The leaked schedule includes a possible launch date for the NVIDIA GeForce RTX 50, the generation of graphics cards that will replace the GeForce RTX 40 and will theoretically hit the market. sometime in 2025. If you look closely at the image, you will see that before this new generation comes for mainstream consumption, the successor to Hopper, the architecture used by NVIDIA in the professional sector, will be launched.

Hopper’s successor will be Blackwell, but what architecture will succeed Ada Lovelace? Well, unfortunately we still don’t have a name, so we can only know her as “Ada Lovelace Next” at this point. However, what I am quite clear about is what we can expect from NVIDIA GeForce RTX 50and I’ll share a preview with you.

The GeForce RTX 50 will repeat the strategy that NVIDIA has maintained since the launch of the GeForce RTX 20, that is, they will already have classic divided into shaders, RT cores, tensor cores and “Optical Flow Accelerator”. The latter is responsible for the acceleration applied to the estimation of moving pixels, necessary for image generation (DLSS 3).

I’m pretty sure NVIDIA it will keep the monolithic GPU core because it finally managed to scale really well with the 5nm node, both in shader density and power per watt consumed, and because it works with a very mature core architecture that the giant has full control over. I also think it’s quite likely that the GeForce RTX 50 will come with a substantial increase in the amount of graphics memory.

However, I have my doubts about the next move NVIDIA will make in using AI to improve gaming performance, although seeing the pace of development the green giant is pursuing, I’m inclined to think that for now the company will focus on improving frame generation, a strategy he was already pursuing at the time with DLSS Super Resolution, which, as our readers remember, received a number of improvements and updates.

Why would NVIDIA release GeForce RTX 50 in 2025?

It’s a simple question to answer, leave room for the market finish digesting GeForce RTX 30, and so that the same situation does not occur with the GeForce RTX 40. The graphics card industry went through a difficult phase due to the cryptocurrency mining boom, you all remember that and you will also remember how it ended.

We’ve gone from a huge graphics card shortage and having to deal with reselling units at exorbitant prices to living in a market where supply has grown, both on the pre-owned and second-hand market, and demand has fallen. substantially. This led to prices having to be forced down, but in the end it was not possible to clear the stock in time and The GeForce RTX 40 eventually co-existed with the GeForce RTX 30.

The launch of the GeForce RTX 50 in 2025 will prevent this from happening again and leave plenty of room for vendors to clear inventory at a reasonable level. On the other hand, it’s clear that this will also give the GeForce RTX 40s a long life, and NVIDIA will almost certainly fill the gap in 2024 with an intergenerational rebuild from Ada Lovelace. Yes, I mean the rumored GeForce RTX 40 Super.