In a bold move that speaks to China’s determination to overcome US technology restrictions, Huawei is preparing to test its most powerful AI processor yet – the Ascend 910D, with customer shipping due as early as May.
The Chinese tech giant is approaching major domestic tech firms to evaluate this new chip, which it claims will outperform NVIDIA’s H100; the industry-leading AI processor that Chinese companies are now barred from purchasing.
The timing is strategically significant. NVIDIA recently disclosed a staggering $5.5 billion financial hit due to Trump administration export controls effectively banning sales of its H20 chip – specifically designed for the Chinese market – without special licenses that are unlikely to be granted in the current geopolitical climate.
Technological Warfare Intensifies
Huawei’s resilience against US sanctions continues to surprise Western observers.
Despite being on a US trade blacklist for nearly six years, the company shocked Washington in 2023 by launching its high-end Mate 60 smartphone with domestically produced processors during then-Commerce Secretary Gina Raimondo’s visit to Beijing.
Now, as NVIDIA shifts manufacturing to American soil with a massive $500 billion commitment to build AI infrastructure in the United States over the next four years, Huawei is countering with its own technological push.
Jensen Huang, NVIDIA’s founder and CEO recently explained:
The engines of the world’s AI infrastructure are being built in the United States for the first time. Adding American manufacturing helps us better meet the incredible and growing demand for AI chips and supercomputers, strengthens our supply chain and boosts our resiliency.
The company is expected to receive the first batch of Ascend 910D samples as early as late May, alongside reports that it plans to begin mass shipments of its current-generation Ascend 910C AI chip to Chinese customers next month.
An Increasingly Siloed AI Landscape
For international enterprises already struggling with AI hardware availability, these developments signal a fragmenting global technology ecosystem. Organizations pursuing worldwide AI initiatives now face different hardware environments across geopolitical boundaries, potentially complicating deployment strategies.
The restrictions extend beyond NVIDIA.
Advanced Micro Devices expects to take charges up to $800 million related to Chinese AI chip sales, while Intel has reportedly informed Chinese clients it will require licenses to sell certain advanced AI processors.
Strategic Implications
Huawei’s push into high-end AI chips is more than just corporate competition – it’s a cornerstone of Beijing’s strategy to cultivate a self-sufficient semiconductor industry. Despite lacking access to some advanced chip-making equipment, Huawei continues developing alternatives to NVIDIA’s products, including its CloudMatrix AI technology designed to rival NVIDIA’s NVL72.
According to industry sources, Huawei aims to produce approximately 750,000 AI chips this year despite US restrictions, potentially testing a China-made extreme ultraviolet (EUV) lithography machine – technology critical for manufacturing cutting-edge semiconductors that has been tightly controlled by the West.
As the silicon curtain between East and West becomes increasingly defined, the AI chip battle may determine which technological ecosystem gains the upper hand in the next generation of artificial intelligence development.
Consumer Impact and Future Outlook
The ripple effects of this technological cold war extend beyond enterprise markets. As AI becomes increasingly embedded in consumer products, from smartphones to smart home devices, the divergence in underlying chip technologies could create two distinct AI experiences globally.
Chinese tech firms have already begun adapting their AI models to run efficiently on Huawei’s Ascend architecture rather than NVIDIA’s CUDA platform.
Meanwhile, rumors suggest Huawei is exploring partnerships with domestic chip foundry SMIC to enhance production capabilities, potentially accelerating China’s timeline toward semiconductor independence.
