It is not the first time that AMD has used the technique of «refried»In its catalog of graphics cards, and in fact it is something that users already got used to in the past with this company. For those who are not used to the term, we will tell you that it implies doing precisely what we have just mentioned: introducing an old “improved” architecture in a new generation of GPU that debuts a different architecture.
Only “top” AMD GPUs will be RDNA 3
According to the source of this rumor (quite reliable if we stick to the precedents), AMD’s upcoming Radeon RX 7000 series could include both RDNA 2 and RDNA 3 architecture GPUs. This theory holds that the company could introduce new 5nm GPUs. based on the new RDNA 3 architecture for the highest end of the generation (Navi 31 and Navi 32), while giving the current RDNA 2 architecture a new life in the lower manufacturer segments.
However, this does not seem like a simple rebrand. Apparently some existing Navi 2X series chips would receive a ‘optical shrinkage” in the 6 nm node, in an attempt to improve its performance-per-watt power consumed ratio. Some of the performance / consumption improvements could, of course, be used to increase operating speeds and thus performance (in this case while maintaining consumption).
These GPUs include the Navi 22, with its 40 RDNA 2 compute units and its 192-bit GDDR6 memory bus, as well as Navi 23, with its 32 RDNA 2 compute units and its 128 bits for the GDDR6 memory bus. The updated Navi 22 chip will power the SKU that will succeed the current RX 6600 XT, while the Navi 23 chip update will work for the lower segment RX X500 range.
If this is true, it is also likely that AMD will take the opportunity to update the video and graphics acceleration components, just as it did with the latest iterations of the Vega series. Meanwhile, RDNA 3 will likely power only the two largest chips of the generation, the Navi 31 and the Navi 32, both at 5 nm, and that will be the heart of the models that will succeed the current Radeon RX 6900/6800 of the firm.
AMD returns to the load with “refried”, are they running out of options?
We have already told you before that it is not the first time -or surely it will be the last- that AMD resorts to refried to fill models of its generations of graphics cards. This is a perfectly valid technique as long as they adequately indicate which architecture equips each graphic (after all, they are simply releasing two architectures and two different lithographs within the new generation of GPU instead of all of it being based on a new one. architecture as expected), but denoting some weakness in the ranks of those in red.
And, after all, when you launch a new generation of graphics cards, what users expect is an increase in performance but also all the benefits of a completely new graphics architecture. If only AMD’s next-gen high-end models will feature RDNA 3 architecture (and will obviously be significantly more expensive than the other models), it can be quite disappointing as it removes the new architecture from the possibilities of many users.
This refried technique also denotes, as we said, some weakness in the brand’s engineering group; As we said, the ideal is to launch a new generation with a new architecture that ranges from entry-level models to the highest-end ones, and failure to do so may mean that they have not BEEN ABLE to do so, be it due to development, cost or time.