NVIDIA's Turing-based 2000 series graphics cards are finally official, and partners are unleashing all manner of custom cards based on the new GPU. EVGA is launching the RTX 2080 Ti and RTX 2080 under a new XC Ultra Gaming series that uses a translucent shroud (with a very Gameboy Color nostalgia vibe) that wraps a dual fan ICX2 cooler in customizable white, black, and red trim and a large multi-heatpipe cooler to pair with the Turing GPU and GDDR6 memory.
EVGA is introducing four XC Ultra Gaming series cards, with two RTX 2080 Tis and two RTX 2080s which differ in price and boost clockspeeds. The graphics cards feature 2.75 slot designs with ICX2 coolers and hydro dynamic bearing fans. EVGA claims the cooler is 14% cooler and 19% quieter. The taller card design reportedly allows for a taller fan hub and thicker blades that can push air through the thicker heatsink without extra noise (whereas its 2-slot cards use a smaller fan hub with more blades to try to balance things). Display outputs include three DisplayPort, one HDMI, and one USB-C VirtualLink.
The EVGA RTX 2080 Ti XC Ultra Gaming comes in two models: the 11G-P4-2383-KR and 11G-P4-2382-KR. Memory clocks on the 11GB of GDDR6 memory is clocked at 14000 MHz on both models, but the $1,199.99 11G-P4-2382-KR features a 1635 MHz boost clock for its 4352 CUDA cores while the $1,249.99 11G-P4-2383-KR takes things up a notch to a 1650 MHz boost clock. Of course, enthusiasts can use EVGA's Precision X1 or NVIDIA's new OC Scanner software to overclock on their own. The RTX 2080 Ti graphics cards have 2 8-pin power connectors.
As far as the RTX 2080 XC Ultra Gaming cards, the $799.99 08G-P4-2182-KR and the $849.99 08G-P4-2183-KR pair a TU104 GPU with 2944 CUDA cores with 8GB of GDDR6 memory clocked at 14000 MHz. The cheaper model features a 1815 MHz boost clock while the higher priced model clocks in at 1850 MHz. EVGA's RTX 2080 XC Ultra Gaming cards use a 6+8 pin power connectors.
EVGA's XC Ultra Gaming cards come with a 3-year warranty and are currently being offered on the company's website. While they were previously available for pre-order, at the time of writing the cards are listed as auto-notify presumably due to the launch window slipping back a week.
What are your thoughts on EVGA's take on Turing?
Related:
Where are the hybrids?
Where are the hybrids?
so basically, what Nvidia has
so basically, what Nvidia has done is not only adjust the relative performance levels of their product lineup, in their favor, but also INCREASE the prices as well. Wow…Nvidia…you guys are gangsters.
Most companies, including Nvidia, do one or the other, but not both, at the SAME DAMNED TIME. For instance, when you go to buy a box of cornflakes, you might notice that while it still costs $3.99, the net wet amount has decreased from 18oz to 15oz. No big deal right? Well, enter Nvidia. They’re not only reducing the net wet amount, but also raising the price to $4.99.
My inner “Tony Montana” has a hard-on right now
nVidia is a BUSINESS not a
nVidia is a BUSINESS not a church of video games.
Actually poor gamers should treat their addiction… or better find a JOB.
“when you go to buy a box of
“when you go to buy a box of cornflakes, you might notice that while it still costs $3.99, the net wet amount has decreased from 18oz to 15oz.”
That’s called Shrinkflation the price stays the same and the total unit weight goes down! Ever notice the Toilet Paper rolls getting narrower and narrower over time. I have an old TP roll that I keep for emergencies and the newer rolls are 1 full inch narrower!
Nvidia has kept the ROP counts the same on its RTX series cards and as such there is not going to be very much in the way of greater FPS increases on any Legacy Gaming Titles compared to the Pascal GPU generation. But Nvidia can charge more because AMD currently does not have the funds to compete dollar for dollar with Nvidia in the R&D investments game and both the GP102 base and the TU102 base dies are really for the professional markets. So all the consumer GP/TU102 variants are actually binned from the Pro Market base die GP/TU102 die that has 96 ROPs binned down to 88 ROPs for the ##80Ti variants.
AMD, in a very short while, is really is going to make more off of its Epyc CPU sales than Nvidia makes from its GPU sales. So AMD has no great financial motivation to field a flagship GPU until maybe the Vega 20 die based defective/non-performant Vega 20 DIEs begin to accumulate and AMD can not use them for the Professional SKUs. AMD is only going to be targeting Integrated Graphics and Mainstream Discrete GPUs. There is no reason for AMD to spend any excess billions when Epyc is really going to make for AMD’s bread and butter. Look for AMD to let the console makers pay for any Tensor Core/AI denoising IP and R&D because it’s the AI based denoising and AI anti aliasing/DLSS that will bring the most performance gains to GPUs and not Ray Tracing as much.
Ray Tracing on Nvidia’s RTX SKUs only is able to be used because of the Trained AI that doing the Denoising on the rather crappy Ray Tracing cores output. Do the math at 16.67 milliseconds how many ray are going to be able to be generated for a scene for any 60 FPS gaming, 33.33nm per frame for 30 FPS frame rates.
Vega 20 is going to get some AI related extensions because the Professional AI markets will demand AI and Compute(FP 64 at 1/2 FP 32 rates). And really the AI based DLSS, AI Deniosing/anti aliasing is really where the performance boost will come from. Ray Tracing will be great for enhanced Reflections, Refractions, ambient occlusion, shadow and lighting effects but that’s a parallel pipeline as Denoising/AA really take loads of processing power done the old shader based way.
Gamers have to realize that if the Professional base DIEs(GP102, TU102, Vega 20) are performant enough then the Professional markets will pay way more for those performant dies that any consumer market ever could. The professional markets can write the costs off as a business expense but the consumer market’s Flagship GPUs all come from the non performant Pro Based Binned DIE variants that build up over time. Nvidia sure had to be really delaying Turing for a long time, because of the mining GPU demand mostly, before Nvidia was motivated to release its RTX branded SKUs. And really Nvidia could have released Turing when it released Volta because how else could you explain there already being enough non performant TU102 dies available to release the RTX 1080Ti so early in the game.
That last Mining Craze really upset Nvidia’s release cadence but Nvidia used that Mining Craze to Keep Pascal for sale at inflated prices for much longer than would normally have happened. Nvidia is always all about the dollar and AMD will happily earn more from its Epyc sales alone while also trying to focus the most on the Professional GPU compute/AI market where the markups are the best. AMD is not really needing to focus on anything but mainstream gaming SKUs until AMD has enough non performant Vega 20 dies filling up the bins that can not be sold in the Pro market. AMD’s all about the mainstream GPU market where the majority of revenues can be had, that and Integrated Vega Graphics and semi-custom console SKUs where the customer pays for the R&D costs.
Edit 33.33nm per frame
To
Edit 33.33nm per frame
To 33.33ms per frame
Edit: RTX 1080Ti
To RTX
Edit: RTX 1080Ti
To RTX 2080Ti
damn it I’m Retarded!
Nvidiot prices are memes,
Nvidiot prices are memes, pewdiepie needs to give it the clap so prices die.
There is still plenty of GTX
There is still plenty of GTX 1080Ti out there at fire sale pricing so double up on that Flagship Legacy Gaming Title Gamers. Dual GTX 1080Ti gaming is nothing to sneeze at.
Nvidiots are more about impressing Vern over at the neighboring rusty doublewide than it is about gaming. Just look at all that LED Bling and Expensive Cases markets.
JHH wants to sell anything green, so let them east Pascal if they can not pony up the dosh for RTX!