Huawei Downplays Chipmaking Power Amid Rising US-China Tech Tensions

Huawei’s founder Ren Zhengfei, has played down the company’s chipmaking capabilities saying the United States

June 11, 2025
Huawei Downplays Chipmaking Power Amid Rising US-China Tech Tensions

Huawei’s founder Ren Zhengfei, has played down the company’s chipmaking capabilities saying the United States is overstating its technological strength as Washington and Beijing engage in tense trade discussions that include key tech export controls. In a rare interview with China’s state run People’s Daily. Ren stated that Huawei’s AI chip, Ascend is “still one generation behind” U.S. technology adding that “the US has exaggerated Huawei’s capabilities we’re not that strong yet.”

The comments arrive at a crucial time when Nvidia CEO Jensen Huang has warned that US restrictions on selling chips to China are giving rise to a “formidable” competitor in Huawei. As Washington bars Nvidia from exporting its most advanced AI chips to China, companies like Huawei are stepping in to fill the gap, increasing their market share and visibility within the Chinese AI landscape.

Huawei has introduced the CloudMatrix 384 AI server as its answer to Nvidia’s dominance. This server links 384 Ascend chips using Huawei’s proprietary optical interconnects to improve bandwidth and processing capacity. Despite this technological advancement, Huawei faces significant challenges, including issues with heat dissipation, weight management, and limited developer adoption.

Although some companies have begun testing the CloudMatrix system, many Chinese AI firms like DeepSeek still rely heavily on Nvidia chips for training large language models (LLMs). Huawei’s chips, while improving, have suffered from technical issues related to distributed computing and training performance. Ren acknowledged these limitations but said Huawei is making progress using cluster computing and stacking technologies to match global standards.

Another major hurdle is the lack of a developer ecosystem comparable to Nvidia’s widely used CUDA platform. Huawei is now investing heavily to overcome these gaps, reportedly spending RMB180 billion ($25 billion) annually on research and development, with RMB60 billion going into basic scientific research that may not yield immediate products but could drive long-term innovation.

Ren also pointed out China’s broader advantages in scaling AI infrastructure, highlighting its world-class power grid and advanced telecom networks. He believes these foundational strengths position China well for AI expansion despite current technology gaps.

With trade talks intensifying and AI taking center stage, Huawei is walking a fine line asserting its potential while tempering expectations, as it seeks to build a more resilient and competitive tech future amid growing global scrutiny.