GeForce RTX 2070 Super & RTX 2080 Super Performance Review

Users shouldn’t expect any significant improvements from the Super chips. Although Nvidia has made a few tiny changes compared to the standard versions, the most important specifications and features are largely the same or at least very similar to previous chips. While the RTX 2070 Super has  256 more shader units (2,560 vs 2,304 @RTX 2070), the RTX 2080 Super only has 128 more CUDA cores (3,072 vs 2,944 @RTX 2080). As you can see in the table below, the clock speeds have even been slightly reduced. The reduction primarily affects the RTX 2070 Super. The GDDR6 video memory is still 8 GB in size and connected via a 256-bit interface.

Important note: Similar to the last generation, there are several TDP versions of the RTX 2080 Super (more information can be found on the GPU page), some of which perform notably different. Depending on the laptop, the Super version can be slower than the normal RTX 2080. You can see this very clearly in the gaming benchmarks using the MSI GT76 (200 watts vs. 150 watts TDP). The results of the XMG Ultra 15, which appear in the benchmarks below, are a nice comparison because its GPU has an identical TDP to the RTX 2080 inside of the MSI GT76. 

The results for the new Super Turing GPUs were, in our opinion, not very super. While the RTX 2070 Super is, in ideal circumstances, up to 20% better than the RTX 2070 (though 10% is more realistic), the RTX 2080 Super can only set its performance apart from the RTX 2080 by a maximum of 10% when using models that have the same TDP. Unfortunately, very few laptop manufacturers state the TDP version used in their products, so customers will normally be in the dark about this. In any case, the TDP has a greater influence on the speed of the RTX 2080 than the type of GPU (Super vs. non-Super), which is somewhat disappointing in our eyes.