@KittyYYuko now also has shared alleged photographs of the upcoming RTX 4060 (Ti) design shortly after leaking the RTX 4070 Ti FE design.
The images from @KittyYukko show a small desktop card with the same design principle as all Founders Edition models based on Ada architecture. However, this card is much smaller and designed to fit into two PCIe slots only. The card is labeled RTX 4060, but there are suggestions that it might be RTX 4060 Ti, possibly because the specs may have been changed, or both cards are tested internally with the same cooler.
The RTX 4060 Ti will feature an AD106-350 GPU with 4352 CUDA cores, 8GB of GDDR6 memory, and a 128-bit memory bus. It is also said to have a reduced TDP of 160W, compared to the upcoming RTX 4070's 200W TDP. On the other hand, the RTX 4060 will use the entry-level AD107-400 GPU, featuring 3072 CUDA cores. The memory configuration is not expected to change, but the TDP will drop to just 115W. Both models are set to feature a PCIe Gen4 x8 interface. While there is no confirmation that NVIDIA plans to launch any RTX 4060 or RTX 4070 series cards as Founders Edition, the leaked images suggest the possibility. Though, it is worth noting that the card's legitimacy remains uncertain.
| Alleged GeForce RTX 40 Series Specs | ||||
|---|---|---|---|---|
| RTX 4070 Ti | RTX 4070 | RTX 4060 Ti | RTX 4060 | |
| Architecture | Ada (TSMC 4N) | Ada (TSMC 4N) | Ada (TSMC 4N) | Ada (TSMC 4N) |
| GPU | AD104-400 | AD104-250/251 | AD106-350 | AD107-400 |
| CUDA Cores | 7680 | 5888 | 4352 | 3072 |
| Base Clock | 2310 MHz | 1920 MHz | ||
| Boost Clock | 2610 MHz | 2475 MHz | ||
| Max FP32 Compute | 40 TFLOPS | 29 TFLOPS | TBC | TBC |
| Memory | 12 GB G6X | 12 GB G6X | 8 GB G6 | 8 GB G6 |
| Memory Bus | 192-bit | 192-bit | 128-bit | 128-bit |
| Memory Bandwidth | 504 GB/s | 504 GB/s | 288 GB/s | 288 GB/s |
| Default TGP | 285W | 200W | 160W | 115W |
| Interface | PCIe Gen4 x16 | PCIe Gen4 x16 | PCIe Gen4 x8 | PCIe Gen4 x8 |
| Release Date | January 5th, 2023 | April 13, 2023 | ||
