Impact of President Trump’s AI Chip Export Controls on China’s DeepSeek and Their Upcoming “R2” AI Model Release

DeepSeek has encountered significant obstacles in securing NVIDIA’s state-of-the-art AI chips, as highlighted by a recent report indicating that U. S.export controls are hampering China’s ambitions in the AI sector.

Impact of U. S.Export Controls on AI Chip Availability for China and DeepSeek’s R2 Release

Since the Trump administration took office, a primary agenda has been to maintain U. S.leadership in the competitive AI chip landscape. This objective prompted the swift implementation of export restrictions on some of NVIDIA’s highly sought-after products, including the H20 AI accelerator. The ramifications of these controls have adversely affected DeepSeek, which is grappling with the performance of its upcoming R2 AI model. A report from The Information notes that CEO Liang Wenfeng has expressed dissatisfaction with the current capabilities of DeepSeek’s large language model (LLM), hampering the readiness for broader adoption among Chinese cloud service providers (CSPs).

Despite efforts by Chinese entities to circumvent these restrictions through various methods, the latest export controls have markedly curtailed the influx of advanced AI accelerators into mainland China, causing disruptions for DeepSeek. Following the success of its previous model, the DeepSeek R1—which notably contributed to substantial market losses for NVIDIA upon speculation about its resource optimization—anticipation surrounding the R2 model was immense.

NVIDIA's AI Chip Renting Services In China Are Much Cheaper Compared To The US, For As Little As $6/Hr 1

According to reports, DeepSeek is currently unable to establish a clear timeline for the R2 model’s rollout in China. This uncertainty stems from domestic CSPs struggling to acquire sufficient computing resources to support the new model. A considerable number of NVIDIA H20 AI GPUs within the domestic market are reportedly allocated to customers operating the R1 model, further complicating the potential deployment of R2 due to limited availability of accelerators. It is evident that U. S.export controls are effectively limiting China’s capacity for AI growth.

In previous disclosures, it was suggested that DeepSeek had connections with the Chinese military and was attempting to acquire AI chips through “shell companies”across Asia, underscoring the lengths to which the firm is willing to go to access NVIDIA’s premium AI technology. However, they continue to face hurdles in this endeavor. NVIDIA has yet to provide a viable resolution for the domestic market, and with alternatives like Huawei, DeepSeek finds itself unable to depend solely on these options for large-scale deployments.

Source&Images