Along with the official introduction of the GeForce RTX 2080 Super, the first super-family leader reviews were made online. As announced, the new product has about 5-10% better results than last year's GeForce RTX 2080. This was achieved using the full version of the Nvidia TU104 core as well as faster GDDR6 memory chips.
GeForce RTX 2080 Super uses the Samsung K4Z80325BC-HC16 chipset with 16 Gbit / s capacity per contact. In the case of a new card, their potential is not fully disclosed, as Nvidia decided to limit the effective speed of the video buffer to 15.5 Gbit / s.
According to the online edition of AnandTech, this was done in order to simplify the issue of new graphics cards for the AIB partners of the "green" chip maker. The vast majority of GeForce RTX 2080 Super models use the same PCB as for similar versions of the GeForce RTX 2080, which were equipped with less powerful GDDR6 chips. If Nvidia decided not to reduce the effective speed of the video buffer, it is likely that some alternative design solutions would not end the stability test.
On the other hand, GeForce RTX 2080 Super with well-designed printed circuits can count on guaranteed overclocking of GDDR6 memory at 16 GHz or higher.