[Hardware]Nvidia GeForce RTX 3060 Ti Founders Edition Review

[WilkerCSBD]

GeForce RTX 3060 Founders Edition Features Introduction
This third model of the Nvidia Ampere saga brings the improvements of this generation to the performance of the RTX 2080 Super with a price of around 400 Euros and a lower consumption, much more manageable than that of its older sisters, with 200 watts of TDP consumption. . It also has a more compact format and a much more accessible entry price.

With a performance only 10% slower than an RTX 3070, the review of the Nvidia Geforce RTX 3060 Ti will not reveal new levels of performance, but it will bring us closer to the performance of the higher-end models for a fraction of the price and consumption. With 200w, and an ITX-type PCB design, it is also a good basis for different integrators to finally be able to incorporate more compact designs to a range that is not yet meeting the needs of PC users who are looking for solutions with less consumption and smaller volumes.

The design of the Nvidia reference model is very similar to that of its older sisters, with a large laminated heatsink and two axial fans to cool it that this time are on the same side. It also uses the same 12-pin connector, although for consumption you will never need more than an 8-contact connector and in fact the adapter that it brings is designed precisely for that.

Technical characteristics of the NVIDIA GeForce RTX 3060 Ti FE:
Memory: 8GB GDDR6
Graphics Engine: GA104 Nvidia Ampere
Bus: PCI Express 4.0.
Memory interface: 256-Bit
Base Frequencies: 1350 MHz
Boost frequency: 1665 MHz
Memory speed: 14Gbps
Total screens: 4
HDCP support: 2.2
TDP consumption: 200w
Connectivity: DisplayPort x 3 (v1.4a) / HDMI 2.1 x 1
Recommended power supply: 500w
Power connectors: 1 x 12 PEG pins
Dimensions: 240 x 112 x 2 slots
New Nvidia Second Generation Ray Tracing Features
This is the second generation of RTX cards from Nvidia, which comes with Samsung's new 8nm manufacturing process, and major improvements in architecture. I am not going to go into much detail, since here we have an extensive article dedicated to all the news about these new GPUs, but I do want to highlight some of the most important elements that characterize this generation.

Among the improvements made in this generation, the RTX 3060 Ti is the shortest of the current models. One of the examples is found in the memory used, which is reduced to GDDR6 (as in the RTX 3070) compared to the latest generation GDDR6X of the RTX 3080 and RTX 3090. The data bus speed is reduced to 14GBps, compared to to the 19Gbps of its older sisters and that, added to the bus also shorter than 256-Bit, in a reduction of 50% of the year of band on the magnificent RTX 3080, but equaling its older sister the RTX 3070. The VRAM o framebuffer also has a 2GB cut to a healthy and reasonable 8GB of VRAM.

This model also enjoys the new second generation RT Core that are capable of doubling the calculation power of triangular intersections over the previous generation. In addition, the possibility of calculating widespread effects such as Motion Blur by hardware is introduced. The Tensor Core also enter its third generation and does so by reducing the number of units available in the GPU, but still quadrupling the processing capacity of the entire system in terms of AI-oriented sparse matrix calculation.

The RTX 3060 Ti will really compete with the RTX 2080/2080 Super, but it is not the generation it succeeds to, its target is undoubtedly the RTX 2060 Ti / Super, to surpass it easily and the competition cards of It is on AMD RDNA2 architecture that is yet to reach the market.

The most notable differences are found in the configuration of the GPU, with its improvements in the manufacturing process and architecture that make this card achieve much higher performance, up to 30-40% over the reference models of the previous generation while maintaining a price very similar. Among the differences we see more than twice the number of CUDA units, going from 2304 to 5888 shading motors.

The classic rasterization units also increase from 64 units to 80, always comparing it with its natural predecessor, the RTX 2060 Super. The Tensor Core are down from 272 units to just 152 but noticeably more powerful than the previous generation. The number of SM units also changes from 32 to 38 in the new model, which also translates into an increase in the RT Core units to the same figure and also a notable increase in the CUDA units where we more than doubled from 2176 units up to 4864.

The data bus is the same, with 448GBps of bandwidth, and the working frequencies are also very similar. We have 1410MHz of base frequency in RTX 3060 Ti, by 1470MHz in RTX 2060 Super, and a frequency