NVIDIA GeForce RTX 5050 & RTX 5060 Ti Specifications Revealed: 5060 Ti Features 4608 Cores & Up to 16 GB GDDR7, While 5050 Offers 2560 Cores & 8 GB GDDR6, Both with 128-Bit Interface

Recent leaks have unveiled crucial specifications for NVIDIA’s forthcoming GeForce RTX 5060 Ti and RTX 5050 graphics cards, reinforcing earlier hints shared just days ago.

Key Specifications Unveiled for NVIDIA GeForce RTX 5060 Ti and RTX 5050

The eagerly anticipated GeForce RTX 50 series is set to make waves in both mainstream and entry-level markets, with the 60-class offering (RTX 5060 Ti) and the 50-class variant (RTX 5050).Insight into these specifications has been provided by prominent leaker @kopite7kimi on X, delivering noteworthy insights into performance metrics.

According to the leaked information, the NVIDIA GeForce RTX 5060 Ti will utilize the GB206-300 GPU die, while the RTX 5050 will be based on the GB207-300 architecture. The RTX 5060 Ti, being the faster of the two, will boast 4608 CUDA Cores, representing a performance increase of approximately 6% over its predecessor, the RTX 4060 Ti. Additionally, it will be available in two VRAM configurations: 8 GB and 16 GB, both utilizing advanced GDDR7 memory. The 16 GB version is anticipated to launch in late March, with the 8 GB variant following in April concurrent with the RTX 5060 (Non-Ti) release.

GeForce RTX 5060 TiPG152-SKU10/15GB206-300-A14608FP32128-bit GDDR7 8/16G180W

— kopite7kimi (@kopite7kimi) March 9, 2025

Both graphics cards, except for the RTX 5050, exhibit a consistent memory bus width of 128 bits. This design choice results in a memory bandwidth of 448 GB/s, which is around 55% superior compared to the RTX 4060 Ti. The GeForce RTX 5060 Ti will make use of the PG152-SKU10/15 board alongside the GB206-300-A1 GPU die, while the RTX 5050 remains committed to the GDDR6 standard, promising memory speeds of 20-22 Gbps for over 300 GB/s of total bandwidth.

The thermal design power (TDP) for the NVIDIA GeForce RTX 5060 Ti is currently pegged at 180W, representing a 20W increase from its predecessor. On the entry-level front, the RTX 5050 marks NVIDIA’s return to offering a 50-class card after omitting it in the RTX 40 series lineup. This discrete GPU will utilize the Blackwell architecture and feature a PG152-SKU50 board alongside the GB207-300-A1 die with 2560 CUDA Cores. Previously, NVIDIA resorted to using XX7-class dies in their RTX 4060 models, but there’s hope that the trend will change this time around for the RTX 5050.

GeForce RTX 5050PG152-SKU50GB207-300-A12560FP32128-bit GDDR6 8G130W

— kopite7kimi (@kopite7kimi) March 9, 2025

It’s worth noting a slight adjustment from earlier reports regarding the RTX 5050; the anticipated TDP is now confirmed at 130W, a marginal difference from the previously speculated 135W. Nevertheless, this model will include 8 GB of VRAM, while the 16 GB variant is expected to have a total board power (TBP) of 135W, mirroring the uplift seen in the RTX 4060 Ti 16 GB variant.

Though the exact launch date for the RTX 5050 remains unannounced, the RTX 5060 series, which includes both the RTX 5060 and RTX 5060 Ti, is projected to debut shortly, with retail availability anticipated next month.

NVIDIA GeForce RTX 50 Graphics Card Specifications (Preliminary)

For further details and imagery, refer to the source.

Related Articles:

AMD Introduces FSR 4 Support for Four New Games in Adrenalin 25.6.2 Drivers, Resolves Multiple Issues for Radeon GPUs

16:07June 12, 2025

Intel Arc Battlemage GPU Featuring PCIe 5.0 x16 Interface Introduced to PCI-SIG

13:23June 12, 2025

Stream the AMD “Advancing AI” Keynote Live: Discover Next-Gen EPYC, Instinct, and AI Product Lineup

13:11June 12, 2025

Next-Gen AMD CDNA Architecture Utilizing “GFX1250” Graphics IP, Followed by UDNA GPU Generation

12:31June 12, 2025