Tuesday, November 15th 2022
NVIDIA GeForce RTX 4080 Now Available, Starting $1200
NVIDIA today formally launched the GeForce RTX 4080 graphics card, its second fastest offering from the GeForce "Ada Lovelace" generation. With a starting price of USD $1,200, the card is positioned a notch below the RTX 4090 flagship, a whole $400 cheaper. It is technically supposed to succeed the RTX 3080 12 GB, while an RTX 4080 12 GB variant was supposed to succeed the RTX 3080 10 GB. NVIDIA cancelled the RTX 4080 12 GB as it heaped bad press due to its specs being significantly different from those of the RTX 4080 16 GB, making this the only SKU with the name RTX 4080.
The GeForce RTX 4080 is based on the 4 nm "AD103" silicon, and armed with 9,728 CUDA cores across 76 streaming multiprocessors. It gets 304 4th generation Tensor cores, and 76 RT cores, besides 112 ROPs. Although it has generationally more memory at 16 GB, its memory bus is narrower at 256-bit GDDR6X. NVIDIA attempted to compensate for this with use of faster 22.4 Gbps-rated memory, and architectural improvements such as larger caches on the silicon, to speed up the memory sub-system. NVIDIA is launching not just the Founders Edition card, but also its partners are launching custom-design boards. Every partner's lineup we've come across thus far includes at least one SKU priced at the $1,200 baseline. The cards should be on the shelves tomorrow (November 16, 2022).
We have a large number of reviews for you today, which include the NVIDIA RTX 4080 Founders Edition, ZOTAC RTX 4080 AMP Extreme AIRO, ASUS ROG Strix RTX 4080 OC, MSI RTX 4080 SUPRIM X, Colorful RTX 4080 Ultra White OC, Gainward RTX 4080 Phantom GS, MSI RTX 4080 Gaming X Trio, and the PNY XLR8 RTX 4080 Verto OC
The GeForce RTX 4080 is based on the 4 nm "AD103" silicon, and armed with 9,728 CUDA cores across 76 streaming multiprocessors. It gets 304 4th generation Tensor cores, and 76 RT cores, besides 112 ROPs. Although it has generationally more memory at 16 GB, its memory bus is narrower at 256-bit GDDR6X. NVIDIA attempted to compensate for this with use of faster 22.4 Gbps-rated memory, and architectural improvements such as larger caches on the silicon, to speed up the memory sub-system. NVIDIA is launching not just the Founders Edition card, but also its partners are launching custom-design boards. Every partner's lineup we've come across thus far includes at least one SKU priced at the $1,200 baseline. The cards should be on the shelves tomorrow (November 16, 2022).
We have a large number of reviews for you today, which include the NVIDIA RTX 4080 Founders Edition, ZOTAC RTX 4080 AMP Extreme AIRO, ASUS ROG Strix RTX 4080 OC, MSI RTX 4080 SUPRIM X, Colorful RTX 4080 Ultra White OC, Gainward RTX 4080 Phantom GS, MSI RTX 4080 Gaming X Trio, and the PNY XLR8 RTX 4080 Verto OC
66 Comments on NVIDIA GeForce RTX 4080 Now Available, Starting $1200
I doubt RTX 4090's price to come down from the $1600, but I do expect RTX 4080's price to drop at $1000 and we might see an RTX 4080 Ti in it's place for $1200.
The 3000 series can exist next to this just fine....which is probably by design thanks for their cryptocurrency supporting shenanigans.
Pls dont support this, it ruins it for all of us.
Current lowest pricing - price difference +185%, performance difference +77%.
NVIDIA GeForce RTX 3080 Grafikkarte günstig kaufen | Black Friday 2022 | Preisvergleich bei idealo
NVIDIA GeForce RTX 4090 Grafikkarte günstig kaufen | Black Friday 2022 | Preisvergleich bei idealo
NVIDIA GeForce RTX 3080 Specs | TechPowerUp GPU Database
I prefer not to buy anything again for a decade and instead play the whole lot of older games I didn't have the time yet to play, rather than congratulate them by buying their overpriced hardware...
4080 is in the price range where 4080 to should be.
So my choice is still either 4090 or simply go cheaper and look over at amd what there cards can do at a lower price.
The prices in EUR will be even more ridiculous, just fricking no.
RTX 3080 = $699
RTX 4080 = $1199
Don't try and defend it. Nobody should defend it. Nobody should accept it.
Yes it's a powerful card but it can only be truly great compared to the last gen. That's how GFX cards work. And is it worth the 90% price uplift on the previous gen? No. It's not.
A massive gripe of mine is the massive size of AIB cards, why does a 320w card need to be so much larger than last gen's 320w card, they're sizing them out of a lot of systems as well as pricing them out of budgets.
This exotic silicon is priced out of reach for many, me included, but it`s siblings will hopefully be more economically convenient. Financially balanced (not necessarily rich) people that gaming is their hobby and their free time worth to than a lot (that is, when you have small to medium children). This quite big segment of people will put the extra cash just to get the "best experience" in the scarce time they have only for themselves.
If you comparer it to other hobbies (subdividing, skiing, watching the stars with telescope, flaying drones or RC cars ect), than 1000-2000$ is not that much for the equipment (the GPU) that enable you to MAX your experience. Diminishing return- getting less when paying more near the top (when talking product tier) . It`s not new, don't try to twist it as a conspiracy to milk you. It is very common, well known and practically obligated business practice. Think what will happen if it was the other way around- getting more performance while adding less money to have the best.
And I agree- size, and more correctly volume, is the new BIG problem poses by GPU`s, mostly NV GPU`s.
Many will struggle to install them in an existing case.
One thing is to upgrade your PSU, but to replace the whole case will, rightfully, deter many people.
I hope AMD's
volunteer marketing departmentpeople do slowly start to realise that, but I won't hold my breath.Or are the cards just produced in really low quantities?
Just wished the rest of the stack downwards also got some good cuts. You gotta back that up with some evidence.
Shred Nvidia? well buying PhysX and making it proprietary is anti consumer and anti progress.
DLSS being a closed off standard is anti consumer and anti progress
Gsync....same story again.
Back in the day, AMD gave them flac for the fake amount of Vram on the GTX970, easy jabs sure but nothing was twisted to be negative, it was just a fact of false advertisement.
AMD meanwhile made open standards for all of that (well physx just went away really)
AMD also championed Mantle, which later became Vulcan, which again was totally open for all to use and had some positive changes planned (audio focussed improvements) and arguably also contributed how DX12 ended up being.
I think the current remarks regarding DP 1.4 are perhaps a bit silly, but at the same time its also silly for Nvidia to charge this much and then not offer the latest tech.
I also think the 7900XT pricing is too close to the XTX but specs wise there is a bigger gab between the 7900xt and 7900xtx then there was between the 6800xt and 6900xt.
Soooo yeah, what exactly did AMD do that was so bad or quote "much worse"? I hope you scrolled purely indeed for that availability info, I cant imagine anyone willing to risk their hard earned cash with Newegg again.....good luck never getting your multiple hundred dollar product and not being able to do anything about it.....
I checked Ebay under the "Sold" category and there are already 4080s that sold for from $1,600 to $1,900.
The leather jacket idiot clearly lost his mind...
- The SAM debarkle - making an unused part of an existing spec seem like their invention and originally gatekeeping support.
- Ryzen 5000 on pre 5 series chipsets debarkle and caving
- The 6500XT. What an insult.
- Zen 3 and 4 pricing for 6 core models
- 7900XT pricing to upsell the 7900XTX - this is horrendous, the 7900XT is more cut down relative to the 7900XTX than the 6800XT was to the 6900XT, yet the naming is off, it should be a 7800XT, and most people compare upwards, 20% slower for 10% less money, but it's also that it's almost a 40% price hike for the second top product. Yikes.
And that's all since Lisa Su took the reigns.Look I won't get drawn into an argument with you about it, because I'm sure from your perspective nvidia is heaps worse right? There must be great reasons AMD did those things, and more, and people have an intense desire to defend then. I'm coming from the perspective that they all do shitty, greedy things, pretty much solely in the desire of increased revenue and profits.
So I stand by my general approach, weight each products unique proposition against your needs, and buy accordingly. If I was to draw a moral line on who to buy a GPU from, I wouldn't but it from any of them. This one just seems especially daft, I'm sorry. They had to single handedly create this new wave of Upscaling and generate massive appetite for it both in competitors and consumers, before some people like to accuse them of holding the industry back, I reject that entirely.
Plus look at hubs coverage of xess, its plainly obvious why it only works on RTX cards, nobody would benefit, not nvidia, not the customers.
"making it seem like their invention"...yeah I dont get the issue there either, so what? 3 seconds later others can say its just resizable bar and again, COMPETITION CAN and did follow suit, its not like AMD blocked it from being used by anyone else.
"originall gatekeeping support", what exactly do you mean by that?
"Ryzen 5000 on pre 5 series chipsets debarkle and caving" You do understand why this was not originally a thing right? the fact that motherboards were never made with this much cpu support in mind? literally not having enough memory to store profiles for all different processors so you have to do some odd stuff to get that to work which can easily leave users with an unusable board unless they buy/borrow a compatible cpu real quick to fix it?
And yeah, consumers complaining and a company as a reaction complying with their wishes....what a horrible thing to do...thank god most companies just give consumers the middle finger....
6500XT is an insult but then the entire cryptocurrency/Covid19 period was a big insult, it shouldnt have been released, that is indeed a strike against them.
Zen 3 pricing being an issue is just nonsense, AMD sold their stuff super cheap in order to get some mindshare back from YEARS of Intel domination, when they finally convinced enough people (including the very companies they need to work with for example with motherboards) that they are worth a damn again, they are still not allowed to ask a bit more? they have to stay in this super low bracket even though their products beat the competition? come on now.
Zen 4 pricing, yes, that is too high, Intel is winning there and AMD should have answered with a slight price drop, heck I dont even get that AMD did not start with some 3DX models to keep/take the performance crown, just odd all around.
literally mentioned the 7900xt vs 7900xtx thing in the post you are replying to.
"from your perspective nvidia is heaps worse right?" yes and I backed that up with arguments.
On DLSS, they HAD to because they are heavily banking on selling the world the RTX brand and tech, and I dont mind that at all, what I mind is them then locking it exclusively for their hardware and shoveling buckets of cash to developers to PLS PLS PLS implement it, which they have to do because that is a lot of effort a developer has to go through purely to appease the Nvidia buyers, it wont help for the AMD buyers (ok thats barely 15% of the market if we look at steam) nor the console crowd (that is a much bigger market).
If they just made it open and asked simply to have mentioned in the settings DLSS by Nvidia, advertisment for them, gamers know where its from, heck it would even be optimised for their hardware and the competition just has to work to see if they can make it run decently on their hardware (same as XESS), then all would be fine, but nope, typical Nvidia, proprietory nonsense that cant last.
Gsync is dead because Freesync/Adaptivesync/VRR is here, PhysX is dead because we moved on to generic solutions that work for all, we cant have games in the future that have an entire page of the exact same tech just for every vendors version of it....that would just be stupid and man, think of the waste of human effort and energy for AMD and Intel to now have to make their own versions of something that already exists.......
So yeah, I dont think im daft, I think Nvidia is the one being daft and like I said, anti consumer and anti progress.
Hell if they just shared the tech, who knows what extra improvements maybe AMD and Intel could have brought to the table.
"Plus look at hubs coverage of xess, its plainly obvious why it only works on RTX cards, nobody would benefit, not nvidia, not the customers."
that last sentense I dont get, intel's xess only works on RTX cards? and nobody would benefit if that was not the case?
And yes mods, we are getting off topic, ill leave it for DM's after this.