It seems that green reserves special chips for those graphics.
While working on its new version of the GPU-Z, the Techpowerup site has discovered a phenomenon related to the Nvidia family’s new Turing graphics: each card has an “-A” variant. That means, for example, we have a TU102-400 and a TU102-400-A that match the RTX 2080 Ti. Normally, Nvidia creates only one ASIC code for each product, be it a reference, an edition of Founders, personalized or with factory overclocking. These nomenclatures should basically distinguish professional products from consumer goods. The answer seems to indicate that Nvidia prevents chips from being overclocked by the most basic graphics cards.
In fact, the products with the “-A” would be the models of license plate partners overclocked by the factory, while the models without the “-A” would be those with the recommended price of Nvidia. They are the same chips, but they are specially classified for each feature and price range. That is, this does not prevent users from playing with the frequencies in their home; This only affects the factory configuration of the cards. In the case of “A” graphics, the Nvidia specification states that these cards may have overclocking equal to or higher than their own edition of Founders. This kind of information could be incorporated into our report on what we know about GeForce RTX.