Nvidia unveils the GeForce RTX 2080 Ti, RTX 2080, and RTX 2070

[philehidiot]

That's not quite how I read Phils comment. It's more that Nvidia have created this brand new tech but have priced it out so high adoption is going to be deterred.

Yep. Devs won't put time and effort into high end stuff only a small percentage has when that time can be spent optimising the results from the more widespread tech. The way you change a market is by taking a risk and encouraging mass adoption - the rewards are huge but the risk is high also. Here, Nvidia are offloading the risk onto early adopters.

And when I say about new features I'm not just on about the raytracing - that's the most obvious one and we know there will be DirectX support for it so that's not so bad. Their stupid new metric that they've invented to make it look 6x better than the previous generation is just annoying marketing rubbish. There are other things they have done with their "AI" cores which will need coding for I'm sure. Will devs bother if the tech is confined to only a small percentage of the market? Of course not. If you want to change the market (as they admit they do), you need to accept there's a financial risk involved and you have to shoulder that risk. Increasing the prices to put the risk on the enthusiastic early adopters will reduce market uptake and reduce the chance of your project working. This kind of disruption is about flooding the market with compatible products at reasonable prices and then devs will pick up on the tech.

[Ozaron]

I hope nVidia get stung when the competition finally arrives, whatever form it's in. This is totally predictable.. and yet still absurdity.

Did I miss something or is there no mention of standard TFLOPS? Every feature and metric they measured this with was an abstraction, and compared only (favourably, biased) against Pascal. Not that Pascal has any serious rivals.

[Percy1983]

The problem is these will sell by the bucket load completely validating the price.

Will stay with my RX580.

[gordon861]

We need to see the actual performance comparisons before really making a decision on the price. If it turns out the 2070 wipes the floor with the 1080ti then prices are OK, but I doubt it.

I've recently built a new PC but stuck with the 980 for now, I can probably stay with this until I move up to a large 4k monitor (unless developers really start pushing the new RTX facilities).

[shaithis]

We need to see the actual performance comparisons before really making a decision on the price. If it turns out the 2070 wipes the floor with the 1080ti then prices are OK, but I doubt it.

Even then, normally new releases cost roughly/relatively the same as the launch price of the model it replaced. It seems that recently they have realised they don't need to do that any more and can just keep adding $100+ to each new generation.

[dangel]

I hope nVidia get stung when the competition finally arrives, whatever form it's in. This is totally predictable.. and yet still absurdity.

Did I miss something or is there no mention of standard TFLOPS? Every feature and metric they measured this with was an abstraction, and compared only (favourably, biased) against Pascal. Not that Pascal has any serious rivals.

Single Precision Perf. 13.4 TFLOPs 10.1 TFLOPs 7.5 TFLOPs 8.9 TFLOPs

TI, 2080, 2070, 1080 respectively. 14TFLOPS for the FE on the RTX2080ti (due to higher clocks).

At the risk of being stoned to death i'm thinking of taking a FE TI as a replacement for SLI 1080s. To my mind the 1080ti was a nice bump over the 1080 and the 2080ti one above that. I can also get the card considerably cheaper than usual..

[Nifl]

Oh nice! Time to buy a $1200 GPU that uses 30% of its silicon for 98% of the games that will be released in the coming years! Time to win the lottery or rob a bank because common sense is dead

[eugenius]

Think about it this way: NVidia is doing society a favor by making gaming too expensive for the average teen.

[peterb]

Think about it this way: NVidia is doing society a favor by making gaming too expensive for the average teen.

I don't know - having seen the phones some teenagers seem to carry!

But on a more serious note - the fact that newer (shinier!) cards have been released doesn't mean the existing ones stop working overnight (unless Nvidia has incorporated a kill switch) so what was good enough yesterday is still good enough today. It may not be good enough in a year (or two years time) but that depends on what you are doing with it.

Yep. Devs won't put time and effort into high end stuff only a small percentage has when that time can be spent optimising the results from the more widespread tech. The way you change a market is by taking a risk and encouraging mass adoption - the rewards are huge but the risk is high also. Here, Nvidia are offloading the risk onto early adopters.

And when I say about new features I'm not just on about the raytracing - that's the most obvious one and we know there will be DirectX support for it so that's not so bad. Their stupid new metric that they've invented to make it look 6x better than the previous generation is just annoying marketing rubbish. There are other things they have done with their "AI" cores which will need coding for I'm sure. Will devs bother if the tech is confined to only a small percentage of the market? Of course not. If you want to change the market (as they admit they do), you need to accept there's a financial risk involved and you have to shoulder that risk. Increasing the prices to put the risk on the enthusiastic early adopters will reduce market uptake and reduce the chance of your project working. This kind of disruption is about flooding the market with compatible products at reasonable prices and then devs will pick up on the tech.

But early adopters always take a risk, either financial, or risk that there won't be followers and the new tech will die a death. I tend not to adopt new technology until it has been established for at least 12 months or more, usually because what I have is good enough and the new tech (unless its something really significant) rarely justifies the risk - financial or technical.

[Dashers]

I don't know - having seen the phones some teenagers seem to carry!

But on a more serious note - the fact that newer (shinier!) cards have been released doesn't mean the existing ones stop working overnight (unless Nvidia has incorporated a kill switch) so what was good enough yesterday is still good enough today. It may not be good enough in a year (or two years time) but that depends on what you are doing with it.

This is an important consideration. I'm usually struggling on top end games by the time I'm two generations behind (970), but I'm not currently (although I'll admit I'm behind on my gaming). Even if the bar is raised with these chips, there is plenty of time for prices and competition to change before I have to upgrade.

[philehidiot]

But early adopters always take a risk, either financial, or risk that there won't be followers and the new tech will die a death.

But this is not your standard early adopter risk, this is Nvidia coming out and saying they want to totally disrupt the market and change how games are made. Early adopters will be few and far between. If you want to achieve their goals with widespread disruption you can't expect the average Joe to take on the early adopter risk, you have to do it as the company in order to ensure the market is saturated with the product to a degree with means your goals are achievable. They're wanting their cake and to eat it by changing the market but pushing the risk onto the average consumer. That's different to standard early adopter risk.

[peterb]

But this is not your standard early adopter risk, this is Nvidia coming out and saying they want to totally disrupt the market and change how games are made. Early adopters will be few and far between. If you want to achieve their goals with widespread disruption you can't expect the average Joe to take on the early adopter risk, you have to do it as the company in order to ensure the market is saturated with the product to a degree with means your goals are achievable. They're wanting their cake and to eat it by changing the market but pushing the risk onto the average consumer. That's different to standard early adopter risk.

Only to a degree - and it is assuming that there will be early adoption. Current cards don't stop working overnight so the risk Nvidia are taking is that there will be early adopters. Of course they will be subsidising to developers to create applications that take advantage of the unique properties of the card - but that is a risk too - if consumers are happy with existing products, they won't buy either.

Of course there will always be some who want the latest and greatest regardless... but then "fools rush in where angels fear to tread".

And it comes down to numbers.

[dangel]

Of course there will always be some who want the latest and greatest regardless... but then "fools rush in where angels fear to tread".

...well considering my username, I'm perhaps both

[zaph0d]

One particularly bad thing is that the 2080 & 2080ti both being on the t104 chip, that's the second tier, they've dropped everything down Another tier meaning they're gonna screw the early adopters over again in 6 months or so with the full t100 chip.
To compound this it means that the 2070 is most likely on the t106, the chip that was designed to go into the **60 class for the last couple gens, and before that was the **50 class.
They keep putting lower and lower tier chips into higher tier cards. Can't wait for the AdoredTV spin on this.

[kalniel]

They keep putting lower and lower tier chips into higher tier cards.

So by definition, aren't they now higher tier chips then?

[zaph0d]

They keep putting lower and lower tier chips into higher tier cards.

So by definition, aren't they now higher tier chips then?

No - it's like putting a Fiat engine in a Ferrari, you wouldn't call that a high tier engine just because it's in a more expensive car would you?