Nvidia’s share of AI chip spending in China slips in 1H 2026 as Huawei narrows gap, Reuters reports

Nvidia’s share of the artificial-intelligence chip market in China declined during the first half of 2026, according to a Reuters report that cited an IDC study. The report attributed the shift to continuing U.S. export controls limiting what advanced AI chips can be shipped into China, alongside rapid progress by Huawei as it works to supply more of the country’s AI infrastructure needs.

The Reuters write-up, published via Yahoo Finance, did not provide additional company-specific detail such as the size of the percentage point decline, the chip categories in question, or which particular Nvidia product lines were most affected. It also did not spell out whether the change reflected slower demand for the newest Nvidia accelerators, a change in pricing, or redistribution of buyers toward alternative suppliers.

Reuters said Huawei’s position in the market improved over the period, framing the development as a “catch-up” dynamic. In practical terms, that means Chinese customers that had relied heavily on Nvidia for training and inference hardware faced more competition from domestically developed solutions, particularly for workloads where buyers may be seeking supply continuity under export restrictions.

U.S. export controls have been a key structural factor shaping the global AI supply chain for years. For a supplier like Nvidia, the controls can constrain the range and performance of chips that can legally be sold into China, which in turn can affect customer roadmaps for building data-center capacity. Even when shipments remain possible, the economics and performance trade-offs can influence procurement decisions toward alternative vendors.

Huawei’s progress, as described by Reuters, underscores how competition in China’s AI hardware market is not just about performance. It also involves integration with local software stacks, ability to scale manufacturing and delivery, and readiness to support enterprise deployments under regulatory and logistics constraints.

Nvidia, for its part, has historically positioned its data center GPUs as foundational accelerators for training and inference, with an ecosystem that includes software tooling and networking. Reuters’ report points to a competitive environment in which that ecosystem may have to contend with a growing number of locally supported alternatives, especially in segments constrained by the export regime.

A limitation for investors and industry watchers is what the cited IDC-based coverage does not disclose. The Reuters report does not provide the underlying IDC methodology, the exact market definition used for “AI chips,” or the specific numbers needed to judge how much share loss is attributable to export controls versus competitor performance, pricing, or customer switching behavior. Without those figures, it is difficult to measure whether Nvidia is losing share across the board or only in particular subcategories.

What to watch next is whether Nvidia updates guidance or commentary around China-related demand and how quickly alternative accelerators from Huawei and others capture new deployments. In parallel, attention is likely to remain on whether export-control rules tighten or loosen, and on whether Nvidia can adjust its product mix or customer support to sustain competitiveness under those constraints.