Monday, December 28th 2020
NVIDIA's Next-Gen Big GPU AD102 Features 18,432 Shaders
The rumor mill has begun grinding with details about NVIDIA's next-gen graphics processors based on the "Lovelace" architecture, with Kopite7kimi (a reliable source with NVIDIA leaks) predicting a 71% increase in shader units for the "AD102" GPU that succeeds the "GA102," with 12 GPCs holding 6 TPCs (12 SMs), each. 3DCenter.org extrapolates on this to predict a CUDA core count of 18.432 spread across 144 streaming multiprocessors, which at a theoretical 1.80 GHz core clock could put out an FP32 compute throughput of around 66 TFLOP/s.
The timing of this leak is interesting, as it's only 3 months into the market cycle of "Ampere." NVIDIA appears unsettled with AMD RDNA2 being competitive with "Ampere" at the enthusiast segment, and is probably bringing in its successor, "Lovelace" (after Ada Lovelace), out sooner than expected. Its previous generation "Turing" architecture saw market presence for close to two years. "Lovelace" could leverage the 5 nm silicon fabrication process and its significantly higher transistor density, to step up performance.
Sources:
Kopite7kimi (Twitter), 3DCenter.org (Twitter), VideoCardz
The timing of this leak is interesting, as it's only 3 months into the market cycle of "Ampere." NVIDIA appears unsettled with AMD RDNA2 being competitive with "Ampere" at the enthusiast segment, and is probably bringing in its successor, "Lovelace" (after Ada Lovelace), out sooner than expected. Its previous generation "Turing" architecture saw market presence for close to two years. "Lovelace" could leverage the 5 nm silicon fabrication process and its significantly higher transistor density, to step up performance.
43 Comments on NVIDIA's Next-Gen Big GPU AD102 Features 18,432 Shaders
Yeah - lets have all the board partners dump the 3xxx line while they are selling every single card and the demand hasn't slowed down one bit. Then force AIB to go spend money on designing new cards and more marketing.
This is WCFtech level stuff article.
AmpereVolta and Turing were.Stop acting like only gamers buy these chips.
So far with the 2000 series and 3000 series out, my 2x 1070's in SLI work better then a single card in 2160P which is what I game at.
I could see 8k a possibility with two MCM video cards in SLI to run 4320P maybe. But it's highly unlikely that Nvidia or AMD want us gamers to have such nice things.
There main focus for the series they have now is 1440P which to me is a step backwards.
Nvidia aren't stupid, they know these 3000 series cards are selling like hotcakes, issue is simply getting supply out.
Day of the Innocent Saints (rough translation)Holy Innocents' Day, which is similar to April Fools.Also, Nvidia isn't going to TSMC, they are staying with samsung, so unless Samsung improves their node they are still going to be limited by the process more than architecture again, and have a even more power hungry chip, that if larger will mean more defect dies, and thus higher priced.... Competition is glorious, AMD is held back by wafer supply, Nvidia is held back by wafer supply, and we the consumers have to pay to play.
Reviewer and people claiming to own one of those cards was actually hypnotized!
1) This article ( or Kopite7kimi twitt for that matter ) give 0 information when this transition might happen , which automaticaly makes your rant absurd .
2) It depends what '' anytime soon '' means to you which hopefully you do understand is not an universal metric . Replacing Ampere early 2021 seems indeed highly unlikely but replacing Ampere late 2021 or early 2022 is very very likely . If anything else Nvidia CEO himself has declared that Nvidia is targeting a yearly release schedule . So yeah it all depends what ''anytime soon'' means to you , but if you expect Ampere to last more than a year or a year and a half tops , then you are delusional , especially given Nvidia is finaly facing some kind of concurence from AMD ( as far as rasterized perf goes ) .
And 5nm Lovelace won't be out soon, there will be Ampere refreshes long before that. 12GB 3060 and 20GB 3080ti will be out ca. q1 2021 and it won't take long from there to have full refresh. Be it 3000 Super's or just sku's with the ti moniker on them.
Looks like something that surely won't happen in 2021, and 2022 is in question, too - yeah, they'll said early 2022 (or something) and it will become something like 'summer 2020 for Ampere'.
Also like rumours that make me think that I should wait RDNA3, because why-not, it's round the corner...
Both companies made good products - well, for people investing ludicrous amounts of money, still waiting for something that will interest me, to the point of buying it - no way they'll miss the opportunity to exploit it for at least a year... Surely, they are developing new stuff and that article can as well be correct, but the dates aren't close at all. There was huge investment in Ampere and RDNA2, no way anyone will give it up and say "no, 'Linda' is replacing the whole line, see ya" (it's AMDs fault they don't have easily mockable architecture name, like in Duron days :) )
Sure by consumer's pov current market sucks big time. But it's more because of absent for true new gen mid-range cards rather than pricier high end(I don't count 3060ti as mid-range). Ancient Polaris and lack luster 16 -series Turings just wont cut it anymore. Sure one can have good deals from 1-gen navis, but those aren't really mid-range either and yeah i think they are EOLled anyway to get room for other 7nm products. So what market really needs is a new $200-$250 mid-range champ, perfect 1080p and capable 1440p card as rx 480 or gtx 1060 were at the beginning.