2 Sources
[1]
Nvidia seeks more China sales as banned AI chips sell on black market
The U.S. allowed Nvidia to resume sales of its downgraded H20 AI chip to Chinese clients on Monday. That's not enough for CEO Jensen Huang, who is already pushing for permission to sell more advanced processors to the world's second-largest economy. "Whatever we're allowed to sell in China will continue to get better and better over time," Huang said this week at the China International Supply Chain Expo in Beijing. "Technology is always moving." But while Washington debates easing restrictions on chip exports to China, there's already a thriving black market that has made the policy discussion largely academic. Research from the Center for a New American Security (CNAS) estimates that between 10,000 and several hundred thousand banned Nvidia chips may have been smuggled to China in 2024 alone. Chinese firms are already planning to install more than 115,000 restricted Nvidia AI chips in some three dozen data centers across the country's western deserts -- chips they cannot legally purchase without U.S. government licenses that haven't been granted. The smuggling operation appears sophisticated and widespread. One smuggler reportedly handled an order worth $120 million for servers containing 2,400 banned Nvidia H100s destined for China, according to The Information. Chinese businessmen have openly bragged online about obtaining hundreds of restricted H200 GPUs, while authorities in Singapore arrested three people suspected of diverting AI servers worth $390 million. Huang has consistently maintained there's "no evidence of any AI chip diversion," arguing that the massive servers are "nearly two tons" and easy to track. But Commerce Under Secretary Jeffrey Kessler directly contradicted those claims. "It's happening," he told lawmakers earlier this year. "It's a fact." The smuggling networks have become increasingly sophisticated. As of this month, more than 70 distributors were openly marketing restricted processors, with many offering delivery within weeks. The chips flow through a complex web of shell companies, third-party resellers, and intermediaries across Southeast Asia. Malaysia has emerged as a particular concern, with the country's imports of advanced GPUs surging over 3,400% in early 2025, prompting new permit requirements for AI chip exports. Malaysian authorities said they "will not tolerate the misuse of Malaysia's jurisdiction for illicit trading activities." Smugglers have used creative methods to move the hardware, including hiding chips in shipments labeled as tea or toys, and even packing them alongside live lobsters. Despite the logistical challenges, the evidence suggests the networks are moving substantial quantities of hardware across borders. The implications extend beyond trade policy. Chinese AI companies such as DeepSeek have already demonstrated impressive results that have caught Washington's attention, and Trump officials are now investigating whether the company accessed restricted chips through intermediaries to achieve those breakthroughs. Those Chinese-made tools are already gaining global traction: DeepSeek's models are now being tested internally by major banks, including HSBC and Standard Chartered, while Saudi Aramco has recently installed DeepSeek in its main data center. Even American cloud providers like Amazon Web Services, Microsoft, and Google offer DeepSeek to customers. If China's most advanced AI capabilities are already being built on smuggled top-tier chips while the government simultaneously funds development of domestic alternatives, it could undermine the entire rationale for export controls designed to preserve U.S. dominance in AI. Once the chips are out there, whether through legal or illegal channels, it becomes a new race to see whose AI models get adopted worldwide. "The No. 1 factor that will define whether the U.S. or China wins this race is whose technology is most broadly adopted in the rest of the world," Microsoft President Brad Smith said at a Senate hearing in May. "Whoever gets there first will be difficult to supplant."
[2]
NVIDIA's H20 AI Chip Ban Was Reportedly Lifted to Counter Huawei's Rising Influence in China and Beyond, Threatening U.S. Dominance
NVIDIA's H20 restriction uplift had a much deeper purpose than relieving Team Green, as AI czar David Sacks says it was a move to counter competition from China. NVIDIA's CEO, Jensen Huang, recently disclosed that the company expects its H20 AI chip to be available for China once again following the easing of US export controls. Now, when you talk about why this move happened apart from the trade deal between China and the US, AI czar David Sacks has disclosed a pretty interesting reason, one which Jensen has been discussing for several months. In an interview to Bloomberg Business, Sacks revealed that one of the reasons for lifting the H20 AI chip ban was to counter the growing influence of Huawei in Chinese markets. You just don't want to hand Huawei the entire Chinese market when NVIDIA is capable of competing for a big slice of it with a less capable chip. Again, we are not selling our latest, greatest chips to China, but we can deprive Huawei of the market share in China, which can be used to scale up to compete with the U.S. globally. Sacks previously mentioned that Huawei's CloudMatrix rack-scale solution is capable of competing with NVIDIA and that Chinese alternatives are a growing threat to the US's AI dominance. Interestingly, Sacks agreed to Huang's narrative that if US chips aren't being offered to the world, China would ultimately take over, and this does show that NVIDIA's CEO's opposition against the recent rounds of US export controls has worked out, since the Trump administration has adopted his narrative. The H20 AI chip ban getting lifted is just an example of how the U.S. sees an emerging threat from China's AI progress. Now that NVIDIA is expected to have access to China, domestic CSPs see more interest in American AI technology, despite Huawei's efforts to expand its market share. Sacks also supports the idea of Gulf states getting access to America's AI chips, claiming that if NVIDIA and others are deprived of the GCC market, China would ultimately take over, similar to the telecom sector. So, it's clear that the Trump administration had to make a choice here, whether to let Chinese tech flow into global markets, or ease the export controls on NVIDIA. It would be interesting to see how the markets evolve once NVIDIA resumes its business in China, since the firm is expected to introduce multi solutions for Beijing, in an attempt to regain lost market share.
Share
Copy Link
The U.S. eases restrictions on Nvidia's H20 AI chip sales to China, aiming to counter Huawei's growing influence. Meanwhile, a thriving black market for banned AI chips poses challenges to export controls.
The U.S. government has allowed Nvidia to resume sales of its downgraded H20 AI chip to Chinese clients, marking a significant shift in its export control policy. This decision comes as part of a strategy to counter the growing influence of Chinese tech giants like Huawei in the global AI market 1 2.
Source: Quartz
Nvidia's CEO, Jensen Huang, welcomed the move, stating, "Whatever we're allowed to sell in China will continue to get better and better over time." Huang is already pushing for permission to sell more advanced processors to China, emphasizing the continuous evolution of technology 1.
While the U.S. debates easing restrictions, a thriving black market for banned AI chips has emerged, potentially undermining the effectiveness of export controls. Research from the Center for a New American Security (CNAS) estimates that between 10,000 and several hundred thousand banned Nvidia chips may have been smuggled to China in 2024 alone 1.
The smuggling operations appear sophisticated and widespread:
The decision to lift the ban on Nvidia's H20 chip sales to China is seen as a strategic move to maintain U.S. dominance in the AI sector. AI czar David Sacks explained, "You just don't want to hand Huawei the entire Chinese market when NVIDIA is capable of competing for a big slice of it with a less capable chip" 2.
This move aligns with Nvidia CEO Jensen Huang's argument that if U.S. chips aren't offered to the world, China would ultimately take over. The Trump administration appears to have adopted this narrative, recognizing the emerging threat from China's AI progress 2.
Source: Wccftech
Chinese AI companies like DeepSeek have already demonstrated impressive results, catching Washington's attention. These Chinese-made tools are gaining global traction:
The global adoption of AI technologies is becoming a crucial factor in determining the leader in the AI race. Microsoft President Brad Smith emphasized, "The No. 1 factor that will define whether the U.S. or China wins this race is whose technology is most broadly adopted in the rest of the world" 1.
As the situation evolves, it remains to be seen how Nvidia's reentry into the Chinese market will impact the global AI landscape and whether the easing of restrictions will effectively counter China's growing influence in the sector.
Google rolls out an AI-powered business calling feature in the US and enhances its AI Mode with Gemini 2.5 Pro and Deep Search capabilities, revolutionizing how users interact with local businesses and conduct online research.
13 Sources
Technology
20 hrs ago
13 Sources
Technology
20 hrs ago
Nvidia and AMD are set to resume sales of AI chips to China as part of a broader US-China trade deal involving rare earth elements, sparking debates on national security and technological competition.
3 Sources
Policy and Regulation
5 hrs ago
3 Sources
Policy and Regulation
5 hrs ago
Calvin French-Owen, a former OpenAI engineer, shares insights into the company's internal workings, highlighting its rapid growth, secretive nature, and innovative yet chaotic work environment.
5 Sources
Technology
21 hrs ago
5 Sources
Technology
21 hrs ago
OpenAI has added Google Cloud to its list of cloud providers, joining Microsoft, Oracle, and CoreWeave. This move aims to meet the escalating demand for computing capacity needed to run AI models like ChatGPT.
7 Sources
Technology
13 hrs ago
7 Sources
Technology
13 hrs ago
Recent reports from AI safety organizations reveal that leading AI companies are ill-prepared for the potential dangers of developing human-level AI systems, with most scoring poorly on safety and risk management assessments.
2 Sources
Technology
4 hrs ago
2 Sources
Technology
4 hrs ago