Nvidia GeForce RTX 5050 Launches at $249: Specs, Features, and More

Today, Nvidia unveiled the GeForce RTX 5050, the smallest model in the RTX 50 series. It will go on sale in the second half of July and start at $249.

Two thirds of the RTX 5060, which debuted in April, have 2560 CUDA cores, while the RTX 5050 has a new GB207 die based on the Blackwell architecture. Like previous 50-series GPUs, it also has 4th generation RT cores and 5th generation Tensor cores. This allows it to support features like the future Reflex 2 and DLSS4 with multi-frame generation.

The base clock speed of the RTX 5050 GPU is 2.31GHz, with a boost to 2.57GHz. Since it comes with 128-bit 8GB GDDR6 memory, it is the only 50-series card that does not yet have GDDR7. Additional features include one 9th generation NVENC encoder, one 6th generation NVDEC decoder, and three DisplayPort 2.1b and one HDMI 2.1b connection. The card requires 130W of power and can accept a PCIe Gen 5 cable of at least 300W or a single 8-pin connector.

Nvidia has that the new card is 60% quicker on average in raster than the RTX 3050, which was released more than three years ago, but they only have a few largely useless graphs that make it difficult to determine the relative performance of the RTX 5050 compared to older Nvidia cards. In terms of performance, the card appears to be about average compared to the RTX 4060 from two years ago.

There should be the RTX 5050 available by this time next month. All of the major board partners, including ASUS, Colorful, Gainward, Galaxy, GIGABYTE, INNO3D, MSI, Palit, PNY, and ZOTAC, will sell the card; however, Nvidia will not be offering a Founders Edition variant.

The RTX 5050 is coming today in laptops with a starting price of $999, and a mobile version will also be available.