RTX 2080 TI vs GTX 1080 TI Benchmark

[Creig]

I certainly hope AMD can ship a Halo product soon so we get some more sanity in these prices.

Or people can just stop throwing money at Nvidia at every new launch. No sales = price drop.

 

[Snowdensjacket]

I game at 1440 21:9 but I'm still using an rx480 which is totally not good enough for this new monitor. Waiting for reviews and for the 2000 series prices to drop to the non-FE prices. I'll look at picking up a 2080 during black Friday time. Wife is about to have another baby, induction scheduled in exactly one week, so I have no rush to get a card immediately.

From what I've read about DLSS I'm really looking forward to seeing what it can do. Ray tracing? Ehhh I don't think it's going to be ready for awhile.

 

[madcap magician]

This is supply and demand in practice. With no competition from AMD, prices will most likely stay at the current levels through Christmas.

 

[alxlwson]

Can we do a who actually games at 4K, 2K, HD Pole on here? I am honestly wondering how many of our readers truly use a 4K desktop monitor.
PS: I game at 2K+ 21:9 1440.

Done!

 

[Zarathustra[H]]

Can we do a who actually games at 4K, 2K, HD Pole on here? I am honestly wondering how many of our readers truly use a 4K desktop monitor.
PS: I game at 2K+ 21:9 1440.

Forum polls are generally not very accurate, as they are opt in, and not a true sample. (Thus people with more impressive setups are more öikeöy to self report than those without, skewing the results) but I guess it couldn't hurt.

 

[Oldmodder]

So what is the price / FPS ratio, to me at least it would be immensely silly to pay 2 X for 10 - 20 FPS,,,,,, even if you where a 1440 / 4K games ( which i am not personally )

 

[HockeyJon]

Would love to upgrade, but flat out refuse to pay those prices, sorry.

 

[chrispix]

I am torn. I am still running a GTX 980. I should have jumped on the reduced cost 1080Ti a few weeks back, but looks like they have bounced back a bit in price..

I was planning on ordering when they announced, but $1200 was $200 over my max that I wanted to spend. If the 2080Ti for $1000 can't deliver the same performance as two 1080Ti SLI, then I guess my mind is made up, buy one 1080Ti now, and buy one when price drops a bit more.. But still paying $500 for a last gen video card still hurts..

Decisions decisions..

 

[Snowdensjacket]

I am torn. I am still running a GTX 980. I should have jumped on the reduced cost 1080Ti a few weeks back, but looks like they have bounced back a bit in price..

I was planning on ordering when they announced, but $1200 was $200 over my max that I wanted to spend. If the 2080Ti for $1000 can't deliver the same performance as two 1080Ti SLI, then I guess my mind is made up, buy one 1080Ti now, and buy one when price drops a bit more.. But still paying $500 for a last gen video card still hurts..

Decisions decisions..

There are good deals on 1080tis in the FS/FT forum right now. Under $500

 

[Snowdog]

Waiting on [H]'s numbers before even looking at any other sites. Sticking by what I said last thread. If this thing kicks ass in rasterization I can probably ignore the RTX nonsense, but maybe not at these prices.

Any legit sites reviews are fine with me. That channel looks questionable, they supposedly had RTX 2080 3 weeks ago. For a channel less than year old with less than 600 subs, they sure got lucky getting a RTX 2080 3 weeks ago, and now an RTX 2080Ti...

Edit, reading the comments from 3 weeks ago. They just took the numbers off the NVidia slide and made up some videos to go with it.

So almost certain, they just did the same thing again.

 

[CrazyRob]

Did anyone notice their Rainbow 6 Siege numbers from the second video? from 60FPS on a 1080ti to 150fps on the 2080ti. What kind of black magic had to happen to pull that off? Some sort of bottleneck on the 1080ti?

 

[SixFootDuo]

Real numbers? Just stop. Seriously. The world is not out to get you ... I promise.

Whatever benchmarks we get will be within a frame here or a frame there.

You guys literally have video proof and you're still in weird some state of denial.

 

[Mad Maxx]

Not paying an arm and a leg for it. Period.

WERD!!

 

[Wolf-R1]

Not paying an arm and a leg for it. Period.

Just like the slew of 1080 based cards that have been in excess inventory I see Nvidia having the same problem. Sure plenty of them will fly off the shelves at first but over the first three months I think demand will wane. The holiday's not withstanding.

 

[Auburn_Tigers]

Nvidia played a lot of suckers out there. They let the idea float around of a 50% increase, knowing damn well it was a lie. They never even tried
to correct that lie. Why do you think the benchmarks were so hard to come by to begin with, because they knew, if the benchmarks got leaked, then
the pre-order sales would be shit at that price, so they let the lie roll on, and the suckers kept buying.

To the ignorant post, about people hating. I like to think of it more along of the lines, that you were sucked into paying
700 more dollars, for less performance boost, than you had in the last gen cards.

 

[bexxxxxx]

Look, same site, posted 2 weeks ago:

Definitely legit, right?

 

[Hakaba]

Yeah it’s not really nvidia job to go around and stop the hype train, it’s the people’s job to do that. Consumers should be more intelligent and vote with their wallet or wait on reviews. Also the official stance from Nvidia on the 1080 to 1080 ti as announced on stage in public may have been 1-2% off depending on titles.

Some people do not believe in upgrading from generation to generation others do and others are upgrading from 6xx series cards. If benchmarks pan out then it might be the best upgrade for them and to skip the 1080.

I do not see myself upgrading and do not see a reason on hating anyone who does. If a person wants to spend 1300+ on new untested tech, cool. Calling them idiots or ignorant and whatever else that floats around in this thread seems childish.

 

[castlefire]

Can we do a who actually games at 4K, 2K, HD Pole on here? I am honestly wondering how many of our readers truly use a 4K desktop monitor.
PS: I game at 2K+ 21:9 1440.

I play on 4K DCI much higher then4k UHD

 

[jnemesh]

We all need to stop comparing the 1080ti to the 2080ti, the 2080ti is NOT the successor to the 1080ti, it's the new "Titan".

 

[Elf_Boy]

IQ and other cheats aside allowing NV (OR AMD/Intel) to pick the games to Bench is not a 'fair' comparison.

heavy driver optimization and working with the game developer can significantly improve performance.

Yes driver optimization is part of the whole GPU packaged - picking the games allows for specific and better optimization.

 

[SixFootDuo]

We all need to stop comparing the 1080ti to the 2080ti, the 2080ti is NOT the successor to the 1080ti, it's the new "Titan".

If true, that would be scary. Your logic would mean that the actually 1080 ti replacement would be ......... slower?

 

[SixFootDuo]

Look, same site, posted 2 weeks ago:

Definitely legit, right?

Sure it is.

It's well known that many 2080 ti cards have been in the wild for weeks. The numbers we are seeing are from these cards. No doubt there were double and triple digit cards out there. Common sense would tell anyone that if the numbers were horrible we would have heard about these horrible numbers.

In fact, reviewers got their cards a week ago. Embargo ends tomorrow or wed when we will start seeing reviews I think. So in a day or two.

Also, I'm gonna call it the way I see it.

Most people that have been negative about these numbers are people that are salty and sour because they don't support a mindset / lifestyle that allows then to save / set aside money to buy one. Many of you out there would have bought one had you had the money. End of story.

 

[focbde]

No regrets in nabbing that 1080 the other week.

Same here.. got one for not very much second hand...

 

[IcePickFreak]

Allow me to sumarize this thread (ok, any GPU thread)

 

[Snowdog]

Sure it is.

It's well known that many 2080 ti cards have been in the wild for weeks. The numbers we are seeing are from these cards. No doubt there were double and triple digit cards out there. Common sense would tell anyone that if the numbers were horrible we would have heard about these horrible numbers.

In fact, reviewers got their cards a week ago. Embargo ends tomorrow or wed when we will start seeing reviews I think. So in a day or two.

Also, I'm gonna call it the way I see it.

Most people that have been negative about these numbers are people that are salty and sour because they don't support a mindset / lifestyle that allows then to save / set aside money to buy one. Many of you out there would have bought one had you had the money. End of story.

It's not legit. When asked about source of previous numbers this is what the channel owner responded:

Source Syndicate3 weeks ago
Original Blog Post - https://goo.gl/1VxB4M

Basically, they just took some vague indications from a NVidia blog, made up some fake videos to go with it.

They have no cards. It's just fabrication based on numbers from blog posts.

 

[GoodBoy]

IQ and other cheats aside allowing NV (OR AMD/Intel) to pick the games to Bench is not a 'fair' comparison.

heavy driver optimization and working with the game developer can significantly improve performance.

Yes driver optimization is part of the whole GPU packaged - picking the games allows for specific and better optimization.

Hmm? You want the drivers optimized to improve performance... I mean I want my games to run as fast as they possibly can, pretty sure the rest of us do too.

I think what you were trying to say was that the vendor will choose games for the reviews based on ones that are "optimized", and that that is somehow unfair.

But really, those are the ones that best showcase the new cards capabilities. To me it's completely expected that they would do that. I look at all of those, but I also expect reviews of performance in games I play. An across the board look. The games that are not yet optimized likely will be very soon anyway (unless you have an AMD card, then you might have to wait, but even they have gotten better about this). This means that the games that have been optimized are the best predictors of the performance we will get from these cards, hence why I do not see any issue with that.

All that said, those supposed performance videos from the youtube video posted above with the synthesized voice seem pretty sketchy to me... from the little I remembered of the 2080ti vs 1080ti, the movement in BfV looked identical to the first video with the 2080 vs something... so yeah I wouldn't trust that website/channel in the slightest.

 

[Dayaks]

Sure it is.

It's well known that many 2080 ti cards have been in the wild for weeks. The numbers we are seeing are from these cards. No doubt there were double and triple digit cards out there. Common sense would tell anyone that if the numbers were horrible we would have heard about these horrible numbers.

In fact, reviewers got their cards a week ago. Embargo ends tomorrow or wed when we will start seeing reviews I think. So in a day or two.

Also, I'm gonna call it the way I see it.

Most people that have been negative about these numbers are people that are salty and sour because they don't support a mindset / lifestyle that allows then to save / set aside money to buy one. Many of you out there would have bought one had you had the money. End of story.

I totally agree it’s some kind of weird human thing. You see it elsewhere. Someone can’t afford a tropical vacation, they bash it, not worth the money, ect. They get a better job and the first thing they do is go on a god damned tropical vacation.

Anywho it is strange because nVidia only gave out the drivers a few days ago to reviewers.

We really need the full review. DLSS is important.

The new CUDA cores also have varied performance boost depending on the game. Battlefield should be an indicator of a best case scenario.

 

[nimer]

Can we do a who actually games at 4K, 2K, HD Pole on here? I am honestly wondering how many of our readers truly use a 4K desktop monitor.
PS: I game at 2K+ 21:9 1440.

I game on a PG27AQ 27" 4k, i can get 45 fps with 80% max graphics settings on any and all games. happy days 1080ti hybrid

 

[Fixall]

The jump from the GTX1080-Ti to the RTX2080-Ti is looking more like the jump from the GTX980-Ti to GTX1080 than GTX980-Ti to GTX1080-Ti. Extrapolating that, I assume that the RTX2080 may be almost as fast as a 1080-Ti, while a 2070 may in fact, be MUCH slower than a 1080-Ti, likely landing a bit above the GTX1080-Vanilla.

Witcher 3:
2080 Ti - 77fps
1080 Ti - 44fps
75% increase (numbers from OP @4k)


1080 Ti - 97.4fps
980 Ti - 56.5fps
72% increase (numbers from [H] @2k)


1080 - 75.4fps
980 Ti - 56.5fps
33% increase (numbers from [H] @2k)




Battlefield 1:
2080 Ti - 87fps
1080 Ti - 59fps
47% increase (numbers from OP)


1080 Ti - 112fps
980 Ti - 75fps
49% increase (numbers from [H] @2k)


1080 - 97.8fps
980 Ti - 75fps
30% increase (numbers from [H] @2k)


Using the numbers available in the video for the Witcher 3 and Battlefield 1 (old benchmarks not available for SotTR), the 2080 Ti looks to be a 75% improvement over the 1080 Ti in the Witcher 3 and a 47% improvement in Battlefield 1.
Using numbers from HardOCP, the 1080 Ti is a 72% improvement over the 980 Ti in the Witcher 3 and a 49% improvement in Battlefield 1.
Also using the numbers from HardOCP, the 1080 is a 33% increase over the 980Ti in the Witcher 3 and a 30% improvement in Battlefield 1.


According to the available benchmarks, the 2080 Ti looks to be almost exactly the same improvement over the 1080Ti as the 1080 Ti was over the 980 Ti.


TL;DR
Your math is shit.

I agree with you Rauelius, the performance hike from the previous gen GTX is nowhere near what it has been for previous generations if the numbers on the video are to be believed.

And how do you figure that?

 

[///AMG]

Can we do a who actually games at 4K, 2K, HD Pole on here? I am honestly wondering how many of our readers truly use a 4K desktop monitor.
PS: I game at 2K+ 21:9 1440.

I have 3 4k Monitors on my desktop, 2x 28" and one 40". I dont really game on my desktop anymore though, thats why I was fine with a 1070 I bought a month after launch. If I did I would consider the 1080 Ti but then ultimately decide the value proposition isn't that great.

 

[twonunpackmule]

Same here.. got one for not very much second hand...

Yup. Going to see what the next set of GPUs do. I think I can hold out.

 

[stamsek]

Forum polls are generally not very accurate, as they are opt in, and not a true sample. (Thus people with more impressive setups are more öikeöy to self report than those without, skewing the results) but I guess it couldn't hurt.

The polls in Genmay are always spot on and 100% accurate with a great deal of detail and thought put into them.

 

[IdiotInCharge]

We all need to stop comparing the 1080ti to the 2080ti, the 2080ti is NOT the successor to the 1080ti, it's the new "Titan".

It is what Nvidia says it is- it's their marketing. And yeah, we'll keep comparing it to the top-end previous generation consumer card, thanks.

 

[kamxam]

I'll probably end up getting one, but most likely will wait till the price drops to the supposedly original MSRP. Or if (Or when) a Titan version comes out i may go for that depending on it's performance. My Titan XP is doing fine atm, so while i have the itch to upgrade i can wait a while.

 

[exlink]

Decided to cancel my EVGA RTX 2080 Ti FTW3 pre-order today once I found out that ray tracing won't even be enabled until an October Windows update. That along with the review embargo being pushed until pretty much the last minute made me want to cancel as I saw no benefit in pre-ordering. I'll wait until some performance numbers come out in November with matured drivers and the official "release" of ray tracing before I drop $1400. Until then my 1080 Ti will continue slaying.

 

[Gmok Bonecrusha]

Can we do a who actually games at 4K, 2K, HD Pole on here? I am honestly wondering how many of our readers truly use a 4K desktop monitor.
PS: I game at 2K+ 21:9 1440.

I game at 4k with my 1080ti. Monster Hunter World which I am all in for right now, max settings gets about 40-55 FPS.

 

[RJ1892]

Though the numbers look realistic. I'm still taking this with a grain of salt, from what I heard the drivers were not even released as of this weekend to reviewers. A "reviewer" with 600 subscribers on youtube and no website, makes me feel they didn't even receive a review sample. Nvidia has a tight grip on this launch and are only providing drivers to approved by Nvidia reviewers.

 

[Hagrid]

Though the numbers look realistic. I'm still taking this with a grain of salt, from what I heard the drivers were not even released as of this weekend to reviewers. A "reviewer" with 600 subscribers on youtube and no website, makes me feel they didn't even receive a review sample. Nvidia has a tight grip on this launch and are only providing drivers to approved by Nvidia reviewers.

On the "reviews" how do we know what settings were done on the 2080ti vs 1080ti? Any of them could be fudging the numbers. I will take a big tub full of salt.

 

[reaper12]

Witcher 3:
2080 Ti - 77fps
1080 Ti - 44fps
75% increase (numbers from OP @4k)


1080 Ti - 97.4fps
980 Ti - 56.5fps
72% increase (numbers from [H] @2k)


1080 - 75.4fps
980 Ti - 56.5fps
33% increase (numbers from [H] @2k)




Battlefield 1:
2080 Ti - 87fps
1080 Ti - 59fps
47% increase (numbers from OP)


1080 Ti - 112fps
980 Ti - 75fps
49% increase (numbers from [H] @2k)


1080 - 97.8fps
980 Ti - 75fps
30% increase (numbers from [H] @2k)


Using the numbers available in the video for the Witcher 3 and Battlefield 1 (old benchmarks not available for SotTR), the 2080 Ti looks to be a 75% improvement over the 1080 Ti in the Witcher 3 and a 47% improvement in Battlefield 1.
Using numbers from HardOCP, the 1080 Ti is a 72% improvement over the 980 Ti in the Witcher 3 and a 49% improvement in Battlefield 1.
Also using the numbers from HardOCP, the 1080 is a 33% increase over the 980Ti in the Witcher 3 and a 30% improvement in Battlefield 1.


According to the available benchmarks, the 2080 Ti looks to be almost exactly the same improvement over the 1080Ti as the 1080 Ti was over the 980 Ti.


TL;DR
Your math is shit.




And how do you figure that?

Except those figures are shit. Take the Witcher 3, look at any review, Witcher 3 gets 60+ fps at 4k not 44, even the FE edition. The best available benchmarks are the ones on VideoCardz and even those figures are best case scenario as they come directly from Nvidia. Those figures, for the most part put the performance increase exactly where Tom Petersen said it would be, 35 to 45% faster.

And the guy that owns the Channel has already admitted to faking numbers already. So the latest numbers from him cannot be trusted either.

 

[Fixall]

Except those figures are shit. Take the Witcher 3, look at any review, Witcher 3 gets 60+ fps at 4k not 44, even the FE edition. The best available benchmarks are the ones on VideoCardz and even those figures are best case scenario as they come directly from Nvidia. Those figures, for the most part put the performance increase exactly where Tom Petersen said it would be, 35 to 45% faster.

And the guy that owns the Channel has already admitted to faking numbers already. So the latest numbers from him cannot be trusted either.

I have no idea what settings the guy in the video used to get the fps numbers for BF1/TW3 and I have no idea if the guy faked it by tweaking the settings for each card or something. That's why I said "Using the numbers available in the video..."

HardOCP's own testing shows that the 1080 Ti averages 48fps @4k with Ultra settings and HairWorks (HairWorks AA turned off). I certainly trust [H]'s testing over some benchmarks at VideoCardz.