Huawei’s chips are a generation behind those of the Americans

Ren Zhengfei, Huawei

In a rare and candid interview with China’s People’s Daily, Huawei CEO Ren Zhengfei (pictured) acknowledged what industry insiders have long suspected: the company’s chips remain a generation behind their U.S. counterparts. But according to Ren, that hasn’t stopped Huawei from finding creative ways to stay competitive — and in some cases, even leap ahead.

Ren’s remarks, reported Tuesday by Chinese state media, mark the first time Huawei’s leadership has publicly commented on their advanced chipmaking efforts. It’s a topic loaded with geopolitical weight as U.S.-China tech tensions continue to simmer, with trade officials from both nations currently engaged in talks in London.

A Workaround Culture Powered by Cluster Computing

Despite facing relentless U.S. export restrictions since 2019 — measures designed to stall China’s technological and military rise — Huawei has invested heavily in R&D, pouring a staggering 180 billion yuan ($25 billion) annually into research. A third of that, according to Ren, goes toward theoretical studies, underscoring the company’s belief that “without theory, there will be no breakthroughs, and we will not catch up with the United States.”

Ren detailed how Huawei is supplementing the limitations of its single-chip technology by embracing cluster computing — essentially linking hundreds of less-powerful chips together to perform like a high-end chip. “We use mathematics to supplement physics, non-Moore’s law to supplement Moore’s law,” Ren explained, referring to the decades-old principle predicting the rapid advancement of chip technology.

It’s an admission of where Huawei stands, but also a message of defiance. “There’s no need to worry about the chip problem,” Ren insisted. “Software is not a bottleneck for us.”

AI Chips: Closing the Gap with Nvidia

Perhaps the most significant development has been in AI computing. Huawei’s Ascend series of AI chips have gained serious traction in China, stepping into a space once dominated by Nvidia, the global leader in AI chips. While Nvidia’s products are still considered more powerful, U.S. restrictions have barred the company from selling its most advanced chips to China, creating an opening for Huawei.

In April, Huawei unveiled AI CloudMatrix 384, a high-performance computing system that connects 384 Ascend 910C chips. Designed for training AI models, the system reportedly outperforms Nvidia’s highly-touted GB200 NVL72 system on select benchmarks.

Semiconductor analyst Dylan Patel of SemiAnalysis even suggested last month that China, led by Huawei, now possesses AI system capabilities capable of surpassing Nvidia’s — a bold claim that signals just how seriously Huawei is pushing its AI ambitions.

A Measured Humility, or Strategic Understatement?

Interestingly, Ren struck a tone of humility in his interview, claiming the U.S. has “exaggerated Huawei’s achievements.” He positioned Huawei as just one player among many in China’s semiconductor space, though the company’s recent performance suggests it’s anything but ordinary.

“Our single chip is still behind the U.S. by a generation,” Ren admitted. But in the same breath, he outlined how Huawei’s unconventional strategies are already achieving practical, real-world results.

The timing of Ren’s comments is no coincidence. With U.S. and Chinese trade officials deep in negotiations, technology restrictions — especially around semiconductors and AI — remain a central issue. The People’s Daily article, prominently featured on its front page, sends a clear message: Huawei isn’t backing down, and China’s tech ambitions are alive and well.

Source: Reuters, June 2025.

China’s AI Chip Race Heats Up

It’s worth noting that while Huawei is drawing most of the headlines, it’s far from alone. Chinese firms like SMIC, Alibaba’s T-Head, and Cambricon are also aggressively developing AI chips and chipmaking technologies. According to a recent report from Counterpoint Research, China’s domestic AI chip market is projected to grow by over 50% annually, reaching $7.6 billion by 2027. In other words, the AI chip race is just getting started — and while Huawei might still be a step behind today, it’s assembling the tools, talent, and infrastructure to catch up fast.