• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

GeForce GTX 1180, 1170 and 1160 coming in August. Prices inside

Page 8 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Status
Not open for further replies.
960 to 1060 was a jump all the way from 28nm to 16nm- biggest change in GPU transistor tech in years. Whereas this is a change from 16nm to a slightly tweaked version of 16nm (12nm). There's going to be some power efficiency/frequency improvements, and some architecture improvements, but don't expect a jump as big as the one from 960 to 1060.

The 960 was also just a ridiculously bad card. A barely 10% improvement over the 760, what a piece of crap that card was.

I'm kind of wondering if I'm just going to run my 970 until it dies. Mine is almost four years old and the most reasonable upgrade sounds like it'll be a 2060 that'll be a little weaker than a 1070 (which is a little less than 60% faster than my 970) for $300. Blech. When I got my 970 for $340 it was about 120% faster than the four year old GTX 570 that came out at $330.
 
How is it that Nv manages to fabricate at 7nm and Intel struggles with 10nm (answer like im a moron please)?
Those terms are really nebulous now. There's really not much difference in what Intel labeled 10nm, and what others labeled 7nm. You can call them the same class.
 
I am betting only the upper models get the RTX treatment. When you get to smaller dies, it probably won't make sense to devote a section of the die to a really small (useless) RT silicon that takes away from needed Shader performance.

So I expect there will be RTX 2080Ti, RTX 2080, RTX 2070, GTX 2060, GTX 2050...
 
Yea that would make more sense. Looking forward to August 20th to see the lineup. I only want the Ti or most performant model below the Titan name. Otherwise we would be confused if there were no more Ti editions..actually that would be a way for them to make more money since people would buy based on the new lineup thinking it was the latest.
 
I am betting only the upper models get the RTX treatment. When you get to smaller dies, it probably won't make sense to devote a section of the die to a really small (useless) RT silicon that takes away from needed Shader performance.

So I expect there will be RTX 2080Ti, RTX 2080, RTX 2070, GTX 2060, GTX 2050...
This makes a lot of sense to me. It would be nice to build up an install base of the RT hardware so that game developers actually consider developing for it. I'm eager to see RT come more into the gaming mainstream
 
I can't see them doing away with Ti, but it may not be released at the same time. This might only be RTX 2080/2070 reveal.
 
960 to 1060 was a jump all the way from 28nm to 16nm- biggest change in GPU transistor tech in years. Whereas this is a change from 16nm to a slightly tweaked version of 16nm (12nm). There's going to be some power efficiency/frequency improvements, and some architecture improvements, but don't expect a jump as big as the one from 960 to 1060.

You are right but the gp104 of the 1070 was 331mm2 while the gf110 of the 570 was 520mm2.

Nvidia can always increase die sizes a little.

30% bigger die sizes, plus gddr6, plus whatever tech advancements, new compressions, caches and whatnot could bring some considerable performance advancements.
 
Not sure why are you comparing GP104 to GF110. That they both have brand names ending in 70 is irrelevant. It's comparing midrange to high end.

A better example that they can increase die sizes is GK104 to GM204. 294mm2 to 398mm2. But can they pull off the efficiency magic of the Maxwell jump again? More cores, higher frequency, and approximately same power consumption.
 
Well I buy xX70 cards and my example was specific to show how Nvidia can play with die sizes. The 570 was bigger and cheaper. Then they decided to change die sizes per tier. Now that the Fab process does not allow me have, maybe they will go back to big dies.
 
They changed the price and name of each tier. The successors to the 570 (cut down high end chip) are the Titan/780, 980 Ti, and 1080 Ti.

As I said though, you are right that they can increase size as they did from Kepler to Maxwell.
 
I need a single card to outperform 2x980TI and it looks like the next TI might pull it off. I'm itching to get away from SLI. If we are taking guesses at launch order, 2080 announced next week, $3k titan two weeks from now, $1K "gaming" Titan a month before holidays and then April for TI card.
 
GTX 2050, 2060 / RTX 2070 & 2080
https://www.youtube.com/watch?v=mWtD2bCgSuU
If they follow the same schedule of the 10xx series, the "2070 Ti/2080 Ti" models
won't appear for probably another 6 months or more.

Videocardz have proper leaks of both 256bit 2080, and 352bit 2080Ti (in addition to the Titan) device IDs. It looks, both are coming at the same time, even if it is hard for me to believe that nVidia is about to sell 754 mm2 gpu to gamers.
 
Videocardz have proper leaks of both 256bit 2080, and 352bit 2080Ti (in addition to the Titan) device IDs. It looks, both are coming at the same time, even if it is hard for me to believe that nVidia is about to sell 754 mm2 gpu to gamers.

They could always charge a grand or more for the Ti and like 3 grand for the Titan.

We all know nVidia loves to double dip but perhaps due to mining the gap won't be long enough to do that.
 
Videocardz have proper leaks of both 256bit 2080, and 352bit 2080Ti (in addition to the Titan) device IDs. It looks, both are coming at the same time, even if it is hard for me to believe that nVidia is about to sell 754 mm2 gpu to gamers.

I think it's rather convenient that the "leaks" showed up after NVidia revealed their their next generation Quadro parts, and anyone could make some "leaks" based on that info.

We are less than a week away from the real info.
 
They could always charge a grand or more for the Ti and like 3 grand for the Titan.

We all know nVidia loves to double dip but perhaps due to mining the gap won't be long enough to do that.
I also think the transitioning to 7nm in the middle of a cycle is screwing up things.
 
I also think the transitioning to 7nm in the middle of a cycle is screwing up things.

Well you had mining screw up the launch.

But yes, I imagine you will see a 7 nm x100 part by the end of the year or early next year and gaming parts in the middle of next year.
 
Status
Not open for further replies.
Back
Top