?

 

[Gintama69]

Why would anyone in their right mind upgrade from a 2080 ti to a 3080 to play at 1080p medium settings?

As I mentioned I play at 1440p res only, so obviously I'm only really interested in the performance difference at 1440p..
Having said that Superposition is very accurate gaming benchmark and if 3080 is only 4% faster than 2080Ti(16Gbps version) then it'll only be 9-12% faster @ 1440p, i.e. not very noticeable even @ 1440p.

A handful of reviews of the 3080 that included 2080 Ti benchmarks show me that it gets anywhere from 10% to 30% better performance at 1440p, but hovering more towards the 25%-30% range. Either way, I don't think it's worth it to buy the next generation if you're already at the top tier.

Well according to Gamers Nexus, 3080 FE is 22-26% faster than Asus 2080Ti Strix, or ~13%/24%/25% faster @ 1080p/1440p/4K with gg gaming CPU(10700K).
But 3080 FE is only 16% faster than Gigabyte 2080Ti Aorus if you have crap gaming CPU(3950X) according to BitWit.

https://imgur.com/5UElXjL

View: https://imgur.com/5UElXjL


So I think Gamers Nexus results are pretty accurate, however the review was done at the launch before 3080 stability issue became known(i.e. crash at high boost clock due to use of 6 cheap caps).
Since then NVidia has release new driver to prevent this issue(probably reducing the boost clock speed, and therefore reducing the overall performance).
I heard people got ~31K on Superposition with 10900K and 3080 at stock speed at the start, but now they seem to only get ~29K on Superposition with same setup.. and that's why I asked someone with 3080 to post the current Superposition score.

 

[Zerk2012]

I only play at 1440p resolution so 1080p or 4K performance is irrelevant for obvious reasons.

But considering 3080 is only ~4% faster than 16Gbps 2080Ti(which is one I have atm) @ 1080p, I don't think it's gonna be much of an upgrade even at 1440p.. maybe 9-12% performance increase won't be hardly noticeable on most games.

That is just using one game engine they all respond different, it's like running just one game to get results and saying that is all you will get in all games.

 

[hotaru.hino]

Well, at the end of the day I don't really know what you're trying to get at. If you're trying to claim the 3080 is "only" 4% faster than the 2080 Ti and you're using a single benchmark with settings weighed more towards better CPU performance, then it's not a compelling argument.

 

[Gintama69]

That is just using one game engine they all respond different, it's like running just one game to get results and saying that is all you will get in all games.

Well I've used Superposition bench for many different cards, 1060, 1080, 1080Ti, 2060, 2070, 2080S, 2080Ti's, etc and the difference in score were very noticeable between all the cards. (All @ default 1080p medium settings)
So it's very weird to see only 4% difference between the 2080Ti(16Gbps) and 3080 scores, when the difference should be at least ~10-15%. (even considering 1080p resolution and faster 16Gbps model)

If stock 10900K and 3080 scored ~31K like some claimed at the start then that sounds about right(cos ~11% difference), so I'm maybe guessing NVidia has gimped the 3080 performance with new drivers. (due to known crash issue at high speed with cheap caps)

 

[hotaru.hino]

Since you seem to be wanting to use Superposition for some reason as the basis of your argument, here's two scores from their leaderboard for the 1080p Medium preset

• https://benchmark.unigine.com/results/rid_038a021cdc7a42649fbe4127eb43566d
• https://benchmark.unigine.com/results/rid_117b5fa3dd5142faa4691d2c4dca38a7

Both used the same drivers, which are the latest versions (456.71). One managed to get 30K with a i9-9900K and an EVGA card. The other got 28.4K with an i7-8700K and an MSI card.

But again, you're also using a preset that's weighted more towards CPU performance. It's like claiming the RTX 2080 Ti can't convincingly outperform (if at all) older video cards like the GTX 1080 or GTX 980 because it doesn't score high enough in 3D Mark's Ice Storm test.