NVIDIA Announces GeForce RTX 2070, RTX 2080 and RTX 2080 Ti at Gamescom 2018

Subject: Graphics Cards | August 20, 2018 - 01:58 PM |
Tagged: turing, tensor cores, rtx 2080ti, RTX 2080, RTX 2070, rtx, rt cores, ray tracing, quadro, preorder, nvidia, gtx, geforce

Amazon.com has its preorder and product listing pages for the various GeForce RTX cards live!

* Update *

NVIDIA's pre-order page is now live, as well as info on the RTX 2070! Details below:

View Full Size

*Update 2 *

Post-Founders Edition pricing comes in a bit lower than the Founders pricing noted above:

View Full Size

* End update *

Much to no one's surprise after the launch of the Quadro RTX lineup and the many subsequent leaks, today NVIDIA officially announced their next generation GeForce RTX 2080 and 2080 Ti gaming GPUs.

Just like we saw with the Quadro RTX lineup, NVIDIA is designating these gaming-oriented graphics card with the RTX brand to emphasize their capabilities with ray tracing.

Through the combination of dedicated Ray Tracing  (RT) cores and Tensor cores for AI-powered denoising, NVIDIA is claiming these RTX GPUs are capable of high enough ray tracing performance to be used real-time in games, as shown by their demos of titles of Battlefield V, Shadow of the Tomb Raider, and Metro: Exodus.

Not every GPU in NVIDIA's lineup will be capable of this real-time ray tracing performance, with those lower tier cards retaining the traditional GTX branding.

Here are the specifications as we know them so far compared to the Quadro RTX cards, as well as the previous generation GeForce cards, and the top offering from AMD.

  RTX 2080 Ti Quadro RTX 6000 GTX 1080 Ti RTX 2080  Quadro RTX 5000 GTX 1080 TITAN V RX Vega 64 (Air) RTX 2070
GPU TU102 TU102 GP102 TU104 TU104 GP104 GV100 Vega 64 ?
GPU Cores 4352 4608 3584 2944 3072 2560 5120 4096 2304
Base Clock 1350 ? 1408 MHz 1515 ? 1607 MHz 1200 MHz 1247 MHz 1410
Boost Clock 1545
1635 FE
? 1582 MHz 1710
1800 FE
? 1733 MHz 1455 MHz 1546 MHz 1620
1710 FE
Texture Units ? ? 224 ? ? 160 320 256 ?
ROP Units ? ? 88 ? ? 64 96 64 ?
Tensor Cores 576 576 -- 384 384 -- 640 -- ?
Ray Tracing Speed 10 GRays/s 10 GRays/s -- 8 GRays/s 6? GRays/s -- -- -- 6 GRays/s
Memory 11GB 24GB 11GB 8GB 16GB 8GB 12GB  8GB 8GB
Memory Clock 14000 MHz  14000 MHz  11000 MHz 14000 MHz  14000 MHz  10000 MHz 1700 MHz 1890 MHz 14000 MHz
Memory Interface 352-bit G6 384-bit G6 352-bit G5X 256-bit G6 256-bit G6 256-bit G5X 3072-bit HBM2 2048-bit HBM2 256-bit G6
Memory Bandwidth 616GB/s 672GB/s 484 GB/s 448 GB/s 448 GB/s 320 GB/s 653 GB/s 484 GB/s 448GB/s
TDP 250W
260 FE
300 watts 250 watts 215W
225W FE
? 180 watts 250W 292 175W
185 FE
Peak Compute ? ? 10.6 TFLOPS ? ? 8.2 TFLOPS 14.9 TFLOPS 13.7 TFLOPS ?
Transistor Count ? ? 12.0 B ? ? 7.2 B 21.0 B 12.5 B ?
Process Tech 12nm 12nm 16nm 12nm 12nm 16nm 12nm 14nm ?
MSRP (current) $1200/
$1000
$6,300 $699 $800/
$700
$2,300 $549 $2,999 $499

$600/
$500

We hope to fill out the rest of the information on these GPUs in the coming days during subsequent press briefings during Gamescom.

One big change to the RTX lineup is NVIDIA's revised Founders Edition cards. Instead of the blower-style cooler that we've seen on every other NVIDIA reference design, the Founder's Edition RTX cards instead move to a dual-axial fan setup, similar to 3rd party designs in the past.

These new GPUs do not come cheaply, however, with an increased MSRP across the entire lineup when compared to the 1000-series cards. The RTX 2080 Ti's MSRP of $1200 is an increase of $500 over the previous generation GTX 1080 Ti, while the GTX 2080 sports a $200 increase over the GTX 2080. These prices will come down after the Founders Edition wave pricing passes (the same was done with the GTX 10xx launches).

Both the Founder's Edition card from NVIDIA, as well as third-party designs from partners such as EVGA and ASUS, are available for preorder from retailers including Amazon and Newegg starting today and are set to ship on August 27th.

Source: NVIDIA

August 20, 2018 | 02:11 PM - Posted by Danomite (not verified)

Can you add the 1070 to that comparison chart?
Thanks!

August 20, 2018 | 02:14 PM - Posted by Danny Lim (not verified)

Impressive specs! I will buy RTX 2080 Ti aftermarket solution from board partners on Newegg such as ASUS, EVGA, MSI, Gigabyte, etc.

August 20, 2018 | 02:17 PM - Posted by Moyenni (not verified)

Well I won't cause I'm not made of money.

August 20, 2018 | 02:48 PM - Posted by Brogen

Same as a cpu, I'm not paying over a grand for a single component. These are hype prices. I'll stay with gtx 1080 sli for the time being. It's good for another 3 years.

August 20, 2018 | 03:08 PM - Posted by WayneJetSki

lol yeah same here. And I feel like by the time the Raytracing stuff is more common to where I really want a raytracing card the next-gen ray tracing cards will be out.

Hopefully those ray tracing chips will be useful for something besides just ray tracing.

August 20, 2018 | 07:16 PM - Posted by MakeWithTheWhitePapersJHH (not verified)

Ray Tracing is a compute task that was orginially done on CPU cores. So are there really dedicated Ray Tracing cores that generate those GigaRays or are they just more like others GPU compute cores. Nvidia needs to make with the RT Core whitepapers.

Any of Nvidia's or AMD's GPUs Via CUDA or OpenCL respectively can accelerate Ray Tracing on those older GPU cores' Shaders and AMD's compute(ACE/Shaders) can do the Ray Tracing Calculations on AMD's GCN based SKUs.

Nvidia is going to have to do more deep dives into their RT cores so readers can get some idea of what exactly the RT cores bring to hardware based ray tracing over doing that same task via CUDA on Nvidia's older generation of GPUs or AMD's GCN based GPU's that use OpenCL.

Nvidia's Tensor Core/AI based denoising is really what's going to take that Limited Ray Traced grainy output and denoise that in the milliseconds time frame allotted in order to make that Hybrid Ray Traced output presentable for gaming graphics. And those GigiRays/s metrics need to be calculated down to Numbers of Rays per milliseconds because at high frame rates there are no full seconds per frame so only some few milliseconds worth of Rays will be possible at High FPS frame rates. It's those Tensor cores doing the quick and efficient Tensor Core pre trained DeNoising/Convolutaional filtering that makes Nvidia's Hybrid Ray Tracing possible or the output would just look like a grainy mess.

August 20, 2018 | 02:43 PM - Posted by Kiteman

Can you add a link to the pre-order page at Nvidia? Or any vendor? Because nobody seems to have updated their storefront yet.

August 20, 2018 | 02:53 PM - Posted by JoeSixPack (not verified)

Price gouging. They liked the higher prices that cryptocurrency caused and want to continue it indefinitely. They are justifying it with ray-tracing, and they get to say things like RTX performance is such and such many times more powerful than the Titan V, but leave out the fact that they are refering to ray-tracing performance, not the performance you'll see in actual games. By the time ray-tracing is prevalent in games and you'd actually feel deprived without it, it'll be five years down the road these initial RTX boards will be obsolete. Price gouging, but most will fall for it. If I'm wrong oh well sorry, but at least I'm not buying the initial hype train.

August 20, 2018 | 02:55 PM - Posted by Kenworth

https://www.nvidia.com/en-us/geforce/20-series/
$100-$200 higher than the presentation prices.

August 20, 2018 | 03:10 PM - Posted by Dark_wizzie

Don't buy Founder's Edition and don't pre-order. $200 extra for a basic heatsink that doesn't look as good as the old heatsinks?

They need a larger die for all this new stuff and it's going to cost more. AdoredTV is going to lose his mind.

August 20, 2018 | 03:08 PM - Posted by Everyone (not verified)

What a massive disappointment. No wonder nVidia was so vocal/adament about keeping 10 series pricing high after crypto tanked. Those "from" prices are already way too high, and from products listed so far - are devoid of reality.

600 for a *70 and 800 for a *80? Fuck you

August 20, 2018 | 03:36 PM - Posted by Dark_wizzie

All this talk about how much an 80 or 80ti or 70 card should cost is dumb. The price depends on what the market will bear. As a consumer all you can look at is price/performance because that is the only metric that matters and can be controlled.

Be upset AMD is causing this to happen.

I don't care about price, I care about price/performance.

August 20, 2018 | 03:49 PM - Posted by Everyone (not verified)

Except that those 70/80/80ti tiers represent performance points. It's reasonable to estimate the performance based on generational changes in the past. There's also nothing crazy about the specs, and there was no mention of performance gains by nVidia (outside of ray tracing). Sure, we won't know for sure how it will perform until release, but everything is pointing to one thing.

Every other generation has brought MUCH better performance for the same price - this is not the case this time. There's no generational 'discount', and yeh I'm kind of salty.

You can call it "what the market will bear," but I call it getting fucked in the ass. Trying to redirect the animosity towards AMD is not helpful. Sure, they've been consistently mediocre, but that shouldn't take away from pointing out nVidia's shitty behavior.

August 20, 2018 | 04:36 PM - Posted by Dark_wizzie

That is what companies do. They want all the profit. They all suck and they're all great at the same time. if people's idea of what is or isn't fair is dictated by profits and not price/perf then the bigger die of 2080ti is a counterargument.

And he mentioned 2070 performing better than Xp. That is a decent margin above 1080ti. So you have a $500 card that beats a $1200 card. 2080ti would probably be a good 50% faster than the 2070 which should destroy a Titan V. If 2070 is $500, 50% faster than 2070 for $1000 seems okayish. We get the 2080ti now instead of half to a year later. Historically that comes with a price drop of the 70 and 80 cards. So theoretically we get the option of ti now, or ti later with price drops. Unless Nvidia doesn't do that, in which case that sucks. None of which we can tell today.

There's no value judgement without the performance numbers, so I think the natural response is to withhold judgement. That's why I never pre-order. The model numbers are rough performance posts but they vary. That's why the 980 was a disappointment to me. I mean, here it's not looking great, but... when I go to the forums I'm sure the hate I'm going to scroll through will be INSANE.

I think the larger die size required to fit all the ray tracing stuff would drive up cost. The v100 has a larger die than the 2080ti but it costs waaaaaaaay more, while the 2080ti die should still be huge. Although, I don't even care that Nvidia paid more for a bigger die for RTX. I don't care why they're doing it or how little or much it costs them. I only care about price/perf. And I don't know what RTX really brings to the table. It's a big wildcard.

I think of all my hobbies, gamers are the most agitated by price changes I've ever seen. If people want to boycot Nvidia because they're upset then be by guest. I dunno, I think we should all take a deep breath here. :D Happy life long life.

(Also, it is kind of annoying that they're going with higher prices on FE. I thought we were done with that?)

August 21, 2018 | 04:58 AM - Posted by MrSmith (not verified)

All those performance figures related to Ray Tracing.

As you said companies can charge whatever they want, especially when they have no competition, but that doesn't mean that people should be happy about it.

But why would any customer defend the price hikes in comparison to past estimates? Do people enjoy paying more for products?

August 21, 2018 | 11:12 AM - Posted by TheRealMarketsAndTheDoshToCompete (not verified)

Do people realise that the consumer market is not the market that's paying its fair share for all that expensive R&D. And that's because the consumers do not want to pay the costs of continued large generational performance increases that the consumers themselves have mistakenly come to expect.

Nvidia's got itself a natural monopoly and has often times leveraged that monopoly in an unfair/ilegal manner. But that IP for continued GPU performance improvments cost billions of dollars to develop and if Nvidia can not use its Professional Market Markups to help pay for the costly R&D then the consumer market will have to pay its fair share.

You can properly blame the Retail channels for taking advantage of the Mining craze more than you can fully blame the GPU makers, Nvidia or AMD. It's an illegal restraint of Retail Trade for either Nvidia or AMD to enforce any MSRP pricing and that S(Suggested) part of that MSRP acronym says its all.

Nvidia, or AMD for that matter, need to be raising their wholesale DIE prices higher up to the highest market demend pricing that is allowed. Because it's not before the consumers are so turned off by the cost that they will not purchase the product if the Retailers add too much onto their retail markup pricing over their wholesale pricing. Nvidia(to a lesser degree with its FE editions) and AMD are mostly not directly involved in the wholesale pricing of products directly sold in the retail channels that contain Nvidia's or AMD's GPU dies.

That's mostly the third party AIB card makers that are directly supplying the retail channel with their finished AIB GPU card products. Nvidia and AMD mostly supply the GPU Processor Dies and some other hardware(GSync Scalar Chips etc.) to the AIB makers and monitor makers(GSync). So Nvidia and AMD mostly sell GPU dies via wholesale OEM Trey Pricing or on purchase contract basis to their AIB partners.

Nvidia started that FE edition direct marketing to consumers precisely because Nvidia has that natural market monopoly and the AIB makers can not complain much because Nvidia has that large natural market monopoly. Natural Monopolies are not illegal per se but any natural market monopoly holder that uses is monopoly to enforce unfair market practices on any market does run the risk of being declared an abusive natural monopoly market share interest and that's a violation of antitrust regulations/laws.

Really Nvidia's FE branded products are no different than Apples branded PC/Laptop/phone/tablet products and Nvidia is free to sell directly to the consumer if it so wishes.
Apple is a Trillion Dollar market cap company because Apple is a retailer more than Apple is a technology company. Nvidia Knows that its GPU only dependency has it at a disadvantage relative to AMD and Intel who make the CPUs that allow all those GPUs the chance to work on most all the PCs/Laptops sold on the market currently and for over many decades.

Nvidia is still more dependent of Consumer GPU sales for the majority of is revenues but Nvidia's non consumer related sales are growing revenues at a faster pace. AMD's has got its CPU sales to make most of its revenue stream with AMD's RTG not producing enough revenues on its own to be able to compete directly with Nvidia currently. AMD is smartly, and from a very sound business model standpoint, focusing on Epyc and its Vega based Radeon Pro WX/Instinct compute/AI GPU sales to the professional market where the markups and revenue potential is much higher.

The Wall street Quants are looking at every little bit of Gross Margin basis point percentage increase in AMD's, any Company's Gross Margins, in order to estimate AMD's future stock share price value so the big hedge funds and institutional investors know where to focus more of their funds' investments towards. So Lisa Su/Her Management Team will continue AMD's professional market focus and gamers have no choice in the matter if they want AMD to at some point obtain the necessary revenue streams to compete with Nvidia, and soon Intel's equally deep pockets in the discrete GPU market.

AMD is focusing on APU Vega Graphics and Mainstream Vega, then Navi, Graphics for the mainstream consumer market where the revenue streams potential are greatest in that non professional Consumer/GPU market. Nvidia's Integrated Graphics market share is limited relative to Intel's and AMD's. And just look at that Bastard stepchild Intel/AMD offspring and that has Nvidia very much trembling in its boots on the discrete Mobile GPU market side of the equation where Nvidia has Discrete Mobile market share to defend. Even AMD's Zen/Vega APUs have some potential to cost Nvidia sales in the low end discrete GPU market.

AMD needs to sit back and let Nvidia drive up GPU prices while AMD can come in and charge just enough less that AMD can still make a reasonable return on its investments and obtain the necessary revenues to more fully compete against Nvidia in the GPU market. AMD's Lisa Su did so very much state that AMD was non longer looking to be the low priced GPU market seller any longer.

August 21, 2018 | 12:46 PM - Posted by Rocky123 (not verified)

Yea I am sure he meant that the 2070 performs better than the Titan Xp in ray tracing but if he did not state that then he was just using marketing fluff to confuse everyone into thinking the 2070 can match the Titan Xp when that is impossible considering the 2070 only has 2304 Cuda Cores.

So yep Ray Tracing is what he meant 100% but like I said he never explained that on purpose because he got what he wanted everyone saying oh look the 2070 matches the mighty Titan Xp.

I find Nvidia kind of cringy to watch and a bit of a wind bag so I never watched the live stream at all. He reminds me of a used car salesman kinda like Raja that went over to Intel from AMD does as well.

August 21, 2018 | 12:48 PM - Posted by Rocky123 (not verified)

I meant to say I find Nvidia CEO kind of Cringy I left out CEO..lol

August 20, 2018 | 03:23 PM - Posted by djnoob (not verified)

Those prices are ludicrous. The launch price for the 1080Ti was $699. The $999 price for retail cards represents over a 42% increase series to series. No thanks.

Consider the fact that Nvidia has been posting record profits and their position in the market, this price increase is greedy AF.

Been a Nvidia user for over 15 years. For the first time I'm seriously considering switching. If it wasn't for my gsync monitor I'd already decided to switch.

August 20, 2018 | 03:33 PM - Posted by toxinate (not verified)

They got you already.

August 20, 2018 | 03:39 PM - Posted by Dark_wizzie

They have a larger die.

The 1080ti launched at $700 many months after the 1080 came out, after which the 1080 dropped. Now we're getting it on launch instead. While it remains to be seen if there will actually be price drops this time around, it's not apples to apples.

August 22, 2018 | 12:12 AM - Posted by Luthair

Unless nvidia is intending to replace the Titan line with the 1080 Ti its an apples to apples comparison. nvidia knew what the die size would be and they chose to go ahead with the design.

Vote with your dollar.

August 20, 2018 | 04:25 PM - Posted by Randy (not verified)

"G-Sync Monitor"

That's why I try to avoid buying into proprietary tech. Not only did that cost you about $200 more for your monitor.....got Apple'd by Nvidia....new products you don't love but feel locked in to buy because of the environment you already bought into.

I know this is not always easy to avoid and I'm not saying you did anything wrong....my comments are more towards my despise of this business practice. There needs to be an open standard for things like Gsync and FreeSync.

August 21, 2018 | 12:55 PM - Posted by Rocky123 (not verified)

Yea I have been a PC gamer since as far back as the late 80's and have noticed the graphics cards prices climb over the years but they at least used to go up just slightly each gen. The most I paid for a video card was like $650CAD-$750CAD then after those few cards I started buying used for a while and got a couple great cards.

My current card the only reason I got it was because I have a store and my supplier found out it was my Bday coming up so he gave me a Sapphire tri-x 390x 8GB at his cost price which was $335CAD and the retail was like $569CAD-$649CAD. The strange thing was I had asked him about the best Nvidia graphics card...lol

August 20, 2018 | 03:28 PM - Posted by Anonymous409 (not verified)

I look forward to AMDs cards whenever they should up to fix the unrealistic pricing.

August 20, 2018 | 06:05 PM - Posted by Randy (not verified)

Vega did nothing to Nvidia's pricing.....in fact I would say so far AMD has helped make this happen by not being as competitive in this space as they are in CPUs. Here's to hoping they get there however.

August 20, 2018 | 07:43 PM - Posted by MakeWithTheWhitePapersJHH (not verified)

Just give AMD the $5,000,000,000 needed and then they can whip somthing up. AMD should continue to focus on the Pro GPU market where the Vega dies command a better markup and only focus on competing with Nvidia's mainstream GPU SKUs.

AMD also needs higher GPU prices like Nvidia as that's how Nvidia pays those billions for the R&D costs. So AMD should let Nvidia have the Flagship GPU Market. AMD can get more Vega graphics installed based via its Raven Ridge APU's that have Zen take Vega graphics along for the ride in the integarted graphics market. Intel's got that market currently but Raven Ridge/Vega Graphics is getting a slice of that pie. I'd rather see some Vega/Navi discrete mobile GPU offerings from AMD than any Flagship GPUs and also some beefer Zen/Vega APUs for the home system builder market instead of only the console market.

Nvidia should have charged more because people will pay and that would mean that AMD also would be able to charge more and be able to properly fund more competative offerings against Nvidia's.

Everyone blames AMD but no one's ponying up the Dosh for AMD to compete with Nvidia on a level playing field. Being competative means matching dollar for dollar the R&D invesments and that includes software/driver support and paying the games makers' costs to design their games to make use of any new GPU hardware features. Nvidia has to pay those Games Companies' develpoer hours or they will not target Nvidia's hardware features, ditto for AMD. Someone please show AMD the MONEY(billions of $) to afford this competition.

August 21, 2018 | 01:03 AM - Posted by thename (not verified)

I agree with wall of text poster here. AMD did managed to match nVidia in performance the past and nVidia responded by matching their price. Consumers then flock towards The Way It's Meant to be Played™. The same thing will repeat next year. Blaming AMD instead of nVidia for the current price of GPUs only self-perpetuates the issue. It's only logical for AMD to refocus on the professional and semi customs market where there is return on investment. Your only real hope now is Intel's GPU coming to fruition in 2020.

August 20, 2018 | 04:14 PM - Posted by Anonymously Anonymous (not verified)

More interested in what performance the 20* series cards have in games that don't have ray tracing.

August 20, 2018 | 10:08 PM - Posted by SpitballingWithJK (not verified)

Until you get the ROP/TMU counts on the 20** series to compare them to the Pascal series variants you can not even begin to estimate what how non ray traced games will perform. If the 2080 still only has 64 ROPs and the same TMU counts, then what other tweaks of the ROP/Cache and the Shader core tweaks will be there to make performance better. Better clocks will help, so will larger L2 cache sizes and improved Shader cores. The 12nm process for TSMC what are the listed improvments there also.

The complete specifications are still not out for the 20** series and missing the TMU/ROP couts is a big F--K-UP on Nvidia's part amd makes them look stupid. Jeff Kampman at TR is taking a stab at Turing's performance and maybe he can be asked about current non RTX optimuzed titles running on those RTX GPU Turing variants. In that article's tables Kampman at least includes the TMU/ROP numbers from the previous Pascal 1080/1080Ti SKUs with a big(?) there to indicate that that's currently unknown for Turing.

(1)

"Spitballing Nvidia's purported GeForce RTX 2080 and RTX 2080 Ti
by Jeff Kampman — 4:57 PM on August 17, 2018 "

https://techreport.com/news/34003/spitballing-nvidia-purported-geforce-r...

August 20, 2018 | 04:29 PM - Posted by Bezzell

These prices, MSRP and Founder's, are completely out of control.

August 21, 2018 | 06:29 AM - Posted by Rustknuckle (not verified)

No shit, here in Norway I paid after taxes and margin for the local vendor $1550 for the 2080ti when I pre-ordered

August 20, 2018 | 04:46 PM - Posted by JediMaster-Allyn (not verified)

Why does this Ray tracing feel like it's going to be an Nvidia Gameworks or Hairworks type of thing...lol

August 20, 2018 | 06:14 PM - Posted by WaitingForDeeperDivesOnTheTuringIP (not verified)

Well Nvidia did not hold the GTX 2080Ti's release up this time around so that tells me that maybe Nvidia must expect some sorts of answer from AMD. Or it could be that either Nvidia expects better competition from AMD sooner or Nvidia is just competing more with its past GTX/Pascal offerings. So that will have to wait for any of AMD's 7nm offerings and probably the Navi GPU micro-arch to arrive.

Also according to anandtech:

"Meanwhile the RTX 2070 will not support SLI; this being a departure from the 1070 which did offer it." (1)

That NVLink based SLI for dual GPU usage needs a closer look ASAP from Anandtech/Others as I'm very interested in any dual GPU scaling issues that may have been resolved via NVLink technology compared to previous SLI bridge IP. Maybe NVLink in the hardware will allow the drivers and software to see any dual GPUs as just one larger logical GPU for GPU scheduling and vastly improve milti-GPU scaling efficiencies for any NVLink based multi-GPU usage from Nvidia's NVLink enabled SKUs.

(1)

"
NVIDIA Announces the GeForce RTX 20 Series: RTX 2080 Ti & 2080 on Sept. 20th, RTX 2070 in October

by Ryan Smith on August 20, 2018 4:00 PM EST "

https://www.anandtech.com/show/13249/nvidia-announces-geforce-rtx-20-ser...

August 20, 2018 | 06:21 PM - Posted by Odizzido2 (not verified)

I will wait for benchmarks but my initial reaction is....meh?

August 20, 2018 | 06:32 PM - Posted by AnonymousUser (not verified)

Yikes, that's pricey

August 20, 2018 | 08:34 PM - Posted by Anonymice (not verified)

$1200? Thats about what I paid for TWO 1080ti s.

August 20, 2018 | 08:43 PM - Posted by Hood

The reason for the high price is because NVIDIA screwed up and made too many 10 series cards. If everyone who needs an upgrade goes for the 1070, 1080, and 1080 Ti(new & used), after a couple months they can get a decent card for $200-$400, and wait a couple years for ray tracing games to mature before worrying about it. With their new Pascal card, their favorite games will run better than ever, and the Turing-optimized games will take a year or more to fill out titles that matter to anyone. With millions of unsold 10 series cards,and millions of former mining cards flooding the market, all we have to do is be patient, hold off from paying the high prices, and they will get lower, on everything ( even the 20 series will have to drop if they're not selling well).
This is a unique situation as far as overstocked and over-priced at the same time - something has to give, it's not going to be me that gives in and pays full list for a 2-year-old card, and certainly not the inflated list for the new cards.

August 20, 2018 | 09:18 PM - Posted by 50004000 (not verified)

We more than likely will not see large scale adoption of Ray Tracing technolgy unless its adopted by the consoles which are currently made with amd gpus. Which means that these cards will be judged on their DirectX 11/12 performance for years to come. I am expecting reasonably performace gains at 1440p/2160p over Pascal but not 30%-50% improvements in AAA titles. You noticed today that they spent little time talking about actualy gaming peroformance, which I believe will be telling when the reviews drop.

Pcper please include 4k resolution test for the 2080ti cards as I am really curious to see if we actually get any gains over the 1080ti with this card.

August 20, 2018 | 09:31 PM - Posted by RayTracingOnGPUsIsNotNEW (not verified)

With Nvidia's deep pockets you will see Nvidia pay to get that adoption with or without any console help.

And that's part of why Nvidia's new RTX GPUs cost more, you have to pay for the hardware/software/firmware engineers some way. You'll have to Pay to Game with Rays.

Nvidia's go the Green to get the adoption part done PDQ!

August 21, 2018 | 06:37 AM - Posted by Rustknuckle (not verified)

The reason I think this will be adopted on a large scale in time is that ray tracings main benefit is not for consumer, but for developers. It should allow large amounts of work that today is done manually to be automated leading to cost and time savings that may either be putt into other parts of development or just taken out as lower development costs to the joy of shareholders.

August 21, 2018 | 06:25 AM - Posted by Anonymous23cs1 (not verified)

Lol all they did was renamed titan the Ti so people will think this is it and wont wait 6 months for cheaper $799 version. Wonder what will replace Ti in 6 months for $799.

August 21, 2018 | 10:00 AM - Posted by Martin (not verified)

To everyone commenting about pricing, just look at GDDR6 (which is new and expensive) and more importantly - die sizes. Those are seriously huge.

TU104 (2080) is somewhere between 400 and 500 mm2 and TU102 (2080Ti) is 754mm2. Compare this to GP104 (1080) and GP102 (1080Ti) which were 314 and 471 mm2.

August 23, 2018 | 05:31 PM - Posted by boniek (not verified)

Who cares how big the dies size is or how expensive GDDR6 is? No consumer should care. Only thing that matters is price to performance ratio.

August 21, 2018 | 04:49 PM - Posted by Zorkwiz

If anyone was holding out for a new GPU who really needs an upgrade, the 1080/1080Ti prices in the next couple of weeks will be a steal. I saw a Zotac Ti today for $526 on Amazon.

The fact is that while they are calling this newer TSMC process "12nm NFF", that it is essentially a tweaked 16nm process much as Intel has been dong for the past few years. The next real jump in clocks and ROP/CUDA performance for traditional rasterizers may come when they switch to their "7nm" process in the next year+.

I'm definitely not expecting a big leap in traditional game performance from RTX 20xx based on their silence, and the Ti pricing is just ludicrous compared to 1080 Ti at launch.

I wonder if Tom will make himself available for this gen's PcPer review/benchmarking session, or if the 10-minute podcast Ryan released with nothing but RTX hype will be all we hear from him till the next revision and die shrink...

This is a launch to profit from pent-up demand and nothing more. Even the games shown with RTX support coming were revealed to be mostly implementing DLSS with the tensor core and not even Ray Tracing. Not exactly a huge benefit and we don't know what, if any, performance hit comes with using the Tensor core or ray tracing capabilities alongside the rest of the chip, especially in terms of stable clocks.

I look forward to the true reviews, but I'm not optimistic at this point that this was worth a 28-month wait.

August 21, 2018 | 07:27 PM - Posted by zgradt

Wow. For a long time in the 90s and 2000s, my graphics card buying philosophy was to replace it when one was released that was twice as fast as the one I had and less than $200. A lot of times this meant the former flagship card had dropped below $200. I think I finally had to give this up around the time I got a Radeon 5850. It's just gotten worse since.

August 21, 2018 | 10:32 PM - Posted by Oscar Castillo

Is the shared frame buffer with NVlink exclusive to the Quadro RTX?

Post new comment

The content of this field is kept private and will not be shown publicly.
  • Lines and paragraphs break automatically.
  • Allowed HTML tags: <a> <em> <strong> <cite> <code> <ul> <ol> <li> <dl> <dt> <dd> <blockquote><p><br>
  • Web page addresses and e-mail addresses turn into links automatically.

More information about formatting options

CAPTCHA
This question is for testing whether you are a human visitor and to prevent automated spam submissions.