NVIDIA expands its RTX 30-series lineup with the launch of RTX 3060 Ti graphics card. It is based on NVIDIA’s new Ampere Architecture and slated to topple last years RTX 2060 Super.
The NVIDIA RTX 3060 Ti uses a slightly cut down version of GA104 GPU originally seen in RTX 3070. It has 38 SM units with 4864 Cuda cores, 152 (3d Gen) Tensor cores and 38 (2nd Gen) RT cores.
It has 152 tecture units, 80 ROPs, 1665MHz GPU boost clock speed and 8GB GDDR66 Video memory. Memory is clocked at 14Gbps with 256-bit memory interface and 448GB/s memory bandwidth. It has TDP of 200W.
NVIDIA claims that RTX 3060 Ti is faster than RTX 2080 Super. It has support for real-time Ray tracking and will easily handle any RTX or DirectXX Raytracing enabled titles.
It also comes with support for NVIDIA’s DLSS (Deep Learning Super Sampling). It is an AI-enhanced super sampling solution that leverages machine learning for image reconstruction at higher resolution. The has NVIDIA Reflex to reduce system latency and NVIDIA Broadcast for streamers.
The NVIDIA RTX 3060 Ti price is set at $399 and will be available from December 2. Ones from NVDIA will come with Founder Edition tag. Other OEM’s like EVGA, ASUS, Gainward, Galaxy, MSI, Palit, PNY, ZOTAC and Colorful will stock and factory overclock cards as well. [via]