• 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.

Discussion Ada/'Lovelace'? Next gen Nvidia gaming architecture speculation

Page 37 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Die sizes haven't been confirmed, but going off of rumored sizes we now have:
- AD102, ~600mm2 = $1599 ---> $2.67 $/mm2
- AD103, ~380mm2 = $1199 ---> $3.16 $/mm2
- AD104, ~300mm2 = $899 ---> $3.00 $/mm2

It's obvious that RTX 4080 in all of its forms is priced at least $100 too high. The issue is they don't want to kill off Ampere sales entirely... and that's where RTX 4080 lives. I fully expect AMD to take Nvidia head-on in the $600 - $1200 price range, with Nvidia being 100% capable of lowing the price to compete (ideally, for them, only after Ampere clears inventory).
 
Saylick said:
Yeah, but that would erode the price of Ampere. They needed to build out a pocket where the remaining inventory could sell out at the highest price possible... This behavior is transparent AF.
Click to expand...
Not to mention that Samsung 8nm has huge capacity and is still producing cards. So why not even keep them in production as Samsung sells its 8nm wafers at a massive discount versus TSMC 4N. TSMC 4N is likely capacity constrained so this feels like production and fab capacity optimization versus inventory problems.

Ampere "inventory" NVIDIA has already sold to board partners and other 3rd parties, so they wouldn't keep them in the lineup only to help others sell inventory. Just as they introduced a new 2xxx series in 2021 (on old TSMC 12nm) they are basically reintroducing 3xxx series and will keep these in production long enough that 4N becomes cheaper and they can start to build lower end 4xxx on 4N.
 
Maybe my reading comprehension fell off a truck but when are we going to get final specs. So to clarify we are getting a PCIe gen 4 card, that will require a new 3.0 PSU for required power cable, and most likely have beta drivers that need to be tweaked for DLSS.

If AMD brings out a Gen5 card in November they will be a better option going forward. If your building a new system using AM5 do you really want a Gen 4 card, especially if a serious hobbyist.

R81Z3N1
 
R81Z3N1 said:
Maybe my reading comprehension fell off a truck but when are we going to get final specs. So to clarify we are getting a PCIe gen 4 card, that will require a new 3.0 PSU for required power cable, and most likely have beta drivers that need to be tweaked for DLSS.

If AMD brings out a Gen5 card in November they will be a better option going forward. If your building a new system using AM5 do you really want a Gen 4 card, especially if a serious hobbyist.

R81Z3N1
Click to expand...
When has the PCIe generation been the limiting factor in a x16 card?
 
swilli89 said:
What the hell is this presentation? It feels like an investor slide deck. Literally 4 minutes spent on new video cards as they pertain to video games and its been over 50 minutes on robotics/datacenter/car computers and other abstract things. Am I missing something or am I crazy?
Click to expand...
NVIDIA GTC (GPU Technology Conference) is a global AI conference for developers that brings together developers, engineers, researchers, inventors, and IT professionals. So not really a gaming focus.
 
NV has left a hole here large enough for AMD to drive a dump truck through. A competitive N31 @ $1200 would likely bring in more cash than AMD's wildest dreams while still seriously undercutting the NV halo price in a huge way.

NV's "everything but gaming" heavy focus in this presentation felt like a shockingly transparent display of no confidence in their top end product to actually compete in traditional gaming scenarios.
 
tamz_msc said:
RTX Remix enabled modding is the thing I'm most excited about. This looks really promising.

www.nvidia.com

NVIDIA GeForce News

Introducing NVIDIA RTX Remix<br/>#RTXON
www.nvidia.com
Click to expand...
Leave it to nvidia to make dishonest comparison by making "off" pictures so dark you can't see s**t. I was excited about it too, but I literally can't see supposed benefits because originals are so dark.

Anyway. disappointing, but not unexpected. This feels like Turing 2.0. Higher power consumption, higher prices just because, relabeling 4070 into 4080, selling midrange 4070 chip for $900 with only 12GB of RAM. This is the kind of stuff that apple pulls. Will definitely be waiting till RDNA3 reveal. Will have to see how things shake out, what RDNA3 looks like, what used prices are going to look like after miners start dumping their cards. As someone has said, I just may have to give up being an enthusiast gamer if this is what future looks like - I ain't paying $900 for 12GB "upper midrange" card. Good thing I have plenty other hobbies other than gaming. If RDNA3 isn't priced better looks like my best choice would be to pick up used 6000 series card on the cheap after miner selloff. Time will tell. But right now I'm with Linus Torvalds when it comes to nvidia.
 
Ahahahaha holy hell what a mess

"up to 2 to 3 times faster!"
What the hell does that even mean!?!
Their presentation is littered with marketing doublespeak, extreme cherry-picking and turd glitter (DLSS3).

nVidia, it's still a turd.

RIP hardcore PC gaming, it'll only go down hill from here. Unless AMD serves up a silver bullet or two into nVidia HQ with RDNA3. As they infrequently do.
 
Last edited:
So wait, after all this time we have a 4080 12GB at $899 MSRP that just ties the 3090 Ti with DLSS3.0 off? Am I reading that correctly?
The same 3090 Ti that's only like 10% faster at 4k than a 12GB 3080 Ti, a card currently selling for under $750?

Woof.
 
fleshconsumed said:
It is. I'm not disputing it. I'm just saying it's not an apples to apples comparison when one picture is pitch black and the other one is bright as day.
Click to expand...
It cannot be an apples to apples comparison. It is almost like a modded remake of the assets in the original game, using tools that NVIDIA has developed.
 
biostud said:
When has the PCIe generation been the limiting factor in a x16 card?
Click to expand...

I agree I run a pcie-4 card in a gen 3 motherboard. But if your spending that kind of money you would expect the best. Not to mention new specs like Display Port 2.0 looks like to me the safe money is to wait.

Pick up a 3 series at a supper discount use FSR to upscale if needed and wait it out.

R81Z3N1
 
Kaluan said:
Ahahahaha holy hell what a mess

"up to 2 to 3 times faster!"
What the hell does that even mean!?!
Their presentation is littered with marketing doublespeak, extreme cherry-picking and turd glitter (DLSS3).

nVidia, it's still a turd.

RIP hardcore PC gaming, it'll only go down hill from here. Unless AMD serves up a silver bullet or two into nVidia HQ with RDNA3. As they infrequently do.
Click to expand...
05f6e5ba8aa901aebc366db3745f6484e574596be01fc7f6e4687b32b1629fae.jpg
 
MrTeal said:
So wait, after all this time we have a 4080 12GB at $899 MSRP that just ties the 3090 Ti with DLSS3.0 off? Am I reading that correctly?
The same 3090 Ti that's only like 10% faster at 4k than a 12GB 3080 Ti, a card currently selling for under $750?
Click to expand...

It'd probably be better at lower resolutions but it does look like there's a serious memory bottleneck.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top