Nvidia H20
AI accelerator chip for the Chinese market
The Nvidia H20 is a graphics processing unit designed for artificial intelligence workloads. It was created to comply with United States export regulations for sale in China. Nvidia resumed applying for licenses to sell this chip in mid-2025 after a period of restricted access.
- Nvidia resumed H20 sales applications in July 2025
- Chip designed for compliance with US export rules
- Targets AI workloads in the Chinese market
- Nvidia market share in China fell to 50 percent
What it is
The Nvidia H20 is a specific model of graphics processing unit intended for artificial intelligence tasks. It belongs to the company's broader hardware lineup but was tailored for a particular regulatory environment. The chip serves data centres and enterprises that require accelerated computing power for machine learning models. Its primary distinction lies in its compliance with international trade rules rather than unique architectural features described in public records.
How it came about
Nvidia developed this processor to maintain its presence in the Chinese semiconductor market. United States export controls limited the sale of high-performance AI chips to the region. The company adjusted its product offerings to meet these legal requirements. In July 2025, Nvidia announced it would resume sales after receiving assurances from the US government regarding license approvals. This move followed discussions between company leadership and officials in both Washington and Beijing.
Inside the chip
The H20 functions as a graphics processing unit optimized for AI workloads. While specific technical details such as core counts or memory bandwidth are not provided in the available records, it operates within the ecosystem of Nvidia's AI software stack. The design prioritizes compliance with export restrictions while delivering sufficient performance for digital twin applications and smart factory logistics. It supports the foundation models and open-source research tools promoted by the manufacturer.
How it was made
Information regarding the specific manufacturing process node, transistor count, or fabrication facility is not available in the provided sources. The chip is produced by Nvidia, though the exact foundry partner and physical packaging details remain unstated in the current material. The focus of public reporting has been on its regulatory status rather than its silicon implementation.
Where it went
This chip targets the Chinese market, which consumes a significant portion of global semiconductors. It is used in data centres for artificial intelligence training and inference. Applications include digital twin technology for smart factories and logistics systems. The processor supports businesses seeking to integrate AI into their operations while adhering to local and international trade laws. It helps maintain productivity in sectors that rely on advanced computing resources.
Versions and family
The H20 is part of a strategy that includes other compliant products, such as the RTX PRO GPU. These variants are designed to meet specific regulatory thresholds. The H20 represents a distinct tier within the company's export-compliant lineup, separate from unrestricted global models.
Why it matters
The availability of the H20 reflects the complex relationship between technology trade and national security policies. Nvidia reported that its market share in China dropped significantly during the period of strict export controls. The resumption of sales aims to recover some of this lost ground. Industry observers view the return of these chips as beneficial for both the supplier and Chinese tech firms that depend on imported hardware for their infrastructure.
Problems and reception
Industry representatives in China described the resumption of H20 sales as positive news. They noted that the large semiconductor market in the region is essential for global companies. The CEO of Nvidia characterized previous export control assumptions as flawed. The decision to re-apply for licenses was seen as a pragmatic step to restore business ties and support the continued development of AI capabilities in the region.
Questions people ask about Nvidia H20
What is the Nvidia H20 chip?
The H20 is a graphics processing unit designed by Nvidia for artificial intelligence workloads. It was specifically created to comply with United States export regulations, allowing it to be sold legally in the Chinese market.
Why did Nvidia stop selling the H20 initially?
Sales were restricted due to United States government export controls on high-performance AI chips. These regulations aimed to limit the transfer of advanced semiconductor technology to China, forcing Nvidia to adapt its product lineup.
When did Nvidia resume sales of the H20?
Nvidia announced in July 2025 that it would resume applying for licenses to sell the H20 in China. The company stated that the US government had assured them that licenses would be granted for these compliant chips.
Good reading and sources
There is no English Wikipedia article on Nvidia H20 yet, so the facts on this page come from the sources listed above: tomshardware.com/pc-components/gpus/alibaba-ai-chip-goes-head-to-head-, global.chinadaily.com.cn/a/202507/16/WS6876e6c1a31000e9a573c378.html. Some are the subject's own pages, so its claims about itself are reported as its claims. EveryChip wrote the text in its own words, with AI assistance, and compares it with the sources to make sure no passage is copied.
It may contain mistakes, and figures change over time. Nothing here is investment or buying advice. If something is wrong, tell us. Last refreshed 2026-11-14.