For many weeks now, there have been a lot of crazy rumors about the next Nvidia RTX 4000 graphics cards. Without confirming it, we think it’s important to point out that one of them has been coming back with insistence for some time. The next generation of RTX cards would only be an update of the engraving process and an optimization of the architecture. The older generation will remember that we experienced something similar when we switched from Maxwell in 28nm to Pascal in 16nm.
RTX 4000 ( Ada Lovelace ) : a simple evolution of Ampere ?
These speculations are reinforced when we take into consideration the track record of Nvidia’s move to Samsung Foundry. This choice, dictated by cost reasons, has clearly proved disastrous : performance, consumption, temperatures and efficiency have not been up to expectations. This rumor is not new since it was already mentioned by other credible leakers in 2021. So obviously if this information is confirmed, it does not mean that the next cards will stagnate in terms of performance. We have all only been able to catch a glimpse of the potential of Ampere and in particular its rise in frequency. On the other hand, these cards will undoubtedly face the more advanced designs of the future Intel and AMD.
