Following the publication of a leaked presentation slide for the upcoming RTX 20 SUPER refresh for laptops, it turns out that several SKUs have been benchmarked on 3DMark. Spotted by @_rogame, we have already reported on the RTX 2070 SUPER, which appears to have been tested in a laptop also containing an Intel Core i9-10880H processor.
While @_rogame has not clarified what hardware with which the other three SUPER GPUs have been paired, a few specifications for the cards can be ascertained. According to the leaker, NVIDIA is prepping Max-Q versions of the RTX 2070 SUPER and RTX 2080 SUPER, too.
Unsurprisingly, the top-tier model will be the RTX 2080 SUPER, which will feature 8 GB of GDDR6 VRAM clocked at 14 Gbps. The card will apparently run on a 150 W TDP and at a 1,365 MHz base clock. There is no word on boost clocks, though. The same applies to the two Max-Q cards.
Meanwhile, both the RTX 2070 SUPER Max-Q and RTX 2080 SUPER Max-Q will have an 80 W TDP, along with 8 GB of GDDR6 VRAM running at 11 Gbps. However, @_rogame claims that the former will have a 900 MHz base clock, with the latter clocked at 735 MHz.
The three cards apparently achieved the following scores in 3DMark Time Spy, Fire Strike and 3DMark 11 Performance:
GPU | Time Spy | Fire Strike | 3DMark 11 Performance |
---|---|---|---|
RTX 2080 SUPER | 9,861 | 24,113 | 32,821 |
RTX 2080 SUPER Max-Q | 7,938 | 18,871 | 25,712 |
RTX 2070 SUPER Max-Q | 7,336 | N/A | 22,639 |
There are too many unknowns to draw any meaningful insights from the performance of these GPUs. They could well be engineering samples, for example, and we have no details on what CPU and RAM these GPUs have been benchmarked. The leaked presentation slide on which we reported this month suggests that NVIDIA expects the RTX 20 SUPER series to offer moderate gains over the original RTX 20 cards.