Huawei AI Technology Push Gains Momentum in China

Huawei Technologies is strengthening its position in China’s technology market as higher revenue, growing demand for artificial intelligence equipment and continued investment in domestic computing infrastructure support the company’s broader recovery.

The Chinese technology group reported first-half 2026 revenue of 483.6 billion yuan, up 9.6% from the same period a year earlier. However, net profit fell 36% to 23.81 billion yuan as higher input costs, increased research and development spending and inventory investment weighed on earnings.

The contrasting revenue and profit figures show that Huawei’s expansion is requiring substantial investment.

AI demand is becoming a major growth driver

Huawei’s artificial intelligence business has emerged as an important part of the company’s technology strategy.

At Huawei Connect in Shanghai this month, rotating chairman Eric Xu said the company’s AI computing equipment was in such strong demand in China that Huawei could not currently produce enough systems to meet domestic requirements.

Xu said the supply constraint was limiting Huawei’s ability to expand international sales on a large scale.

The comments indicate that China’s demand for domestically produced AI computing equipment is becoming an important market for Huawei.

They also highlight a strategic opportunity for the company as Chinese technology firms seek alternatives to advanced processors supplied by overseas companies.

Huawei is accelerating its AI chip roadmap

Huawei has announced plans to release new generations of Ascend AI processors as it expands its domestic computing ecosystem.

The company plans to launch the Ascend 960DT in the first quarter of 2027 and the Ascend 960PR in the third quarter.

Huawei has also said it intends to continue developing new generations of AI processors on a regular cycle.

The strategy is intended to improve both individual chip performance and the ability to combine large numbers of processors into high-performance computing systems.

Huawei says its UnifiedBus technology will allow large numbers of processors to be connected within next-generation AI computing systems.

Large computing systems are central to the strategy

Huawei is increasingly focusing on system-level computing rather than relying only on the performance of individual processors.

At Huawei Connect, the company introduced its Atlas 960E SuperPoD, which it describes as a system designed to accelerate training and inference for very large AI models.

Huawei says the system can integrate up to 4,096 NPUs and is designed for models with as many as 10 trillion parameters.

The company has also introduced computing architectures intended to connect processors across larger clusters.

Huawei says its UnifiedBus architecture can support scaling from individual cabinets to much larger AI computing clusters.

This approach reflects a broader industry trend in which AI performance increasingly depends on the ability to connect processors, memory, storage and networking infrastructure efficiently.

Demand currently exceeds supply

Huawei’s AI ambitions are being supported by strong domestic demand.

Xu said the company did not have sufficient production capacity even to satisfy Chinese demand for its AI computing products.

Huawei therefore does not currently plan to expand fully into international markets with those products, although Xu said the company continues to supply selected countries where demand is particularly strong.

The situation creates both an opportunity and a constraint.

Strong demand can support revenue growth, but insufficient manufacturing capacity limits how quickly Huawei can convert that demand into additional sales.

U.S. export controls have changed China’s AI market

Huawei’s growing role in AI computing is also linked to restrictions on China’s access to advanced foreign semiconductor technology.

U.S. export controls have limited Chinese companies’ access to some advanced Nvidia processors and related technologies.

The restrictions have encouraged Chinese technology companies to develop domestic alternatives and have increased the strategic importance of companies such as Huawei.

Reuters reported that Huawei is positioning its Ascend product line as part of China’s effort to build a more self-reliant AI computing industry.

The development does not mean that Huawei has eliminated the technological gap with leading global chip suppliers.

Nvidia continues to have significant advantages, particularly in software and its CUDA ecosystem, while Chinese companies face constraints involving manufacturing, advanced memory and semiconductor supply chains.

Memory shortages are raising costs

Huawei’s AI expansion is also facing a major supply-chain challenge.

High-bandwidth memory, or HBM, is an important component of modern AI accelerators because it allows processors to access large volumes of data rapidly.

Reuters reported earlier this month that Huawei and other Chinese AI chipmakers had raised prices for current and next-generation processors as HBM costs increased.

Sources told Reuters that Huawei’s indicated price for its Ascend 950DT accelerator had risen above 250,000 yuan, with the increase ranging from 20% to 50% depending on contract terms.

The higher costs illustrate one of the challenges facing China’s effort to expand domestic AI computing.

Developing a competitive processor is only part of the process. Manufacturers also need access to advanced memory, packaging, manufacturing capacity and other components.

Huawei is investing heavily in research and development

The company’s financial results show how much Huawei is spending to maintain its technology push.

Huawei’s first-half research and development spending increased 25% year-on-year, according to Reuters.

The higher spending contributed to the sharp decline in net profit despite the increase in revenue.

Huawei also increased purchases and inventory during the period, while its cash flow turned negative.

That suggests the company’s strategy involves substantial upfront investment as it expands its technology capabilities and prepares for future demand.

AI infrastructure extends beyond chips

Huawei’s strategy is broader than producing AI processors.

At Huawei Connect 2026, the company introduced new cloud services, storage systems, networking technologies and enterprise AI products designed to support large-scale AI deployment.

Huawei Cloud said its latest AI Cluster Service is intended to provide infrastructure for agentic AI, while its Agentic Model as a Service platform is designed to allow different AI models to be deployed as services.

The company also introduced new storage technology designed to support AI inference in hyperscale data centres.

Huawei’s approach is therefore increasingly focused on supplying multiple layers of the AI infrastructure stack.

China’s AI ecosystem is expanding

Huawei’s developments form part of a broader expansion of China’s domestic AI industry.

Alibaba announced this week that it was developing a next-generation AI model with between 5 trillion and 10 trillion parameters while also introducing a new AI chip through its T-Head semiconductor unit.

Alibaba has also set a target for its cloud data-centre capacity to exceed 20 gigawatts by 2032.

Other Chinese semiconductor companies are similarly expanding.

The competition is creating a growing domestic ecosystem involving chip designers, memory manufacturers, cloud providers, AI developers and data-centre operators.

Huawei’s smartphone business remains part of the picture

Huawei’s consumer technology business continues to be an important part of the company’s broader technology presence.

Earlier this month, Huawei launched its new Mate XT2 series of tri-fold smartphones.

The devices use Huawei’s Kirin 9050 Pro processor and were launched with prices ranging from 19,999 yuan to 24,999 yuan.

Huawei was also reported to hold 22.6% of China’s overall smartphone market in the cited data, while maintaining a particularly strong position in the country’s foldable-phone segment.

These figures illustrate that Huawei’s recovery is not limited to one technology category.

However, the company’s growing emphasis on AI computing indicates that enterprise infrastructure and artificial intelligence are becoming increasingly important to its long-term strategy.

Revenue growth comes with higher costs

Huawei’s latest financial results demonstrate the trade-off involved in that strategy.

Revenue increased, but profitability declined sharply.

Higher research spending, input costs and inventory investment can reduce near-term earnings even when a company is expanding its sales and technology capabilities.

For Huawei, those costs are occurring while the company is attempting to develop an increasingly complete domestic technology ecosystem.

The next stage will depend on capacity

Huawei’s current AI position gives it access to a large and expanding domestic market, but production capacity remains a constraint.

The company says demand for its AI computing systems already exceeds its ability to supply the Chinese market.

Its response is to accelerate new chip generations, expand large-scale computing architectures and develop the software, networking, storage and cloud infrastructure required to use those processors.

Whether that strategy produces sustained financial growth will depend on Huawei’s ability to increase production, manage component costs and continue improving the performance of its technology.

For now, the company’s latest figures show a technology group growing revenue while spending heavily to build capabilities for the next phase of China’s AI industry.