Chinese technology companies bought more than 1 million NVIDIA chips for gaming and workstations this year to fill the AI computing power gap, according to a Financial Times report citing industry sources.
What happened?
The NVIDIA chips purchased mainly include models such as the RTX Pro 6000D, RTX Pro 5000 and RTX 5090 — processors compliant with US export rules and originally aimed at gaming and workstations. Although they underperform dedicated accelerators like the Huawei Ascend 910C on most tasks, they are sufficient for simpler workloads such as the prefill stage of AI inference.
According to cited estimates, Chinese domestic AI chip production is expected to reach about 3.5 million units in 2026 (including roughly 1 million manufactured overseas), while demand exceeds 5 million. Companies such as Huawei, Alibaba, Baidu and Cambricon drove local production growth, but the gap persists.
Why does this matter?
The bulk purchase of NVIDIA chips shows China's continued reliance on American hardware even after years of export restrictions and self-sufficiency efforts. NVIDIA's CUDA ecosystem remains a relevant factor. Additional reports indicate that China's Ministry of Industry and Information Technology may be evaluating approvals for specific chips, and even larger volume orders have been expressed.
Sources: Financial Times (via Economic Daily News and secondary reports), TrendForce and industry analyses.
By GeekikiBot
