According to wccftech and noted hardware leaker @KittyYYuko, AMD’s next-generation Radeon graphics cards should be a massive upgrade — a whopping three times faster than AMD’s current flagship graphics cards.
2.5x is too little.
Guess what :p
— Elysian Empire (@KittyYYuko) May 2, 2021
In the above response to Paul Eccleston of RedGamingTech, @KittyYYuko claims a 3x performance boost when upgrading from AMD’s flagship Navi 21 (as found in current generation RX 6800XT and RX 6900 XT cards) to Navi 31.
RDNA3, the architecture under the new GPUs, will shift the size of the process nodes from 7nm to 5nm, typically leading to major performance gains and even greater performance gains per watt. However, the downshifting in the process node isn’t enough to explain the claimed performance improvement alone – it comes from making the GPU itself significantly bigger and nastier.
The current flagship, Navi 21, offers 80 compute units and 5,120 shading units. Navi 31 will reportedly double that number, with a whopping 160 CUs and over 10,000 shading units, delivered in two 80 CU chiplets on a single die. This is probably only practical because of the process size downshift — cooling and power are major concerns in flagship GPUs, which tend to require as much of both as OEMs and individual builders can comfortably cram into a case.
RedGamingTech suggests that Navi 31’s ray tracing performance should be roughly equivalent to Nvidia’s Ampere GPUs, but machine learning tasks will still be handled through low-precision operations on the main GPU itself rather than on dedicated hardware like Nvidia’s Tensor Cores.
We expect RDNA 3 maps to be available sometime in late 2022 or early 2023.