AMD’s RDNA 3 GPUs possible that it isn’t as powerful as first thought


At the very least, the RX 7900 XT and Navi 32 may not be available.

image via hitech

Even AMD’s next-generation flagship graphics card, the Radeon R9 Vega 64, may not be as strong as originally thought, according to the latest report from the grapevine.

Greymon55, a well-known leaker, has revealed updated (alleged) core counts for AMD’s upcoming RDNA 3 graphics cards, according to the latest rumour.


So, AMD has lowered their target core counts for the Navi 31 (flagship) and Navi 32 GPUs, but Navi 33 has remained the same, goes the idea (Greymon55 has floated the number provided here for Navi 33 in the past). All hardware conjecture should be approached with a healthy dose of scepticism, of course.

In other words, how large of a dive have the core statistics taken? The anticipated 15,360 core count for Navi 31, which should be the RX 7900 XT, has been reduced to 12,288, whilst Navi 32 (RX 7800 or 7700) has gone from 10,240 to 8,192. The Navi 33 mid-range CPU is still expected to have 4,096 cores, as originally anticipated.

Is Nvidia currently in a better position to win the raw performance war than it was previously?
Since this new information has surfaced, it’s possible that the previous claim that the RDNA 3 flagship could provide 92 TFLOPS of raw performance is inaccurate, or that the GPU itself is hiding something important. Alternatively, this rumour might be completely false.


73 TFLOPS is what we’re looking at here, and that’s what we’ve heard in the past, since Navi 31 is expected to be clocked at 3GHz (in the past, 2.5GHz has been theorised, but remember that the incoming 6950 XT current-generation refresh is expected to boost above 2.5GHz already, right out of the box).

Although AMD’s next-generation GPUs may not have as many cores as we had hoped for, the overall mood of the signal indicates that we should proceed with caution in light of all the pre-release hoopla from the rumour mill.

But given what we’ve heard about Nvidia’s RTX 4000 models, this might be a cause for concern, since Lovelace could provide a greater performance jump than originally predicted, and it won’t just a simple Ampere upgrade; there will be additional modifications at the architecture level.


When it comes to raw performance and power consumption, it’s possible that Nvidia’s next-generation GPUs may outperform AMD’s RDNA 3 cards, but that might come at the expense of the PSU, which is likely to be significantly more efficient for Team Green’s next-generation cards than for AMD’s RDNA 3 cards.

It’s a worry for Lovelace that gamers may be forced to change their PSUs at the upper end because of power consumption, as well as the chance that AMD’s RX 7000 models may come available first, whilst Nvidia’s RTX 4090 flagship may be the first to be released (not the graphics card anyone with a normal-sized bank balance will be looking at).

If these next-generation GPUs are to be effective, a lot of various and intricate factors must be taken into account, not to mention the unknown influence Intel will have when desktop Arc graphics cards are released, presumably next month.


Leave a Comment