The AMD RDNA3+ architecture was emerging as an intergenerational successor to the current RDNA3, that is, as a revision with small improvements over it. Therefore, a great leap in terms of performance could not be expected, but the importance of this update it was not so much in the power but in the correction of some problems that the new generation of the Sunnyvale giant has been showing since its launch, and in an improvement in efficiency.
I have already told you on previous occasions that the RDNA3 architecture represents an interesting evolution compared to RDNA2. The most remarkable thing is, without a doubt, the performance improvement that it has achieved in ray tracing and the multi-chiplet design that it uses, although it has not met expectations in terms of raw power or efficiency, and this new generation architecture came accompanied of important errors that, in the end, have hindered its launch.
Some of those issues have been resolved with new drivers, but there are others that are present at the hardware level and can only be fixed. with a replacement of the affected units. On the other hand, we have also been seeing information pointing to problems with energy profiles and also to poor performance with meshed shaders, two issues that it would not be possible to resolve via drivers, and that could only be resolved with the release of a patch. of said architecture, that is, with the arrival of AMD RDNA3+.
The truth is that the launch of this new architecture made a lot of sense. On the one hand, RDNA3+ would allow AMD “settle accounts” and correct all those problems that we have been discussing, and on the other hand, this architecture would serve as the basis for the Sunnyvale company to launch a “refreshment” of its Radeon RX 7000 graphics cards between the end of this year and the beginning of next year. Those new graphics cards could arrive identified as Radeon RX 7X50 (Radeon RX 7950 XTX, for example), and they would compete with the renewal that NVIDIA will launch of the GeFoorce RTX 40.
The fact is that according to new information, coming from a very reliable source that already correctly alerted to the problems of the Radeon RX 7000, AMD would have canceled RDNA3+. Said source has been quite harsh in justifying this information, as he has commented that all the RDNA3 graphics cores failed to reach their performance and efficiency targets, and that in general terms it is “bad architecture”.
That’s why it wouldn’t make sense to launch a review, because it would be impossible to correct the deficiencies of RDNA3 with itand AMD would go on to focus directly on RDNA4, a completely new architecture that will allow the Sunnyvale company “Starting over”.
All RDNA3 refresh cancel.
All RDNA3 SKUs miss perform & efficient target.
RDNA3 bad all platform.
Jensen won.
—All The Watts!! (@All_The_Watts) January 22, 2023
The AMD RDNA3+ entry will be discarded in favor of RDNA 4, we explain why it was published first in MuyComputer.
