Specifications for the Geforce RTX Super Series can be found on the web

In August 2018, Nvidia unveiled the new graphics architecture Turing and the Geforce RTX 2000 series, through a trio of graphics cards in the upper performance layer. At the unveiling, the previous architecture had Pascal and the Geforce GTX 1000 series over two years on its neck, but a report claims that more powerful versions of the Geforce RTX 2000 cards are already in the works.

Contents hide

1 Specifications for Geforce RTX Super:

2 Related posts:

According to the information, there are five Turing-based graphics cards that maintain the current name scheme but are expanded with the suffix “Super”. With this, the performance should jump up a notch, while the price is similar. This means, for example, that the Geforce RTX 2060 Super will be able to compete with the Geforce RTX 2070, with the help of an upgraded graphics circuit and more graphics memory.

Now Igor Wallossek from German Tom’s Hardware reports on specifications for three of the upcoming products. At present, the two above-mentioned graphics cards are equipped with the graphics circuit TU106 in a different scaled-down design, where the Geforce RTX 2060 has 1,920 CUDA cores accompanied by 6 GB of GDDR6 memory over a 192-bit wide memory bus.

Read This Now: AMD seizes Battlefield 4 - market exclusivity for Radeon

The super version of the card will instead step up to 8 GB video memory and bus width of 256 bits, which means that the bandwidth increases from 336 to 448 GB / s. In the same vein, the number of cores is increased to 2,176 – just below the 2,304 seen on the Geforce RTX 2070. The latter graphics card in its Super edition is instead endowed with a scaled-down version of the TU104 graphics circuit from the Geforce RTX 2080.

Specifically, 2,560 CUDA cores and an unchanged memory of 8 GB are listed, but the step up to today’s Geforce RTX 2080 with 2944 CUDA cores is thus a long one. It is worth mentioning that the previous information is rumored that a Geforce RTX 2070 Ti Super will also take place in the product family. However, Wallossek does not mention this card and does not touch on any Geforce RTX 2080 Ti Super model.

Read This Now: Test ray tracing performance with Cryengine Neon Noir

Finally, details are listed for the Geforce RTX 2080 Super, which continues to use the TU104 circuit but with 128 extra active cores, for a total of 3,072. This includes 8 GB of graphics memory with a memory frequency increased from 14,000 to 16,000 MHz. Assuming that the bus width is still 256 bits, the memory bandwidth then increases from 448 to 512 GB / s.

Specifications for Geforce RTX Super:

	2080 Super	RTX 2080	2070 Super	RTX 2070	2060 Super	RTX 2060
GPU	TU104-450	TU104-400	TU104-410	TU106-400	TU106-410	TU106-200
CUDA cores	3 072 st.	2 944 st.	2 560 st.	2 304 st.	2 176 st.	1 920 st.
Memory amount	8 GB	8 GB	8 GB	8 GB	8 GB	6 GB
Minnesfrekv.	16 000 MHz	14 000 MHz	14 000 MHz	14 000 MHz	14 000 MHz	14 000 MHz
Memory bus	256-bit	256-bit	256-bit	256-bit	256-bit	192-bit
Bandwidth	512 GB/s	448 GB/s	448 GB/s	448 GB/s	448 GB/s	336 GB/s

Igor Wallossek diligently compares the upcoming Nvidia cards with AMD’s recently unveiled Radeon RX 5700 XT and RX 5700, and believes that the Geforce RTX Super variants are designed to counteract these in terms of performance. Whether the same applies to pricing remains to be seen, but Videocardz believes that the cards with the “Super” suffix occupy the same price level as its predecessor, which in turn receive a reduced price tag.

Read This Now: Radeon "Fiji" the only new graphics processor in the upcoming 300 series

However, no concrete prices and launch dates are available, and in addition, clock frequencies and many other specifications are obscured. It is worth keeping in mind that Nvidia has not presented any official details about the graphics cards, and the information should therefore be taken as unconfirmed information.

Source: Videocardz via Igor Wallossek (German)