Recently, NVIDIA's professional-grade graphics card RTX 6000D for the Chinese market has become the focus of attention in the tech industry.

This GPU, based on the Blackwell architecture, was originally designed as a "special version" to cope with U.S. export restrictions. However, newly exposed test data and market feedback show that both its specifications and performance have been significantly reduced, resulting in low market acceptance.

This situation not only reveals the profound impact of the China-U.S. tech war on the hardware industry but also reflects the challenges NVIDIA faces in its strategy in China.

The RTX 6000D has recently appeared on the Geekbench testing platform for the first time. Its name no longer carries the "PRO" identifier, which is similar to the naming convention of the previous generation of workstation graphics cards.

According to the test records, this graphics card integrates 156 SM units, equivalent to 19,968 CUDA cores. Compared with the standard RTX PRO 6000, which has 188 SM units (24,064 CUDA cores), this represents a reduction of approximately 17%.

In addition, the video memory capacity has decreased from 96GB in the standard version to 84GB (tests show 83GB, and the industry speculates this is a display error). The memory bus width has also been reduced from 512 bits to 448 bits, resulting in a bandwidth drop from 1,792 GB/s to 1,568 GB/s.

The core frequency is also limited to 2430 MHz, which is lower than the standard version's 2600 MHz and above.

These adjustments comply with the restrictions imposed by the United States, aiming to control computing power output, but at the same time, they have significantly reduced the hardware potential of the product.

In the OpenCL benchmark test of Geekbench 6.5, the RTX 6000D scored only 390,656 points on a test platform based on the AMD EPYC 9654 Rocky Linux 9.4 system.

This result is significantly lower than the 450,000 to 500,000 points achieved by the full-spec version, with a performance gap of about 20%.

The benchmark data further confirms the impact of the reduced specifications on real-world applications, significantly diminishing the RTX 6000D's competitiveness in professional scenarios such as AI training and graphics rendering.

Although NVIDIA has high hopes for the RTX 6000D, the market response has been very cold.

It is reported that several major Chinese technology giants have informed NVIDIA of their lack of intention to place orders, mainly due to the product's low cost performance.

After significant specification reductions, the RTX 6000D still has a high price. Coupled with the risks brought by the uncertainty of U.S. policies, customers' willingness to purchase continues to decline.

This situation has disrupted NVIDIA's layout in the Chinese market, forcing it to re-evaluate the planning of its workstation product line.

The original launch of the RTX 6000D stemmed from the U.S. export controls on high-end AI hardware, aiming to prevent advanced technologies from flowing into China.

However, the policy environment may change. Foreign media reports state that the U.S. government is initially negotiating to allow NVIDIA to export the AI chip H200 to China.

Although the U.S. approved the export of H20 in early 2025, China responded calmly due to security considerations. Analysts point out that China is currently more focused on acquiring advanced chip manufacturing equipment to expand domestic production capacity rather than purchasing a single type of chip.

A research report from Jefferies Group emphasizes that China's process of technological independence may be disrupted as a result, but in the long run, the localization strategy remains the focus.

As more test data emerges, the true performance and market positioning of the RTX 6000D will gradually become clear. However, its current lukewarm reception in the market has already served as a wake-up call for NVIDIA.

Under the dual pressures of external policy restrictions and intensified internal competition, NVIDIA not only needs to deal with the constraints of geopolitics on its product strategies but also face the strong advances from competitors like Google's TPU in the AI acceleration market.

How to find a dynamic balance between global layout and localized compliance, while solidifying its dominant position in the AI hardware ecosystem, is becoming the key to NVIDIA's future competition.