The development of an advanced AI processor by Huawei is reportedly presenting a mounting challenge to Nvidia’s global dominance in the artificial intelligence (AI) chip market. This challenge is particularly pronounced in China, a key market for AI technology. The move by Huawei comes amidst escalating tensions between Washington and Beijing concerning technology transfer and semiconductor manufacturing capabilities. Industry analysts are reportedly monitoring developments closely.
Huawei’s New Ascend Chip
According to recent reports from the Wall Street Journal, the Chinese technology giant, Huawei, is preparing to test its new Ascend 910D chip. Company sources reportedly claim this chip could outperform Nvidia’s flagship H100 processor, a leading AI chip globally. Huawei reportedly expects to receive the first batch of advanced AI processor samples by late May, with testing already underway at select Chinese tech companies. This suggests the new chip is nearing a phase of wider evaluation.
Surprising Manufacturing Strides
The advance in the Ascend chip development follows a trajectory of improvement in Huawei’s chip manufacturing capabilities that has reportedly surprised many Western observers. The Financial Times reported earlier this year that Huawei has made significant strides in production efficiency. The newspaper quoted the Financial Times report as saying, “significantly improved the amount of advanced artificial intelligence chips it can produce.” The report also stated that the manufacturing yield of Huawei’s latest AI chips has reportedly increased to approximately 40% from just 20% a year ago. While this yield remains below the industry standard of about 60% for similar processors, it represents a critical threshold that has reportedly made Huawei’s production line profitable for the first time.
Context of US Export Controls
This manufacturing progress is notable given the extensive U.S. export controls designed to limit China’s access to advanced semiconductor technology. In 2022, the U.S. banned the sale of Nvidia’s H100 chip to Chinese customers before it was even launched, a decision that reportedly created a substantial market opportunity for domestic alternatives within China. This ban included previously allowed chips like the MI308 and H20, reportedly resulting in a $5.5 billion write-off for Nvidia.
Implications for Nvidia’s China Revenue
The implications for Nvidia’s revenue from the China market are reportedly significant. According to Grace Shao’s December 2024 analysis, “Deep Dive: Is Huawei like China’s Nvidia plus Google? ,” China has historically accounted for “approximately 20-25% of Nvidia’s annual sales.” As Chinese companies increasingly adopt domestic alternatives like Huawei’s Ascend chips, Nvidia could potentially see a substantial reduction in revenue from what has been one of its most important markets.
Shao was quoted as stating, “As the chip ban took effect, Chinese AI companies began buying and hoarding inventory. However, Huawei’s launch of the Ascend 910B and then 910C completely disrupted the ecosystem in China and emerged as the most viable alternative to Nvidia’s sought-after A100 chip.” The upcoming 910D model could reportedly accelerate this transition towards domestic alternatives.
Huawei’s Resilience and Self-Sufficiency Push
For Western observers, Huawei’s resilience in the face of U.S. trade restrictions has been remarkable. The company’s surprise launch of the Mate 60 smartphone in 2023, featuring a domestically produced processor, demonstrated its ability to innovate despite export controls. The timing of the launch during then-Commerce Secretary Gina Raimondo’s visit to Beijing was reportedly interpreted as a deliberate show of technological independence. Beyond corporate competition, these developments reflect China’s broader ambition to achieve semiconductor self-sufficiency. As reported by the Financial Times, Beijing is providing “state support, with Beijing urging local tech companies to buy more of Huawei’s AI chips and shift away from $3.3tn US chipmaker Nvidia.”
Technical and Software Hurdles
However, technical challenges reportedly remain for Huawei’s AI chips. According to the Financial Times, “AI companies and Huawei researchers have also said that the Ascend 910B did not work well for large-scale model training because of problems with inter-chip connectivity and memory issues.” Huawei is reportedly working to address these limitations in its newer models. The software ecosystem presents another hurdle: Nvidia’s CUDA platform has established itself as the industry standard for AI development. A person close to Huawei’s business reportedly told the Financial Times that Nvidia’s software is “known for being easier to use and capable of faster data processing than Huawei’s offerings,” posing a challenge for Huawei’s alternatives.
Production Constraints and Integrated Approach
Production capacity also reportedly remains constrained for Huawei’s AI chips. Despite the Financial Times reporting plans to produce 100,000 910C and 300,000 910B chips this year, Huawei reportedly prioritizes supplying large state-run cloud providers like China Mobile, leaving other potential customers struggling to secure supplies. Nevertheless, Huawei’s integrated approach to technology development gives it unique advantages. Shao stated, “Huawei has created a comprehensive ecosystem and long invested in the cloud, hardware, and various touch points with enterprises and consumers (far before AI became hot),” highlighting their existing infrastructure.
Geopolitical Implications and Outlook
The geopolitical significance extends beyond corporate competition. Huawei founder Ren Zhengfei told Chinese President Xi Jinping that worries about a “lack of core and soul” (referencing semiconductors and operating systems) eased, adding, “I firmly believe a greater China will rise faster,” according to People’s Daily, reported by Financial Times. This underscores the national importance attached to self-sufficiency. Huawei’s AI processor development signals acceleration of tech decoupling between China and Western markets.
While Nvidia likely maintains global leadership (North America/Europe), the emergence of viable alternatives in China suggests fragmentation of the AI hardware landscape along geopolitical lines. Commercial success of Ascend 910D will indicate China’s progress. If Huawei delivers on claims and addresses limitations, Western companies may need to reassess strategies/supply chain dependencies in a bifurcated global tech ecosystem.
