NVIDIA could well relaunch the “Super” name for its RTX 40 series GPUs. An initiative by the company which shows that it wants to compete with AMD’s advances at all costs.
Competition is ever fiercer between AMD and NVIDIA. If the latter largely dominated the market for a long time, the situation changed with the arrival of AMD GPUs based on the RDNA and RDNA 2 architectures. AMD was largely able to make up for this slight performance delay from the arrival of the cards graphics of the RX 6000 series. All while keeping prices much more attractive for consumers, where prices on the NVIDIA side always tend to be near the heights. The return of the Super name could well make a comeback and put NVIDIA in a more competitive position against its eternal competitor.
The return of Super cards: a strategic necessity?
This information would have been disclosed by a rather serious leaker in the industry: hongxing2020. NVIDIA would therefore prepare three Super versions of already existing cards: the RTX 4070, the RTX 4070 Ti and the RTX 4080. A name that we had not seen appear for a while since the last range of graphics cards bearing this label was the RTX 20 series. One question remains: what should we expect from the innovations brought by these three new cards?
Currently, the RTX 4080 utilizes almost the entire power of the AD103 GPU, as 76 of its 80 stream multiprocessors (SMs) are fully utilized. Suffice it to say that the room for improvement is more than narrow. If rumors are to be believed, the 4080 Super will potentially use AD102 silicon. However, this is not even fully used by the powerful RTX 4090. If the rumor is confirmed, the performance gain for the 4080 Super would be rather interesting compared to its little sister.
The issue of naming and performance
If you find that there is a little confusion among the names, this is quite normal, since these new names could cause confusion. The example of the future 4070 Ti Super perfectly illustrates this observation. Technically, the classic version of the 4070 Ti already uses the entire AD104 GPU. If NVIDIA wishes to upgrade it, this would mean that the Super will have to switch to the AD103 GPU, already used by the 4080. At that time, the room for development is greater, since the 4070 Ti will go from 60 SMs to 76 permitted by AD103. For the 4070 Super, the AD104 could be kept, but it will then have to optimize the use it makes of it. Among the three cards, it therefore seems to be the most complicated to promote and justify to the consumer.
NVIDIA is now forced to strengthen its range in the face of the rise in power of AMD GPUs, which offer very interesting performances at ever more competitive prices. For example, the RX 7800 It remains to be seen if these three new Super cards will one day be released and if they will convince users.
Source : Digital Trends
2