[SOLVED] Disappointed at 3080 performance

[Gintama69]

Hmm now i'm bit disappointed at 3080's performance.. (well not that it really matters cos i don't game much)

According to Gamers Nexus review, 3080 FE is overall 20.9% faster than Asus 2080Ti Strix or 24.3% faster @ 1440p. (13.1%/24.3%/25.4% faster @ 1080p/1440p/4K).
But according to Bitwit review, 3080 FE is overall 16.2% faster than Gigabyte 2080Ti Aorus or 15.3% faster @ 1440p. (11.8%/15.3%/21.6% faster @ 1080p/1440p/4K).

Not sure why there's so much performance difference between the two reviewers especially @ 1440p(only res I care about), +24.3%(Gamers Nexus) vs +15.3%(Bitwit).
But if Bitwit is right then 3080's only gonna be ~10% faster than my 2080Ti.. i.e. unnoticeable except on bench.

Gamers Nexus =

Bitwit =

p.s. if you have 3080, can you please run Superposition bench at stock settings? thanks

 

[Gintama69]

Here's detailed performance numbers from both reviews..

https://imgur.com/a/VmTkPSY

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

 

[Math Geek]

everyone uses something different to benchmark with. different game, settings, hardware and so on will effect the numbers. in the end any single one is not worth taking as the absolute gospel but rather look at a bunch of different reviews and mentally average them. this is a good idea of what you can expect. but again your system will differ in many ways and will NEVER match what anyone else manages.

 

[Gintama69]

everyone uses something different to benchmark with. different game, settings, hardware and so on will effect the numbers. in the end any single one is not worth taking as the absolute gospel but rather look at a bunch of different reviews and mentally average them. this is a good idea of what you can expect. but again your system will differ in many ways and will NEVER match what anyone else manages.

But the difference between the two reviewers are huge, i.e. +24.3% @ 1440p(Gamers Nexus) vs +15.3% @ 1440p(Bitwit).. i.e. obviously they both can't be right.

 

[RTX 2080]

But the difference between the two reviewers are huge, i.e. +24.3% @ 1440p(Gamers Nexus) vs +15.3% @ 1440p(Bitwit).. i.e. obviously they both can't be right.

Actually, there should be every reason to believe that both of them are right.

Notice that Bitwit tested using a stock AMD Ryzen 3950x, whereas Gamers Nexus tested using an Intel i7-10700k overclocked to 5.1 GHz. The 3950x has more cores and threads, but the 10700k is definitely the superior gaming CPU.

So a 10% performance improvement variation between the two when one is using a far superior gaming CPU than the other (among many other testing variables) does not surprise me at all.

 

[Math Geek]

as i said, there are so many variables to how it performs, a different system of course is going to have different numbers. see if you can find something close to your actual system and that should help see where you might be. but again you WILL NEVER match what you see in any review. you may be higher or lower.

you can either enjoy the card with some fun gaming or you can spend all your time trying to figure ut why someone else got 2 extra fps than you. in my opinion one of these options is more fun than the other

 

[Some_Guy_87]

Even if it was 10%, a 700$ card outperforming a 1200$ card by that amount is still an incredible push in my opinion.

 

[sizzling]

It’s not really a like for like comparison. To me it seems likely that there will be a 3080Ti that if per prior releases will be launched in 9-10 month time. That’s the one to compare to if/when it releases. The 3080 should be compared to the 2080.

 

[protostar8]

Even if it was 10%, a 700$ card outperforming a 1200$ card by that amount is still an incredible push in my opinion.

I think you mean an $84,000 card outperforming a $1200 card according to Ebay hahahaha!

In all seriousness though, it is awesome that a card that's $500 cheaper can way outperform the $1200 one.

 

[hotaru.hino]

It feels silly to complain that a $700 card is disappointing when it gets ~20% more performance than a card that goes for $1200 on average.

 

[animekenji]

Gamers Nexus tries to remove CPU bottlenecking as a factor so you can see just how fast the GPU runs when it's going flat out. Steve will oftentimes also try to remove RAM clock speed as a factor by using the fastest memory available. Testing a GPU that is being bottlenecked by the CPU gives a useless result because you're not seeing what the GPU is actually capable of. If you bought the 3080 and are getting poor performance, your CPU is probably not up to the task of powering it. You can alleviate that situation somewhat by moving to a higher resolution, so the workload gets shifted from the CPU to the GPU.

 

[chickenballs]

you are comparing a $750 gpu to a $1400 (Aorus) gpu
you should look at cost per frame