Nvidia with Plan B: RTX 5050 is delayed, RTX 3060 should fill the gap

Nvidia is reportedly planning a surprise return to the graphics card market. Because the RTX 5050 is likely to be delayed, the RTX 3060 is scheduled to fill the gap in June. This would have both advantages and disadvantages for price-conscious buyers.

Comeback of the Ampere architecture

Rumors have been circulating for a long time about a new edition of the GeForce RTX 3060. The old graphics card model was supposed to go on sale again at the end of March. But nothing came of it. There are now increasing indications that Nvidia will bring the two-generation-old hardware back onto the market in June 2026.

The reason for this is an alleged delay in the planned GeForce RTX 5050, which was actually intended to serve the segment for price-conscious buyers. The upcoming Blackwell entry-level card was originally supposed to come with nine gigabytes of video memory and a 96-bit memory interface. It is not known why there are delays now. The return of the RTX 3060 with twelve gigabytes of GDDR6 memory would represent a pragmatic solution to quickly close the gap in the lower price segment.

Production at Samsung instead of TSMC

The leaker shared the information MEGAsizeGPU on X (via WccfTech). According to previous leaks, the company will probably again use Samsung’s 8-nanometer process to produce the older chips. This would relieve the modern 5-nanometer capacities at the contract manufacturer TSMC. The valuable capacities at TSMC are urgently needed for current Blackwell graphics processors as well as lucrative hardware for AI data centers.

In addition, the older GDDR6 memory is cheaper to purchase on the global market than the new GDDR7 standard. The big advantage for customers of a new RTX 3060 edition would be the twelve gigabytes of memory, which is rare in this affordable price range and clearly exceeds the nine gigabytes of the RTX 5050.

Old technology versus new software

The RTX 3060 also has 3,584 CUDA cores and benefits from a 192-bit memory interface. It is also superior to the RTX 5050 in these points. However, this is offset by the outdated technology of the Ampere generation, which no longer delivers the highest efficiency. The architecture does not offer the latest tensor cores or current ray tracing units. In addition, it does not support many newer DLSS functions for AI-supported image enhancement.

This can mean a noticeable performance disadvantage, especially with modern titles. Buyers of an RTX 3060 would definitely have to make compromises here. Whether the old model is a sensible alternative for PC gamers depends largely on the exact pricing. At the moment, any information remains unconfirmed as Nvidia has so far strictly refused to make an official statement on this topic. But it won’t be too long until the supposed release in June 2026.