New information regarding NVIDIA’s mid-range GPU that comes with 16GB VRAM has surfaced.Apparently, the RTX 4060 Ti GPU will have a different version of the graphics processing unit (GPU) and will need more power than its predecessors. This was disclosed in documents that NVIDIA shared with its board partners not long ago. However, it’s important to keep in mind that these are just proposed specs, and NVIDIA hasn’t provided the actual GPUs yet. So, there’s still a chance that these specifications might change.
Even though the upcoming RTX 4060 Ti GPU with 16GB VRAM will have a different product identification number (SKU) than the 8GB model (363 instead of 361), the board design will still be the same (PG190). What’s interesting is that NVIDIA has decided to use a slightly different GPU variant for this particular card, but the change is quite minor – it’s just a small upgrade from AD106-350 to 351. This probably means that the CUDA core count will remain the same, but some physical modifications to the design were still required.
It is not surprising that the power consumption will increase when the memory size is doubled. Nevertheless, the reported increase is only around 5 watts compared to the RTX 4060 Ti 8GB version which has a power consumption spec of 160 watts. At this point, it’s still unclear whether the 16GB variant will have the same memory clock speed of 18 Gbps, but it’s reasonable to assume that it will be the same for now.
It seems like NVIDIA has become more cautious after the launch mishap with the RTX 4080 12GB, so they are being more careful with the design of the different cards under the same name. This should result in the specifications being quite similar between both SKUs. As of now, there are only three changes being discussed – the capacity, GPU variant, and TDP.
The RTX 4060 Ti is scheduled to be released in the second half of July. The 8GB model, on the other hand, is expected to make its debut on May 24th.