According to preliminary specifications, Nvidia’s upcoming mainstream GeForce RTX 4060 Ti graphics card will be less power-demanding than the company’s high-end products, with a short PCB that will fit easily into tiny PCs. Meanwhile, while this device is anticipated to reach the ranks of the top graphics cards, its performance and specifications may be less than what some may desire.
Late on Tuesday, hardware leaker @Kopite7Kimi revealed that Nvidia’s GeForce RTX 4060 Ti will be built on the AD106 graphics processor with 4352 CUDA cores and 8GB of 18 Gbps GDDR6 memory through a 128-bit interface. The reference card is supposed to use a “very short” printed circuit board, making it easier to install into tiny PCs, and it will consume roughly 220W – hardly much by today’s standards.
Nvidia has not yet confirmed or denied any of the specifications for its future devices, so take them with a grain of salt.
Meanwhile, based on the preliminary specifications of Nvidia’s GeForce RTX 4070 announced last week, the difference between the RTX 4060 Ti and the RTX 4070 appears to be pretty large. Assuming similar clocks, we’re looking at at least 26% less compute performance from the 4060 Ti, which is significant.
Given the Ada Lovelace design’s advancements over the Ampere architecture, as well as the higher clocks provided by TSMC’s 4N fabrication technology, we can expect the AD106 GPU and GeForce RTX 4060 Ti board to have much greater compute horsepower than the GeForce RTX 3060 Ti. However, its memory bandwidth of 288 GB/s will be 41% lower than that of its predecessor. The 32MB L2 cache should help to reduce the disparity, but the question is whether it will remove situations in which the new GeForce RTX 4060 Ti lags behind the GeForce RTX 3060 Ti at high resolutions and/or with antialiasing enabled.
Worse, the vanilla GeForce RTX 3060 has 4GB more RAM than the future GeForce RTX 4060 Ti, which is strange. Given the 128-bit memory bus and a small number of CUDA cores, we believe Nvidia developed the AD106, particularly for laptops. Regardless, we were already concerned that the 8GB of VRAM on the 3060 Ti, 3070, and 3070 Ti was becoming a limiting issue two years ago. A mainstream performance card debuting in 2023 with 8GB of memory feels almost insulting to gamers.
Using AD106 for a GeForce RTX x060 Ti part will, of course, enhance the company’s earnings, but from the perspective of a gamer, what matters is performance and the ability to play upcoming games. We expect the 4060 Ti to easily outperform the present 3060 Ti because to its significantly higher clocks, but we’ll have to wait and see how it compares to other GPUs once it’s released.
Also Read:
