Nvidia RTX 5060 reportedly packs 25% more CUDA cores than its predecessor
Following yesterday's leak detailing the RTX 5060 Ti and RTX 5050 specifications, the missing piece of the puzzle has finally fallen into place: the RTX 5060. The same leaker has now provided the full specifications of the upcoming GPU.
According to kopite7kimi (via videocardz), all three cards will use the same PG152 board design, which apparently supports two distinct GPUs (GB206 and GB207) and two memory types (GDDR6 and GDDR7). This suggests a unified platform, potentially streamlining production and reducing costs.
As predicted, the RTX 5060 non-Ti appears to be a slightly less powerful variant of the 5060 Ti. It features 3,840 CUDA cores, approximately 25% more than the RTX 4060, a higher 150W TGP, 35W more than its predecessor, and 8 GB of GDDR7 memory across a 128-bit memory bus. While clock speeds remain undisclosed, the core count increase, faster VRAM, and TGP difference suggest a decent performance uplift, at least more than the rest of the RTX 50 series brought compared to its RTX 40 series counterparts.
Nvidia is reportedly planning to announce the RTX 5060 alongside the RTX 5060 Ti and RTX 5050 next week. These cards are expected to become available starting in April, but we're unsure if they'll all be released simultaneously.
KitGuru says: Of all three reportedly upcoming Nvidia RTX 50 series cards, which one looks the most interesting to you?
The post Nvidia RTX 5060 reportedly packs 25% more CUDA cores than its predecessor first appeared on KitGuru.