When it comes to Nvidia graphics cards, those with the -60 suffix have traditionally been the most popular amongst PC gaming fans. However, some product lines have had an ‘entry-level’ model, usually labelled with the -50 suffix. According to recent leaks, Nvidia is bringing this tier back for the Blackwell GPU family in the form of the RTX 5050.

According to recent rumors, the desktop version of the Nvidia GeForce RTX 5050 graphics card will utilize the GA107 GPU, featuring 2560 CUDA cores and 20 streaming multiprocessors. Leaks from the Geekbench site regarding the laptop version suggest it will use the same GPU chip and also feature 20 Compute Units, making it essentially have the same configuration as the desktop version.

The desktop version of this card is expected to have 8GB of GDDR6 memory, operating at 20 Gbps. The memory type for the laptop version is unknown, but it is likely to use a similar specification. The TDP has not yet been confirmed.

On par with RTX 4060

The Geekbench OpenCL score indicates that the RTX 5050 is roughly 14% slower than the RTX 5060 laptop GPU. The RTX 4060 laptop GPU score on Geekbench OpenCL is 89,206, suggesting it has similar performance to the RTX 5050 laptop. The RTX 5050 also seems to be roughly 19% faster than the RTX 4050 laptop GPU from last generation. These are only preliminary figures, however, especially since only one version of the RTX 5050 has been benchmarked so far.

There is no information to indicate when these graphics cards and laptops will officially launch. Until Nvidia makes an official announcement, this information should be treated as speculation and taken with the appropriate grain of salt. Last generation, the 4050 was a laptop-exclusive GPU, but a desktop version of the 5050 is reportedly in the works.

About the Author

Rebecca Hills-Duty

Writing and journalism experience at VRFocus, UploadVR, The Escapist, HTC, PC Gamer, Tech Radar+ and Dexerto. Can sometimes be found playing with retro tech.