The GeForce RTX 50 series has become one of the big targets of the “rumorology” world, and because of that we have a lot of information that we’ve sifted through on several occasions to tell you only the most plausible details. In this sense, one of the most famous sources industry provided more information which reinforces the base we previously had on this new generation of NVIDIA graphics cards.

First, the rumors and leaks are reinforced saying that the GeForce RTX 50 series will use the Blackwell architecture and that unlike what happened with the GeForce RTX 40 there won’t be two different architectures (Ada Lovelace came to the consumer and Hopper to the professional sector). This makes sense as NVIDIA has already followed through on this bet on previous occasions. For example, there was no difference between the architectures with Ampere and neither with Turing.

Second, NVIDIA is practically expected to use TSMC’s 3nm node with the GeForce RTX 50. This node represents an important evolution compared to the 5nm that was used in Ada Lovelace, and will allow significant improvements in both efficiency and performance levels. On the other hand, it should be taken into account that this node changes will reduce the size of the transistorssomething essential for increase their density.

Finally, there is also an important change in the range of outputs that the GeForce RTX 50 will offer, as this new generation is said to use DisplayPort 2.1 connector. This connector increases the available bandwidth and allows for much higher resolutions and refresh rates. We don’t know if it will be available in all models of that new generation, including the most economical ones, but it would be normal.

Five things we expect from GeForce RTX 50

• Higher performance and better performance values ​​per watt consumed.
• Fourth generation RT cores with improved ray tracing-path tracing performance.
• Fifth generation tensor cores with more AI power.
• Increasing the amount of graphics memory in mid-range and mid-range models.
• New technologies and optimizations for ray tracing and AI, as well as advancements within the NVIDIA DLSS ecosystem.

If all goes according to plan, NVIDIA will introduce the GeForce RTX 50 between late 2024 and early 2025. From what we’ve seen in previous generations, I think it’s most likely that the GeForce RTX 5090 and 5080 will be introduced first. between October and November 2024and that the rest of the models will begin to be introduced in 2025.