
CNBC quoted sources as saying that the Chinese AI startup is negotiating to lease computing power in European facilities to circumvent U.S. export controls.
AI-generated summary
U.S. export controls restrict the export of Nvidia's most advanced chips to China, but companies can still use these chips remotely through overseas data centers, creating a regulatory loophole. Current law does not give regulators broad authority to control companies' use of these chips through overseas cloud services. The White House in July accused Dark Side of the Moon of using Nvidia Blackwell chips remotely through a facility in Thailand.
Sources told U.S. financial media CNBC that Dark Side of the Moon, a Chinese AI startup that has been accused by the White House of using advanced Nvidia chips remotely in East Asia, is trying to obtain these chips through data centers in Europe instead.
CNBC reported that four people familiar with the matter said that Dark Side of the Moon was discussing deals this year to lease computing power from emerging cloud players (neoclouds) in European facilities, which are equipped with Nvidia graphics processors (GPUs). The people spoke on condition of anonymity because they were discussing non-public information.
Nvidia has a near-monopoly position in the high-end chip market, making U.S. export controls an important tool for Washington to maintain its competitive advantage over China's AI. American AI companies such as OpenAI and Anthropic are currently engaged in model training and price wars with Chinese developers, including Dark Side of the Moon, which developed the Kimi series of models.
Three sources told CNBC that Dark Side of the Moon was seeking to lease computing power provided by Nvidia's Blackwell systems. Blackwell is Nvidia's advanced chip, which is currently subject to relevant export control restrictions.
The White House said in July that Dark Side of the Moon used Nvidia Blackwell chips remotely through a facility in Thailand.
U.S. export controls restrict the export of Nvidia's most advanced chips to China, but companies can still use these chips remotely through overseas data centers, creating a loophole. Current law does not give regulators broad authority to control companies' use of these chips through overseas cloud services.
Dark Side of the Moon, Nvidia, the White House, the U.S. Department of Commerce, the Chinese Embassy in the United Kingdom and the Chinese Ministry of Foreign Affairs did not respond to requests for comment.
In the past six months, the technical capabilities of Chinese AI developers have improved significantly, and if they can obtain more rights to use Nvidia chips, it will put greater pressure on U.S. competitors.
If Chinese AI companies want to rent the computing power of Nvidia chips in Europe, they usually rent computing resources from emerging non-Chinese cloud service providers or data center operators who can purchase high-end chips.
Two sources said that Dark Side of the Moon intends to rent computing clusters in Europe to perform inference work, that is, to provide model services to end users. One of the sources said they had also learned that Dark of the Moon was interested in training computing clusters.
A source said that the two parties have negotiated relevant arrangements, including allowing Dark Side of the Moon to rent a locally located computing cluster equipped with at least 8,000 Nvidia Blackwell GPUs. The person described the Chinese AI startup as "very aggressive" in seeking more computing power.
A person familiar with the matter told CNBC that since the White House alleged that Dark Side of the Moon obtained computing power through Thailand, relevant people have been negotiating to obtain computing power in Europe. The Dark Side of the Moon did not respond to requests for comment at the time. (Compiled by: Chen Yuting) 1151106
AI outlook — possibilities, not facts
The U.S. may introduce new regulations on the use of restricted Nvidia chips through overseas cloud services
Likely · Within months
Dark Side of the Moon may eventually rent European computing clusters for inference and training
Possible · Within months

Reuters reported that about 600,000 Nvidia H200 chips are currently stored in Hong Kong because Beijing has not yet approved large-scale imports. Although some chips have entered the mainland through indirect channels, the Chinese government is wary of over-reliance on U.S. technology and has limited the scale of imports.

Hua Hong Grace Semiconductor is scaling up its silicon photonics and power management chip production to capitalize on the rising demand for artificial intelligence infrastructure.

Huawei has unveiled its Qinyun HM650 series desktop computers, the first to run on its proprietary HarmonyOS and feature 100 per cent domestically sourced components, including the Kirin X90 processor, targeting enterprise and government sectors.

Lithuania is strengthening its position in the global semiconductor and AI hardware supply chain as laser manufacturer Litilit opens a new factory in Vilnius, despite ongoing trade and diplomatic tensions with China regarding Taiwan.

The US Federal Communications Commission has voted to stop recognizing Chinese testing laboratories and certification bodies for electronic products, effective December 1, 2028, citing a reliance on Chinese firms for 82% of US equipment-authorization applications.
The 2026 World Internet Conference Wuzhen Summit concluded on November 5 in Wuzhen, Zhejiang. This summit brought together about 1,800 guests from 138 countries and regions around the world, released a number of results such as the "Working Together to Build a Global Open Artificial Intelligence Action Initiative", and discussed artificial intelligence governance and cyberspace cooperation.