Preliminary specifications of Nvidia's performance mainstream GeForce RTX 4060 Ti graphics card indicate that the board won't be as power hungry as the company's high-end offerings, with a short PCB that will fit nicely into compact PCs. Meanwhile, even though this product is likely to join the ranks of the best graphics cards, its performance and specs may be lower than some might desire.
Hardware leaker @Kopite7Kimi late on Tuesday said that that Nvidia's GeForce RTX 4060 Ti will be based on the AD106 graphics processor with 4352 CUDA cores that will be paired with 8GB of 18 Gbps GDDR6 memory using a 128-bit interface. The product will consume around 220W — not a lot by today's standards — and the reference card is said to use a "very short" printed circuit board, which will make it easier to install into compact PCs.
Meanwhile, the GeForce RTX 4060 Ti Founders Edition will continue to use the notorious 12VHPWR power connector. Why does a 220W part even need a 16-pin connector with the potential to deliver 600W? We'll leave that for the readers to debate.
Nvidia GeForce RTX 40-Series Specifications
| Row 0 - Cell 0 | GPU | FP32 CUDA Cores | Memory Configuration | TBP |
| GeForce RTX 4090 | AD102 | 16384 | 24GB 384-bit 21 GT/s GDDR6X | 450W |
| GeForce RTX 4080 | AD103 | 9728 | 16GB 256-bit 22.4 GT/s GDDR6X | 320W |
| GeForce RTX 4070 Ti | AD104 | 7680 | 12GB 192-bit 21 GT/s GDDR6X | 285W |
| GeForce RTX 4070 | AD104 | 5888 | 12GB 192-bit 21 GT/s GDDR6X | 250W |
| GeForce RTX 4060 Ti | AD106 | 4352 | 8GB 128-bit 18 GT/s GDDR6 | 220W |
Nvidia has not confirmed and of the specifications for its upcoming products yet, so take them with a grain of salt. Meanwhile, based on preliminary specifications of Nvidia's GeForce RTX 4070 revealed last week, it looks like the gap between the RTX 4060 Ti and the RTX 4070 will be quite significant. Assuming more or less similar clocks, we are talking about at least 26% less compute performance from the 4060 Ti, which is substantial.
Considering the improvements of the Ada Lovelace architecture over the Ampere architecture, as well as higher clocks enabled by TSMC's 4N fabrication technology, we may expect the AD106 GPU and the GeForce RTX 4060 Ti board to feature significantly more compute horsepower than the GeForce RTX 3060 Ti. However, its memory bandwidth of 288 GB/s will be a whopping 41% lower compared to its predecessor. The 32MB L2 cache should help to mitigate the difference, but the question is whether it will eliminate cases when the new GeForce RTX 4060 Ti will lag behind the GeForce RTX 3060 Ti in high resolutions and/or with antialiasing enabled.
Perhaps worse, even the vanilla GeForce RTX 3060 carries 4GB more memory than the upcoming GeForce RTX 4060 Ti, which is a bit odd. Given the 128-bit memory bus and a relatively small number of CUDA cores, we presume Nvidia designed its AD106 primarily with laptops in mind. Regardless, we already felt the 8GB of VRAM on the 3060 Ti, 3070, and 3070 Ti was becoming a limiting factor, and that was two years ago. A mainstream performance card launching in 2023 and still using 8GB of memory feels almost insulting to gaming enthusiasts.
Using AD106 for a GeForce RTX x060 Ti part will of course allow the company to increase its profits, but from a gamer's point of view, what matters is performance and ability to play games upcoming games. We expect the 4060 Ti will easily surpass the existing 3060 Ti, thanks to its substantially higher clocks, but we'll have to see how it stands up to other GPUs once it launches.
Of course this is all unofficial and preliminary information that may not be accurate or final. Nvidia may change its plans as it gets closer to the GeForce RTX 4060 Ti launch, or the specs might be for a laptop variant. But for now, the GeForce RTX 4060 Ti looks rather pale and may only match the existing RTX 3070 Ti.
