The specifications for NVIDIA’s upcoming GeForce RTX 5060 Ti and RTX 5050 graphics cards have recently been revealed, confirming some previous speculations. These new GPUs are part of NVIDIA’s 50 series lineup, which aims to cater to mainstream and entry-level markets.
According to reliable sources, notably the leaker @kopite7kimi, the GeForce RTX 5060 Ti will utilize the GB206-300 GPU architecture, while the RTX 5050 will run on the GB207-300 die. The RTX 5060 Ti is set to be the more powerful option among the two, featuring 4,608 CUDA cores. This represents an increase of roughly 6% over the RTX 4060 Ti. The RTX 5060 Ti will be available in two memory options: one with 8 GB and the other with 16 GB of GDDR7 memory. Notably, the 16 GB version is expected to hit stores by the end of March, followed by the 8 GB variant in April, which will coincide with the release of the standard RTX 5060.
The memory bandwidth for the RTX 5060 Ti is impressive, reaching 448 GB/s, an enhancement of about 55% compared to the memory bandwidth of the RTX 4060 Ti. Both the RTX 5060 Ti and RTX 5050 are equipped with a memory bus width of 128 bits. The RTX 5050, unlike its more powerful counterpart, will feature GDDR6 memory with speeds between 20 to 22 Gbps, resulting in a total bandwidth exceeding 300 GB/s.
In terms of power consumption, the RTX 5060 Ti will have a thermal design power (TDP) of 180W, which is a slight increase from its predecessor. On the other hand, the RTX 5050 is positioned as an entry-level offering, likely coming in with a lower power requirement. Initially, it was suggested that the RTX 5050 would have a TDP of 135W, but the latest information indicates it will be 130W with an 8 GB memory option.
As for their configurations, information shows the RTX 5050 will include 2,560 CUDA cores and will also have a discrete nature within NVIDIA’s Blackwell microarchitecture. The architecture for the GTX series will also include a PG152-SKU50 board paired with the GB207-300-A1 die.
Here’s a quick summary of what’s expected from the NVIDIA RTX 50 series:
-
GeForce RTX 5060 Ti:
- CUDA Cores: 4,608
- VRAM: Available in 8 GB and 16 GB GDDR7
- Memory Bandwidth: 448 GB/s
- TDP: 180W
- Expected Release: March 2025 for 16 GB option, April 2025 for 8 GB option.
- GeForce RTX 5050:
- CUDA Cores: 2,560
- VRAM: 8 GB GDDR6
- Memory Bandwidth: 300 GB/s (approx.)
- TDP: 130W
- Expected Release: April 2025.
As these graphics cards near their release dates, anticipation is building among gamers and PC enthusiasts alike, especially with the promise of improved performance at more accessible price points. Both graphics cards are anticipated to provide significant enhancements over their predecessors, making them appealing options for those looking to upgrade their gaming setups. With the RTX 50 series, NVIDIA appears to be strategically targeting both performance-oriented and budget-conscious consumers alike.