The new version of the junior model may become Nvidia's response to the memory crisis.
NVIDIA is preparing a new version of the GeForce RTX 5050 video card with an updated memory configuration. According to well-known insider kopite7kimi, the model will receive 9 GB of GDDR7 memory, while the power consumption level will remain the same — about 130 W.
The current version of the RTX 5050 is equipped with 8 GB of GDDR6 memory with a speed of 20 Gbps and uses a 128-bit bus. This configuration provides a bandwidth of about 320 GB/s. In the new modification, NVIDIA is allegedly switching to three GDDR7 memory modules of 3 GB each. As a result, the total volume will increase to 9 GB, but the bus width will be reduced to 96 bits.
At the same time, faster memory should partially compensate for the reduction in the bus. It is expected that GDDR7 will operate at a speed of about 28 Gbps, which will increase the bandwidth to 336 GB/s — approximately 5% more than the current version. Thus, the amount of video memory will increase by 12.5%, and the total bandwidth will increase slightly.
The graphics processor itself, according to the leak, will remain the same. The video card will still use the GB206 chip with 2560 CUDA cores.
Media suggest that one of the reasons for switching to GDDR7 was the situation on the memory market. GDDR6 supplies for the mass segment are becoming less stable, while GDDR7 production, on the contrary, is gradually increasing. The use of 3 GB modules also reduces their number — from four to three.
Rumor has it that the new version of the GeForce RTX 5050 may be presented closer to Computex 2026, which will be held from June 2 to 5 in Taipei. NVIDIA has not yet officially confirmed the existence of such a modification.
