Nvidia Geforce RTX 3060 Ti close to RTX 3070 in new performance results

Nvidia’s new graphics card in the Geforce RTX 3000 series was not allowed to bask in the glory of the news long before reports of upgraded models appeared. The Geforce RTX 3080 performance card is said to have doubled the amount of memory, which even the slightly simpler sibling RTX 3070 is said to have. The middle-class family member, Geforce RTX 3060, has not even had time to launch before information about an upgraded variant circulates frequently on the web.

The graphics circuit used in the Geforce RTX 3060 Ti is said to be the same GA104 used in the big sibling RTX 3070, but with the capacity slightly scaled down. A number of performance results are now being searched on the web and communicated by the constant Twitter leak TUM_APISAK (via Videocardz).

In terms of the results in the 3DMark Fire Strike and Time Spy test suites, the Geforce RTX 3060 Ti manages to reach points of 30,706 and 12,175, respectively. Compared to the corresponding results for the Geforce RTX 3070, this is about 10 percent lower, and the previous generation top card Geforce RTX 2080 Ti typically reaches about 15 percent higher earnings. The recently launched competitor AMD Radeon RX 6800 typically reaches about 40 percent higher scores in 3DMark Fire Strike and 25 percent higher in Time Spy.

When Geforce RTX 3060 Ti testas i Geekbench the middle class shows about 8 percent lower performance than the RTX 3070 in terms of CUDA calculations, but unexpectedly about 10 percent faster when the Vulcan interface is used. On the same unexpected track, it also achieves 30 percent better results than the Radeon RX 6800. In the game and performance test Ashes of the Singularity, the RTX 3060 Ti gets a score of 9,700 in the resolution 1,920 × 1,080 pixels and the graphics setting CRAZY.

It is 14 percent slower than its sibling RTX 3070 and 20 percent slower than the Radeon RX 6800. When the resolution is changed to 4K UHD, the ratio is adjusted slightly as the RTX 3060 Ti is only 8 percent slower than the RTX 3070, but 24 percent slower than the Radeon RX 6800. does not really match the ratio in specifications between Nvidia’s sibling cards, as the RTX 3060 Ti scales down the number of CUDA cores to 4,864, which is a reduction of about 17 percent from the 5,888 CUDA cores in the RTX 3070.

Specifications: Geforce RTX 3060 Ti versus RTX 3060 and RTX 3070

* Specifications are based on leaked information and are speculative until confirmed.

Further comparisons show reported 80 ROP units in the RTX 3060 Ti, compared to 96 in the RTX 3070. In terms of clock frequencies, the newcomer is said to run at a base speed of 1,410 MHz and a boost level of 1,665 MHz, for a total computing capacity of 16.2 TFLOPS . In comparison, the RTX 3070 runs at a base speed of 1,500 MHz and a boost frequency of 1,725 ​​MHz, where a computing capacity of 20.31 TFLOPS means a difference of about 20 percent compared to the little sibling.

According to previous information, the graphics memory measures in at 8 GB GDDR6, the same amount that the standard model of RTX 3070 is equipped with. The memory bus is also 256 bits wide here, and if the memory frequency is the same 14,000 MHz used in the RTX 3070, the bandwidth measures in at 448 GB / s. In terms of price level, the Geforce RTX 3060 Ti is stated to land around USD 399 or approximately SEK 4,300 including VAT. The graphics card will be launched on December 2, 2020.

Read more about Nvidia’s upgraded Geforce RTX 3000 models: