The GeForce RTX 5060 will be one of the most important next-generation graphics cards from NVIDIA, as it will position itself as mid-range and will have a limited pricemaking it attractive to those who are on a tight budget but want to enjoy a good level of performance and all the features and technologies of the Blackwell architecture without having to make a big investment.

The last rumors we’ve seen about the GeForce RTX 50 focused mainly on the top-end GeForce RTX 5090 and GeForce RTX 5080 models, but thanks to all the information and details we were able to find out when NVIDIA Blackwell presented I have enough information to I can share with you an estimate of what we can expect from the GeForce RTX 5060.

GeForce RTX 5060 architecture, design and specifications

The GeForce RTX 5060 will use the Blackwell architecture. It will most likely come with the GB207 GPU made in TSMC 4nm nodeand which uses a monolithic core design similar to that of Ada Lovelace, that is, it has specialized hardware (tensor cores, RT cores and an optical flow accelerator) and also large amount of L2 cache memory.

The GB206 core will reserved for GeForce RTX 5060 Ti, and the GB204 core (it was also identified as GB205) will be the one that powers the GeForce RTX 5070 and GeForce RTX 5070 Ti. This split makes sense as it leaves the GB203 GPU as an option for the GeForce RTX 5080 and the GB202 as the GeForce RTX 5090 GPU.

The GPU GB206 will be less complex than its older sisters and therefore cheaper and easier to manufacture. It would be a perfect choice to boost the GeForce RTX 5060, but what specs would this graphics card have? This is a great question and I will share it with you. an estimate that seems pretty accurate to me and this could easily be accomplished:

• Graphics core GB206 made on TSMC’s 4nm node.
• 42 SM units.
• 5,376 shaders at speeds between 2.8 GHz and 3 GHz.
• 168 texturing units.
• 64 raster units.
• 42 cores of the fourth generation RT.
• 168 fifth generation tensor kernels.
• 128-bit bus.
• 12 GB of GDDR7 memory at 28 Gbps (448 GB/s bandwidth).
• 32 MB L2 cache.

With this configuration and running at 3 GHz The GeForce RTX 5060 would achieve 32.25 TFLOPs in FP32which is a figure that would place it above the GeForce RTX 4070 in raw performance, but behind the GeForce RTX 4080. The performance improvement compared to the GeForce RTX 4060 would be huge, and having 12 GB of graphics memory would be more balanced.

Possible release date and price

The GeForce RTX 5080 and GeForce RTX 5090 are expected to launch later this year. The GeForce RTX 5070 will arrive next, likely in January 2025, with the GeForce RTX 5060 Ti and GeForce RTX 5060 coming later between the first and second quarters of 2025.

Regarding the price will be placed between 300 and 400 dollars-euros. I dare not specify more, because the use of GDDR7 memory will mean an increase in costs, and it is clear that this will eventually affect the selling price of the GeForce RTX 5060.

AI generated images.