The RTX 4090D has fewer CUDA cores than its RTX 4090 counterpart, at 14,592 versus 16,384, and it also has a slightly lower power draw at 425 watts instead of 450 watts, although most other specifications remain the same between the two versions of the chips. But the RTX 4090D is still "about 5% slower in gaming and creativity," Nvidia spokesman Benjamin Berraondo said in an emailed statement to The Verge.
Nvidia launched the graphics processing unit after the United States tightened export restrictions on high-end chips shipped to China. The new restrictions not only prevent Nvidia from selling the less powerful H800 and A800 AI GPUs that it developed for the Chinese market, but also prevent the company from selling the RTX. 4090 in the country.
The GeForce RTX 4090 D was designed to fully comply with US government export controls, Berraondo said, adding that the company engaged extensively with the US government when developing the chip.
The RTX 4090D will be available in China starting in January for 12,999 yen (~$1,836 USD). Its release may help calm the growing demand for powerful graphics cards in China, which has prompted some factories in the country to dismantle and reuse the banned RTX 4090 for artificial intelligence.