Fans of gaming graphics cards from Nvidia froze in anticipation. On August 20 the company should present and open a pre-order for a new generation of graphic cards. Recently, a lot of different rumors have leaked to the web regarding a new family of products, some of which contradict each other.
According to the latest information provided by the portal Tech Power Up, the model 2060 belonging to the mid-price segment will not use the usual prefix GTX in favor of a new RTX. This involves using the latest ray tracing technology. Also, the source shared more detailed characteristics of the video card. The GTX 2060 is manufactured using 12-nm process technology. The operation of the device is provided by the graphics processor TU106, operating at a frequency of 1.4 GHz to 1.5 GHz. The type of GDDR6 memory installed with a capacity of 8 GB (previously reported 5 GB), running at a clock speed of 1.7 GHz. It should be noted that due to the results of tests given by the resource 3DMark, there is a probability of performance increase to the level of the current flagship, represented by GTX 1080.
Nvidia GeForce RTX 2070 acts as the senior model of medium-budget solutions. The graphics card, built on the Turing architecture and developed using a 12-nm process, has at its disposal A tu104 processor, clocked at 1.5 GHz in normal operation, and 1.6 GHz overclocking. Memory type GDDR6, operates at a frequency of 1.7 GHz and has a memory capacity of 8Gb. Bandwidth is 383 GB/s. Estimated time of entering the market, according to various rumors, is the end of October. Accurate information about the price is not available at the moment, but it is expected the amount of $ 400 to $ 500.
Mostly a large number of rumors has gone around flagship models. RTX 2080 and RTX 2080 Ti are the most anticipated devices that use Nvidia Turing. RTX 2080 Ti is inferior in performance to the professional products Quadro RTX 8000 and Quadro RTX 6000. This statement is due to the fact that the game card uses 4352 CUDA cores, against 4608 cores in the Quadro series. Memory type GDDR6 boasts a volume of 11 GB and a frequency of 1.7 GHz. TU102 GPU operates at a frequency from a base 1.3 GHz to 1.5 GHz in turbo mode. RTX 2080 is equipped with a graphics chip TU104, operating at a frequency of 1.5 GHz to 1.6 GHz. The bandwidth of the graphics card has reached 352 GB/s, which exceeds RTX 2070, but still does not reach 616 GB / s in 2080 Ti. The official cost will be known later. But according to some reports, the price reaches $ 800 for RTX 2080.
All the graphics cards are connected via PCIe 3.0 x16. Also all maps have all the necessary types of interfaces: USB 3.0 Type — C (1), HDMI (1) DisplayPort (3).
More information will be available after the presentation of the devices.
Video card          Graphics processor   Memory type   Amount of memory   Clock frequency     CUDA cores
RTX/GTX 2060             TU106                  GDDR6                   8Gb                          1.7Ghz                    1536
RTX 2070                      TU104                  GDDR6                   8Gb                          1.7Ghz                    2304
RTX 2080                      TU104                  GDDR6                   8Gb                          1.7Ghz                    2944
RTX 2080 Ti                  TU102                  GDDR6                  11Gb                         1.7Ghz                    4352

20 August Nvidia has announced prices for the new cards

Video card          Graphic card power   Card price
RTX/GTX 2060              ???                          ???
RTX 2070                        175W                   599USD
RTX 2080                       215W                    799USD
RTX 2080 Ti                  250W                  1199USD

Аs we can see, the prices are very high

https://www.techpowerup.com/