Nvidia SUPER: Only small updates or already an attack on AMDs Navi GPUs?

The Nvidia GeForce RTX 2080 Super is still based on the TU104 chip, but it now features 3072 shader units. This is an increase of 128 shader units compared to the GeForce RTX 2080. More RT and Tensor cores also result in improved Ray Tracing performance. Nvidia also raised the clock of the VRAM significantly for more performance (15.5 GHz, 496.1 GB/s). However, both the size of the VRAM and the memory interface are identical.

The performance of the GeForce RTX 2080S is supposed to be better than the Nvidia Titan XP according to Nvidia. If all the new features are supported, the GeForce RTX 2080S will also be much faster than the GeForce GTX 1080 Ti. This means the RTX 2080 Super is a full-fledged high-end GPU and can handle most games of 2018/2019 smoothly in high settings and 2160p. There is currently no direct competitor for the Nvidia GeForce RTX 2080 Super from AMD. This might also be the reason for the slight delay of the availability.

Nvidia specifies the power consumption of the GeForce RTX 2080 Super with 250 Watts, so slightly more than the predecessor (GeForce RTX 2080 215 Watts). Compared to the similarly powerful Nvidia Titan XP, the consumption is identical, but the performance per Watt was improved significantly.

The Nvidia GeForce RTX 2070 Super is now based on the TU104 chip, which is also used for the faster Nvidia GeForce RTX 2080. The chip of the RTX 2070 Super features a total of 2560 shader units, so 256 more compared to the GeForce RTX 2070 (TU106, 2306 shader units). The VRAM is identical with 8 GB GDDR6 RAM and a 256-bit interface (14 GHz, 448 GB/s). The RTX 2070 Super offers more Ray Tracing performance thanks to 40 RT cores; and the number of Tensor cores was also increased to 320.

The performance of the GeForce RTX 2070 Super should be slightly above the GeForce GTX 1080 Ti according to Nvidia. With the support of new features like DLSS and Real Time Ray Tracing, the GeForce RTX 2070 Super should be much faster than the GeForce GTX 1080 Ti. The new GPU should be able to handle smooth gameplay in most titles from 2018/2019 in high settings and 2160p. The GeForce RTX 2070 Super should compete with AMD's upcoming Navi generation (like the AMD Radeon 5700 XT).

Nvidia specifies the power consumption of the GeForce RTX 2070 Super with 215 Watts (TGP), but the performance per Watt is much better compared to a similarly powerful GeForce GTX 1080 Ti (rated at 250W).

Nvidia took on a popular point of criticism with the GeForce RTX 2060 Super and equips the new graphics card with more VRAM. The Nvidia GeForce RTX 2060 now also features 8 GB GDDR6 VRAM; only the Nvidia GeForce RTX 2080 Ti and the Nvidia Titan RTX offer even more VRAM. The new Nvidia GeForce RTX 2060 Super is based on the same chip compared to the Nvidia GeForce RTX 2060, but now offers more shader units (2176). As a reminder: The GeForce RTX 2060 only featured 1920 shader units. The number of RT and Tensor cores was also increased, which should improve the Ray Tracing performance.

The performance of the Nvidia GeForce RTX 2060 Super should be slightly better than the GeForce GTX 1080 according to Nvidia. If the new RTX features are supported, the Nvidia GeForce RTX 2060 Super should have an even bigger advantage compared to the old Pascal GPU. The performance target of the Nvidia GeForce RTX 2060 Super is the upper mid-range segment and it should be able to handle most games of 2018/2019 smoothly in high settings and 1440p. The direct competitor from AMD is the Radeon RX 5700.

Nvidia specifies the power consumption of the new chip with 175 Watts (TGP). This is slightly more compared to the GeForce RTX 2060 (160 Watts), but the performance per Watt was once again improved and the slightly slower GeForce GTX 1080 consumed a bit more power (180 Watts).