NVIDIA GeForce RTX 3080 Performance Leaked: 24 Percent Faster Than RTX 2080 Ti

Tsing

The FPS Review
Staff member
Joined
May 6, 2019
Messages
11,202
Points
83
nvidia-geforce-rtx-3080-close-up-1024x576.jpg
Image: NVIDIA



The embargo for NVIDIA GeForce RTX 3080 reviews won’t lift until next week, but VideoCardz has decided to spoil us early with a series of alleged performance metrics for green team’s flagship GPU, which releases in about seven days.



You can check out the complete numbers (i.e., 3DMark benchmarks and game-specific tests) below, but the GeForce RTX 3080 appears to be 24 percent faster on average than NVIDIA’s previous flagship, the GeForce RTX 2080 Ti. It also beats out the GeForce RTX 2080 SUPER, GeForce RTX 2070 SUPER, and GeForce RTX 2060 SUPER by 37, 45, and 50 percent, respectively.



More specifically, the GeForce RTX 3080 seems to have no problem blowing past 60 FPS in 4K, even without the help of NVIDIA’s deep learning super sampling...
Click to expand...

Continue reading...
 
Last edited by a moderator:
If true, and it is 25% faster than the 2080Ti, then they will have a hot seller on their hands. Even if gimped at 10GB of vRAM, it'll be a nice upgrade for anyone coming from the Pascal era..
 
At 4K 22-25% faster could be as much as 13-15FPS (assuming 60FPS baseline). Percentage gains always look nice, but I am reminded of the fact that 25% of 10 is only a 2.5 gain. Now if we're talking FPS, a 2.5 FPS gain isn't going to be very noticeable..
 
I don't know what "MaxQ" means, but I'd like to see the difference in SotTR with ray tracing.
 
I guess the 3080 will be my next card. Seems like a big enough improvement over my 1080Ti.
 
Just realized my 2070 Super may fall within the EVGA step up program guidelines so hopefully I can do that as I would just move the 2080 Ti to my back up PC and possibly get a 3080 or 3090.
 
Armenius said:
I don't know what "MaxQ" means, but I'd like to see the difference in SotTR with ray tracing.
Click to expand...

I thought MaxQ was the branding used on the laptop GPUs - basically same thing just with different Boost / temp parameters so you don’t melt down the laptop
 
And here I thought they were supposed to be 50-60% faster. Still 25% is nice... especially since it's launch price is less expensive.
 
Thats great considering the designated price point. In come the miners and the scalpers...
 
Well well well....
www.mooreslawisdead.com

Nvidia’s Ultimate Play

Introduction – Let Them Eat Cake 11/23/20 Update: It has now come to light that Nvidia has been directly selling large quantities of Ampere ASICs to mining firms since Ampere's launch. These firms write their own bios', drivers, and order their own coolers. This explains the conflicting...
www.mooreslawisdead.com www.mooreslawisdead.com

Too good to be true this nicer new Nvidia?
 
Uvilla said:
Well well well....
www.mooreslawisdead.com

Nvidia’s Ultimate Play

Introduction – Let Them Eat Cake 11/23/20 Update: It has now come to light that Nvidia has been directly selling large quantities of Ampere ASICs to mining firms since Ampere's launch. These firms write their own bios', drivers, and order their own coolers. This explains the conflicting...
www.mooreslawisdead.com www.mooreslawisdead.com

Too good to be true this nicer new Nvidia?
Click to expand...

Good read there that makes sense (hopefully not).
 
Nice read and makes sense. But it's going to bite them in the *** if AMD can release to market with volume and compete with the 3080 it will hurt them a lot. Especially if they are competitive vs the vaunted cooler.
 
I do agree it is a dangerous game to play. Very dangerous. I know few things piss me off more than being screwed over in price. So we will see.
 
Your math is way off. The 3080 is about 100% faster than the 2060S @4k max, not only 50%. The graphs read as a percentage comparison of the current gen cards vs the 3080 with the 3080 as "100%". So you can't just take the subtractive difference between the numbers, you have to divide 100 by whatever the lower number is. So for the 2060S, it scores about 50% of the 3080's performance. 100/50 = 2, or double, which is 100% more.
 
I don't think the performance of the card alone can hit up 50% more than 2080ti. It's the DLSS 2.0, AI and stuffs that help the supported game to reach 50% faster, and we don't even have it tested yet. The 25% improvement along with a budget price is also huge. Why don't you stop bitching about it, and wait for an actual test on the game that really support the card's potential (cyberpunk 2077, minecraft, etc.)?
 
I'm just tired of all the BS, I want real benchmarks where everything is known and controlled.
I don't know why nvidia is doing this, and even putting gasoline on the fire.

If you say 30% faster and people get 25% in some games, 30% if others, they will be fine.
But if you preach 50% and they actually only get 30-40% then they'll be disappointed regardless of the fact that the card is already great value (assuming it won't be a paper launch)
 
You must log in or register to reply here.
Become a Patron!
Back
Top