The GeForce RTX 4090 D will use the AD102-250 chip instead of the AD102-300-A1 used in the standard RTX 4090 model. This indicates that the new model will have less computing power and artificial intelligence capabilities than the standard version. Typically, in such cases, NVIDIA reduces the frequency or reduces the number of active function blocks.
Despite this, the RTX 4090 D will still retain support for software-level features and other key features of the original model. The new graphics card to be released in China is aimed at the consumer segment and will not be officially used in the field of artificial intelligence.
In this situation, it is interesting that the price of the modified graphics card will not differ from the regular version of the GeForce RTX 4090. In China, you can buy a stripped-down model for the equivalent of $1600.