On May 21st, reports emerged indicating that during NVIDIA CEO Jensen Huang's visit to China, Chinese customs imposed an import ban on the RTX 5090D v2, a flagship gaming graphics card tailored specifically for the Chinese market. This graphics card represents a second-tier, downgraded version developed by NVIDIA in response to U.S. export restrictions. It features a reduction in memory from 32GB to 24GB, a narrowing of the memory interface width from 512-bit to 384-bit, and a 25% decrease in bandwidth. Despite these downgrades, the price remains steadfast at 16,499 yuan. The ban, initiated by China, caught NVIDIA by surprise. Prior to this, Huang's visit was anticipated by external observers to potentially alleviate the impasse in Sino-U.S. chip trade, especially with expectations that restrictions on the sale of the H200 AI chip to China might be eased. However, the outcome proved contrary: not only did the H200 chip fail to secure approval for purchase in China, but even the consumer-grade gaming graphics card, the RTX 5090D v2, was prohibited from sale. Industry insiders opine that this sales prohibition is not linked to AI computing power regulations, given that NVIDIA has proactively limited the AI computing capabilities of the RTX 5090D v2. Instead, this move seems to signal China's refusal to accept downgraded, differential treatment products from the U.S., while simultaneously firmly advancing the substitution of domestic chips.