Sunday, May 24th 2020
AMD RDNA2 "Navi 21" GPU to Double CU Count Over "Navi 10"
AMD's RDNA2 graphics architecture, which sees real-time ray-tracing among other DirectX 12 Ultimate features, could see the company double the amount of stream processors generation-over-generation, according to a specs leak by _rogame. The increase in stream processors would fall in line with AMD's effort to increase performance/Watt by 50%. It may appear like the resulting SKUs finally measure up to the likes of the RTX 2080 Ti, but AMD has GeForce "Ampere" in its competitive calculus, and should the recent specs reveal hold up, the new "Navi 21" could end up being a performance-segment competitor to GeForce graphics cards based on the "GA104" ("TU104" successor), rather than a flagship-killer.
The RDNA2-based "Navi 21" GPU allegedly features 80 RDNA2 compute units amounting to 5,120 stream processors. AMD might tap into a refined 7 nm-class silicon fabrication node by TSMC to build these chips, either N7P or N7+. The die-size could measure up to 505 mm², and AMD could aim for a 50% performance/Watt gain over the "Navi 10." AMD could carve out as many as 10 SKUs out of the "Navi 21," but only three are relevant to the gamers. The SKU with the PCI device ID "0x731F: D1" succeeds the RX 5700 XT. The one bearing "0x731F: D3" succeeds the RX 5700, with a variant name "Navi 21 XL." The "Navi 21 XE" variant has a PCI ID of "0x731F: DF," and succeeds the RX 5600 XT.
Among the other variants are the "Navi 21 XTX," with PCI ID "0x731F: D0," which could be a limited edition SKU succeeding the slightly beefed up RX 5700 XT 50th Anniversary Edition; the Navi 21 Pro XT and Navi 21 Pro XL with PCI IDs "0x731F:10" and "0x731F:12," marking Radeon Pro W5700X and W5700 successors, respectively. There are also four Apple-exclusive SKUs, the "Navi 21 XTA" and "Navi 21 XLA" client-segment chips targeting next-generation iMac and iMac Pro desktops, and their Pro variants targeting future Mac Pro workstations.
The first-gen RDNA may not ride into the sunset, as AMD is planning to refresh them. The company is probably porting the "Navi 10" silicon to TSMC N7P, to come up with new mainstream SKUs that lack ray-tracing, but will occupy mid-range price-points. This would be similar to NVIDIA positioning half its "Turing" product-stack without ray-tracing, under the GTX 16-series, occupying sub-$300 price-points. The resulting ASICs are the "Navi 10+," "Navi 10 XM+," and "Navi 10 XTE+."
Source:
Hardware Leaks (by _rogame)
The RDNA2-based "Navi 21" GPU allegedly features 80 RDNA2 compute units amounting to 5,120 stream processors. AMD might tap into a refined 7 nm-class silicon fabrication node by TSMC to build these chips, either N7P or N7+. The die-size could measure up to 505 mm², and AMD could aim for a 50% performance/Watt gain over the "Navi 10." AMD could carve out as many as 10 SKUs out of the "Navi 21," but only three are relevant to the gamers. The SKU with the PCI device ID "0x731F: D1" succeeds the RX 5700 XT. The one bearing "0x731F: D3" succeeds the RX 5700, with a variant name "Navi 21 XL." The "Navi 21 XE" variant has a PCI ID of "0x731F: DF," and succeeds the RX 5600 XT.
The first-gen RDNA may not ride into the sunset, as AMD is planning to refresh them. The company is probably porting the "Navi 10" silicon to TSMC N7P, to come up with new mainstream SKUs that lack ray-tracing, but will occupy mid-range price-points. This would be similar to NVIDIA positioning half its "Turing" product-stack without ray-tracing, under the GTX 16-series, occupying sub-$300 price-points. The resulting ASICs are the "Navi 10+," "Navi 10 XM+," and "Navi 10 XTE+."
51 Comments on AMD RDNA2 "Navi 21" GPU to Double CU Count Over "Navi 10"
Maybe this could be a good replacement, provided it is available in most common fonts, but whatever font we are using as default for forum posts shows it fine. It looks identical to me and doesn't make emoticons.
꞉ ꞉D
: :D
That said, TPU's main site doesn't show emoticons, so it's just a forum problem (or unexpected result)
I would expect higher prices that the 5700Xt special for the higher one's.
They also "could" this and "could" that. Maybe they "could" and "could" that as well. There is nothing of value in these reprinted speculations that have nothing concrete to back them up.
Come on, Lisa and Huang, make something that will be worth my attention! Not the energy sucking 5700XT or the gimmicky RT of Nvidia's current GPUs. Shall I give my next GPU money to Intel???
I mean why not do 4 times the streaming processors while we are at it?
Is there any reason why we stop at double? and technical reason? or is it just financial, not giving people all they could have right ffing now?
Like I just dont get that if just more streaming processors is so important, why doesnt AMD just make a 2000 dollar gpu with all the streaming processors in the world and take the performance crown?
4 times the shaders would mean 40 256-shader compute units or 80 128-shader compute units which is not possible on this process node since the die would become prohibitively large.Except than Navi 21 will be at least 100% faster than your GTX 1080.
Maybe I'll manage to hold for a couple of years more, maybe not. :D
They made so many mistakes during the years. Fury X with 4 GB, Vega 64 with compute oriented architecture, RX 5700 XT that competes in the mid-range segment.
Left CrossFire support, have problems with RX 5700 XT's driver stability....
Do you have problems with your RX 5700XT's and AMD driver stability? Because I don't have any problems with my 5600XT nor AMD drivers stability so I'm curious.