The current lineup doesn’t even use GDDR6X and it’s selling like shit. There’s just no way.

Maybe, but it would be unnecessary because RDNA4 caps out at midrange so GDDR6 would suffice. AMD decided to cut RDNA4 high-end to produce more AI chips and earn more money, and give their engineers more time to get multiple GPU chiplets on 1 card working for RDNA5, which is where the real performance boom is at. RDNA3 still has 1 graphics die, only the memory controller and cache is on separate chiplets.

So the 7900XTX will remain AMD’s flagship until RDNA5.

There might be some kind of refresh of the 7900XT(X) with slightly better performance and efficiency, maybe those would use GDDR7 if possible and economical.

The good news for current owners is the 7900 cards have plenty of VRAM to last until RDNA5. The bad news is there will be no competition fir the 5080 and 5090 so expect even higher MSRPs than the 4000 series. $2500 MSRP for a 32GB 5090 wouldn’t surprise me. And $1500 for the 5080, the “gaming flagship”.

If you were waiting for next gen hoping value would improve vs the 4000 series… I hope you have even more patience.

The moment I heard the news about RDNA4 high-end being scrapped and the monster chiplet design was moved to RDNA5, as well as the high AI demand and lack of production capacity I pulled the trigger on a 7900XT because next gen is going to be absolutely bonkers on the Nvidia side and nothing better will be released on the AMD side other than software improvements, maybe a refresh of the 7900 cards but that’s it. This card with 20GB VRAM will last me until RDNA5/RTX6000.

Jayz2cents also made a video a while ago voicing his opinion to buy a GPU now cause it’s only gonna get worse in the coming years. A situation arguably worse than the crypto boom, combined with a lack of competition for 80 and 90 series and Nvidia’s Apple approach… Bad news.

Intel won’t have a truly viable product for general gaming within this timeframe either. Even today their drivers are lightyears behind both AMD and Nvidia with performance all over the place depending on each individual game. And Intel too is making AI chips based on GPUs. The consumer GPUs are like a proof of concept.

or productivity users, there’s… A6000.

A6000 is a lot of money. For productivity users for say, Blender, you can get the same 48GB of VRAM and more compute for a lower cost if you go with dual 4090’s.

GDDR7 memory chips will be in production with either 2 or 3 GB sizes, which means 36GB of VRAM on 384-bit bus could be a possibility for next gen.

It’s because AD102 is already a huge monolithic die, a little over 600mm2, with 814mm2 being the theoretical limit. In short: the bigger the die size the lower the yields, there will never be a gaming GPU much bigger than 600mm2 because then you’re looking at terrifying prices.

A node shrink only helps so much, and you don’t want a GPU sucking 800 watts either. The wider memory bus also takes up extra space. The 5090 is still monolithic so a ~30% improvement in performance sounds plausible.

Even worse, AMD (underatandibly) is “skipping” RDNA4 high-end to both maximize AI production and give their engineers more time to get a chiplet design with multiple graphics chiplets working well.

RDNA5 will likely be high end again, but until then, the 7900XTX or a refresh of it will likely remain the fastest AMD card.

Which means next gen Nvidia pricing will go through the roof. We’re probably looking at a $999 16-20GB 5070(Ti) that matches AMD flagship performance so I wouldn’t be surprised if a 24GB 5080 will be priced at $1500-1750 MSRP as the gaming flagship and the hybrid 32GB 5090 $2500-3000 MSRP. Remember the 3090Ti launched at $2000 despite gaming competition from the $1000 6950XT.

… And people will buy them. RIP GPU prices for the next 3 years, at least.

I would bet on not a significant upgrade as well. Performance per transistor in Ada vs Ampere actually went down, mostly because they spend so much transistor budget on cache. 170% more transistors for 65% more performance, if you assume the full potential AD102 die is 15% faster than a 4090. But even before AMD and Nvidia started playing with massive caches, you could always relatively accurately predict how fast a GPU would be based on transistor count.

Cache is not shrinking in die area with new nodes. If this thing has 128MB, and the rest of the die stays at 600mm^(2), the area dedicated to logic would be smaller than AD102. Unless they start stacking cache that is. Would not shock me, if most, or all the L2 is on a 2nd layer.

I’ve heard 60% more logic density going from 5nm to 3nm, but who knows how optimistic those numbers are as they are probably best case scenario. Can’t but imagine a real GPU application would at maximum reach 40-50% more logic density, and I can’t imagine a performance uplift higher than that, unless they make it a 800mm^(2) die which I don’t believe.

GA102 to AD102 increased by about 80%, but the jump from Ad102 to GB202 is only slightly above 30%,

Maybe GB202 is not the top chip, and the top chip is named GB200.

I mean, you’d expect this die to be called GB102 based on the recent numbering scheme, right? Why jump to 202 right out of the gate? They haven’t done that in the past, AD100 is the compute die and AD102, 103, 104… are the gaming dies. In fact this has been extremely consistent all the way back to Pascal, even when there is a compute uarch variant that is different (and, GP100 is quite different from GP102 etc) it’s still called the 100.

But if there is another die above it, you’d call it GB100 (like Maxwell GM200, or Fermi GF100). Which is obviously already taken, GB100 is the compute die. So you bump the whole numbering series to 200, meaning the top gaming die is GB200.

There is also precedent for calling the biggest gaming die the x110, like GK110 or the Fermi GF110 (in the 500 series). But they haven’t done that in a long time, since Kepler. Probably because it ruins the “bigger number = smaller die” rule of thumb.

Of course it’s possible the 512b rumor was bullshit, or this one is bullshit. But it’s certainly an odd flavor of bullshit - if you were making something up, wouldn’t you make up something that made sense? Odd details like that potentially lend it credibility, because you’d call it GB102 if you were making it up. It will also be easy to corroborate across future rumors, if nobody ever mentions GB200-series chips again, then this was probably just bullshit, and vice versa. Just like Angstronomics and the RDNA3 leak, once he’d nailed the first product the N32/N33 information was highly credible.

GA102 to AD102 increased by about 80%

without scaling DRAM bandwidth anywhere near as much, only partially compensating for that with a much bigger L2.

For 5090 on the other hand we might also have clock increase going (another 1.15x?), and proportional 1:1 (unlike Ampere -> Ada) DRAM bandwidth increase by a factor of 1.5 due to GDDR7 (no bus width increases necessary; 1.5 = 1.3 * 1.15), so this is 1.5x perf increase 4090 -> 5090, which has to be further multiplied by whatever u-architectural improvements might bring, like Qesa is saying.

Unlike Qesa, though, I’m personally not very optimistic regarding those u-architectural improvements being very major. To get from 1.5x that comes out of node speed increase and the node shrink subdued and downscaled by node cost increase, to recently rumored 1.7x one would need to get (1.7 / 1.5 = 1.13) 13% perf and perf/w improvement, which sounds just about realistic. I’m betting it’ll be even a little bit less, yielding more like 1.6x proper average, that 1.7x might have been the result of measuring very few apps or outright “up to 1.7x” with “up to” getting lost during the leak (if there was even a leak).

1.6x is absolutely huge, and no wonder nobody’s increasing the bus width: it’s unnecessary for yielding a great product and even more expensive now than it was on 5nm (DRAM controllers almost don’t shrink and are big).

You should be looking at transistor amount if anything at all, “cuda cores” is only somewhat useful when looking at different products within the same generation.

Its expected to be like Ampere, Ampere was 17% increase in SMs (rtx 3090ti vs rtx Titan) but the SM itself was improved such that they yielded about 33% improvement per SM in ‘raster’ and massive improvements in occupency for RT workloads. So 3090ti ended up 46% faster in ‘raster’ vs rtx Titan.

The TPC and GPC of Blackwell are rumored to be overhauled with a more hesitant rumor about the SM also being improved.

It’s highly likely to be a major architecture update, so core count alone won’t be a good indicator of performance.