A public and anonymous GPU-Z validation confirms the specifications of the GeForce RTX 3060 Ti, NVIDIA’s second graphics card based on the Ampere GA104 GPU, which would have similar or slightly better performance than the RTX 2080. RTX confirmed 3060 Ti, which appears to be made by board partner HP, has 4864 CUDA cores and 80 ROPs.
More importantly, the validation confirmed the graphics card frequencies. The RTX 3060 Ti would have a base clock of 1410 MHz and a boost clock of 1665 MHz, both reported as defaults by the software.
The graphics card would be equipped with 8 GB of GDDR6 memory with a 256-bit bus with a speed of 14 Gbps, the same configuration as RTX 3070, which NVIDIA will officially launch in two days.
NVIDIA has yet to confirm the GeForce RTX 3060 Ti, but the AIBs are very busy preparing for this release, our sources claim. Pricing and release date are yet to be set, but the model is expected to launch in mid-November priced below $ 400.
Listings on Taobao (China) are currently between 2,049 and 2,999 yuan (305-446 USD), although other sources claim that the suggested launch price does not exceed $ 400 USD.
| NVIDIA GeForce RTX 30 Series | ||||
|---|---|---|---|---|
| RTX 3090 | RTX 3080 | RTX 3070 | RTX 3060 Ti | |
| Reference Board | PG132 SKU 30 | PG132 SKU 10 | PG142 SKU 10 | PG190 SKU 10 |
| GPU | 8nm GA102-300 | 8nm GA102-200 | 8nm GA104-300 | 8nm GA104-200 |
| Die Size | ||||
| Transistors | ||||
| CUDA Cores | ||||
| GPU Utilization | ||||
| Tensor Cores | ||||
| RT Cores | ||||
| Base Clock | ||||
| Boost Clock | ||||
| Shader Perf. | ||||
| Memory | ||||
| Memory Clock | ||||
| Memory Bus | ||||
| Maximum Bandwidth | ||||
| Consumption | ||||
| Price | ||||
| Launch Date | September 24 | September 17 | October 29 | Mid November |