NVIDIA's upcoming GeForce RTX 50 "Blackwell" series will include mainstream offerings RTX 5060 Ti and RTX 5060, with new details emerging about their memory configurations. Both cards will utilize a "PG152" board SKU with the GB206 GPU core and feature a 128-bit memory bus interface, similar to their RTX 4060 predecessors.
The cards will incorporate GDDR7 memory technology, capable of reaching speeds up to 28 Gbps. This upgrade provides significant bandwidth improvements, offering 448 GB/s - a 55% increase over the RTX 4060 Ti and 65% more than the RTX 4060. Advanced memory compression techniques will help overcome any limitations of the 128-bit bus interface.
For memory capacity, the RTX 5060 Ti will feature 16GB VRAM, matching its predecessor's upgraded variant. While this larger capacity doesn't necessarily translate to significant performance gains, it helps meet the requirements of modern AAA games, particularly when gaming at 1440p resolution with ray tracing or high-resolution textures enabled.
The RTX 5060 will maintain an 8GB VRAM configuration, positioning it primarily as a 1080p gaming card with some 1440p capabilities through upscaling technology. Despite competitors like Intel's Arc B580 and AMD's RX 7600 XT offering 12GB and 16GB VRAM respectively, NVIDIA appears committed to this specification at launch, though changes could occur as higher capacity G7 modules become more widely available.
Both cards are scheduled for release in late February or March 2025. The preliminary specifications table for the RTX 50 series shows:
All cards in the series will utilize the 16-pin power connector (12VHPWR), with specific TBP (Total Board Power) values yet to be announced for the 5060 Ti and 5060 models.
Price?