Another week, another round of RTX 4080 Super rumors and leaks, but this time the rumors have some teeth. The PCI ID Repository, an independent database of hardware device information, has a new entry for the RTX 4080 Super and its GPU codename The first of Ada Lovelace's refreshes appears to use AD103, which is very disappointing .

We first found this news on BenchLife.info (which has further details from Videocardz), but the gist of this news is that someone has obtained PCI ID details. This information has been added to the PCI ID Repository and clearly indicates that the updated card still uses the AD103 GPU.

The version used in the current RTX 4080 is almost a full chip, with only four streaming multiprocessors (SM) disabled; given that the RTX 4080 already has 9,728 shaders on board, an AD103 full die based RTX 4080 Super, which is based on the full AD103 die, will only have 5.3% more shaders. [Unless Nvidia plans to increase the clock speed of the Super version, the performance increase is hardly worth considering. Even if it does launch with a 10% higher clock speed, the more details we see about the RTX 4080 Super, the more likely it is that it will be a disappointment at launch.

GPU fans may well remember the RTX 2080 Super from 2019. That graphics card was only marginally better than its successor, the RTX 2080, with 4% more shaders, 6% better boost clock, and 11% faster VRAM. However, it was available at the same MSRP as the RTX 2080, so while it was a disappointing refresh, it was not necessarily a bad buy.

If that happens (and we won't have to wait very long to find out) and a lot of people pick up the new RTX 4080 on the back of significant discounts, sales of the RTX 4080 Super are likely to be even worse than the previous model.

I had hoped that Nvidia would use a much reduced AD102 (the chip that drives the RTX 4090) for the RTX 4080 Super. However, that idea is becoming more and more unlikely with each passing week. I should really know better by now.