The overlock team of the Brazilian Teclab magazine has managed to transfer the memory chips of the RTX 2080 SUPER to an RTX 2080 Ti, actually making one RTX 2080 Ti SUPER.
The RTX 2080 Ti is equipped with 11 GB of GDDR6 memory with a frequency of 1750 MHz (14,000 MHz effective). The RTX 2080 SUPER, on the other hand, has 8 GB of GDDR6 memory certified for 2000 MHz (16,000 MHz) which however operate at 1937 MHz (15,496 effective MHz). Basically, this is a 10.7% difference in memory speeds between the two graphics cards.
There is a difference in the total memory capacity between the RTX 2080 Ti and the RTX 2080 SUPER of 3GB, so Techlab had to dismantle two RTX 2080 SUPER instead of one. The team did not use reference models, but rather expensive custom solutions from Galax.
Even more impressive is the “surgical” procedure with which they extracted the memory from the RTX 2080 SUPER, which they then inserted, chip after chip, on the RTX 2080 Ti. An operation that was successful, with the graphics card BIOS that accepted the new memory with open arms and made it work at 1,750 MHz without requiring changes.
In the end, the overclocker team was able to push the memory to 2150 MHz (17,200 effective MHz), an increase of 22.9% and 10.7% respectively compared to the RTX 2080 Ti and the RTX 2080 with standard memories.
Unfortunately, Teclab did not fully test its “RTX 2080 Ti SUPER”, but only carried out a test with Unigine Superposition, in which the graphics card scored 11,460 points with the 1080p Extreme preset. A standard RTX 2080 Ti typically stops between 8600 and 9200 points.