Monday, January 29th 2024
Top AMD RDNA4 Part Could Offer RX 7900 XTX Performance at Half its Price and Lower Power
We've known since way back in August 2023, that AMD is rumored to be retreating from the enthusiast graphics segment with its next-generation RDNA 4 graphics architecture, which means that we likely won't see successors to the RX 7900 series squaring off against the upper end of NVIDIA's fastest GeForce RTX "Blackwell" series. What we'll get instead is a product stack closely resembling that of the RX 5000 series RDNA, with its top part providing a highly competitive price-performance mix around the $400-mark. A more recent report by Moore's Law is Dead sheds more light on this part.
Apparently, the top Radeon RX SKU based on the next-gen RDNA4 graphics architecture will offer performance comparable to that of the current RX 7900 XTX, but at less than half its price (around the $400 mark). It is also expected to achieve this performance target using a smaller, simpler silicon, with significantly lower board cost, leading up to its price. What's more, there could be energy efficiency gains made from the switch to a newer 4 nm-class foundry node and the RDNA4 architecture itself; which could achieve its performance target using fewer numbers of compute units than the RX 7900 XTX with its 96.When it came out, the RX 5700 XT offered an interesting performance proposition, beating the RTX 2070, and forcing NVIDIA to refresh its product stack with the RTX 20-series SUPER, and the resulting RTX 2070 SUPER. Things could go down slightly differently with RDNA4. Back in 2019, ray tracing was a novelty, and AMD could surprise NVIDIA in the performance segment even without it. There is no such advantage now, ray tracing is relevant; and so AMD could count on timing its launch before the Q4-2024 debut of the RTX 50-series "Blackwell."
Sources:
Moore's Law is Dead (YouTube), Tweaktown
Apparently, the top Radeon RX SKU based on the next-gen RDNA4 graphics architecture will offer performance comparable to that of the current RX 7900 XTX, but at less than half its price (around the $400 mark). It is also expected to achieve this performance target using a smaller, simpler silicon, with significantly lower board cost, leading up to its price. What's more, there could be energy efficiency gains made from the switch to a newer 4 nm-class foundry node and the RDNA4 architecture itself; which could achieve its performance target using fewer numbers of compute units than the RX 7900 XTX with its 96.When it came out, the RX 5700 XT offered an interesting performance proposition, beating the RTX 2070, and forcing NVIDIA to refresh its product stack with the RTX 20-series SUPER, and the resulting RTX 2070 SUPER. Things could go down slightly differently with RDNA4. Back in 2019, ray tracing was a novelty, and AMD could surprise NVIDIA in the performance segment even without it. There is no such advantage now, ray tracing is relevant; and so AMD could count on timing its launch before the Q4-2024 debut of the RTX 50-series "Blackwell."
396 Comments on Top AMD RDNA4 Part Could Offer RX 7900 XTX Performance at Half its Price and Lower Power
"The world is wrong and Im right"... lmao
I mean raster performance, yeah but at 450~500? I can't see it going under 600 at best if not more.
It will depend a lot on RTX5000 perf/prices and how its going to compete to them (raster/RTRT).
------------------------------------------- I dont know guys...
Seems like this one is on a constant campaign of shoveling dirt and mud all over AMD.
Almost every post on every thread is on the same direction. Twisting facts, cropping screenshots and many more just to prove how bad AMD is on everything.
if graphics processors could be 'stacked' with >50% performance gains for each GPU added,
I'd probably rock multiple G80s or G92s, for the lulz.
I'd love to see some real representative reviews when both camps have launched their cards and shows where their new products stand in the picking order as well as pricewise. Would give me some food for thoughts on the possible upgrade path.
Truth be told, if Nvidia decided to release a complete AD102 chip today (with all 144 SMs and 96 MB of cache enabled) and upgrade it with the 24 Gbps G6X chips used on 4080 and 4080S, with board designs simply reusing parts from existing RTX 4090 assembly lines, they'd achieve something that could be considered equivalent to a full generational uplift by simply releasing something that isn't cutdown. Yet there is no reason, or incentive to do even that, regardless of branding or positioning. They even played with the idea around, there are reports of "Titan Ada" engineering samples having been built, they just never released that (at 48 GB config) or 4090 Ti (24 GB config).
But I was just wondering when the new lower tier products were coming out with more convinient prices to the average Joe, as Q4-2024 seemed to be the starting season of new products. Does make sense from a marketing point of view, and these companies are always keen on making the most out of these months. Also curious if AMD catches up with RT and ARC closing in on the competition, would be nice (maybe just whisful thinking) ;) .
I usually bring popcorn. I suspect both AMD and Nvidia will stagger their launches from the top down, for profitability reasons. Sell the highest margin SKUs to those with fat wallets and the least patience first, while shifting old inventory at more affordable tiers before updating SKUs lower down the product stack.
For people looking for midrange GPU upgrade, I think were going to see very little of value from AMD until after Christmas and Nvidia's 5060Ti or whatever they call it will probably be Summer 2025. I'd like to be wrong, but that just seems to be the way the last few generations have played out.
Maybe the new Navi 48 and Navi 44 powered graphics cards will be called something else?!
What about Radeon AI 170 Ultra?
The L3 caches will remain rather small - 48 or 64 MB...
videocardz.com/newz/amd-rdna4-gpus-to-feature-48-to-64mb-of-infinity-cache
The 8800XT (Navi 48) matches the approximate shader count, exact memory size, bus width, and cache size of the 7800XT so the cache size isn't really surprising or unexpected.
I think it's quite obvious that RDNA 3 would have been faster with more L3.
128 MB is needed for Navi 48.
RDNA2 had more L3, RDNA4 has the same L3 as RDNA3 and RDNA3.5
More L3 is always better but it's always about balancing cost vs performance. If it adds 15% to the cost and only brings 10% more performance, then it's not worth it.
What's the cost difference in % between RX 7900 XTX and RTX 4090? One is 930 bucks, the other 1740 bucks.
Those missing % in performance would have made the whole lineup more interesting.
What was in the rumors earlier, about AMD not competing in the high-end consumer graphic cards segment in the near future is now confirmed, but with some nuances why this decision was made.
www.tomshardware.com/pc-components/gpus/amd-deprioritizing-flagship-gaming-gpus-jack-hyunh-talks-new-strategy-for-gaming-market
This combined with the newsflash about a 'change' of architecture ( www.techpowerup.com/326442/amd-to-unify-gaming-rdna-and-data-center-cdna-into-udna-singular-gpu-architecture-similar-to-nvidias-cuda ) makes me think not all hope is lost, but it will take time.
But with lot's of it depending on implementation by software developers, so still quite a big uncertainty factor.