54 Sources
54 Sources
[1]
Nvidia said to be developing new, more powerful AI chip for sale in China | TechCrunch
The world's most valuable chipmaker is far from giving up its desire to retain China as a key growth market. Nvidia is apparently putting together a new AI chip meant for sale in China that's half as powerful as its flagship B300 Blackwell GPU, Reuters reported, citing anonymous sources. This new chip, codenamed B30A, would be more powerful than the H20 GPUs the company is currently allowed to sell to China, the report said, though its design will be single-die in nature, unlike the dual-die design used in the more powerful B300 GPUs. The B30A would still have similar features to the H20, like fast data transmission, support for NVLink, and high-bandwidth memory. The development of the B30A seems to be separate from another chip that the company is said to be developing for sale in the country, Reuters reported. "We evaluate a variety of products for our roadmap, so that we can be prepared to compete to the extent that governments allow. Everything we offer is with the full approval of the applicable authorities and designed solely for beneficial commercial use," Nvidia said in an emailed statement. The news comes as the Trump administration in recent weeks has relaxed its stance on letting chipmakers export high-performance AI chips to China, though Reuters cited its sources as saying that approvals for this new chip aren't certain at all. As geopolitical tensions between Beijing and Washington are increasingly affected by the race to develop AI, critics argue that the U.S. must maintain its lead by controlling the supply of necessary technology to China. But Nvidia and its peers -- finding themselves selling the equivalent of shovels during a gold rush -- maintain that ceding the valuable market of China to rivals like Huawei would be tantamount to giving up entirely.
[2]
Nvidia reportedly halts production on its H20 AI chips | TechCrunch
Beijing may have thrown a wrench into Nvidia's plans on making a comeback in China's AI market. Nvidia has instructed its component suppliers to stop production related to its H20 AI chip, according to The Information. The production halt comes after Beijing reportedly warned Chinese companies against using these chips because of potential security issues and fears of backdoors that would give the U.S. access to sensitive data. China's government is urging companies to use domestic chips instead, The Information reported. This comes a month after companies including Nvidia were given the green light to start selling its AI chips designed for China's market. TechCrunch reached out to Nvidia for more information.
[3]
Fragmented ecosystems and limited supply: Why China cannot break free from Nvidia hardware for AI
Can China's attempts to use homegrown chips work for the country's AI ambitions? Last week saw major twists in China's AI landscape: Trump imposed a 15% sales tax on AMD and Nvidia hardware sold to China, Beijing froze new Nvidia H20 GPU purchases over security concerns, and DeepSeek dropped plans to train its R2 model on Huawei's Ascend NPUs -- raising doubts about China's ability to rely on domestic hardware for its AI sector. As part of its recurring five-year strategic plans, China's long-stated goal has been to gain its own technological independence, particularly in new and emerging segments that it sees as key to its national security. However, after years of plowing billions into fab startups and its own nascent chip industry, that country still lags behind its Western counterparts and has struggled to build its own truly insulated supply chain that can create AI accelerators. Additionally, the country lacks an effective software ecosystem to rival Nvidia's CUDA, creating even more challenges. Here's a closer look at how this is impacting the country's AI efforts. China has had a self-sufficiency plan for its semiconductor industry in general since the mid-2010s. Over time, as the U.S. imposed sanctions against the People's Republic's high-tech sectors, the plan evolved to address supercomputers (including those capable of AI workloads) and fab tools. In 2025, China has created several domestic AI accelerators, and Huawei has even managed to develop its rack-scale CloudMatrix 384. However, ever since the AI Diffusion Rule was canned, and the incumbent Trump administration banned sales of AMD's Instinct MI308 and Nvidia's HGX H20 to Chinese entities, the PRC doubled down on its efforts to switch crucially important AI companies to using domestic hardware. As a result, when the U.S. government announced plans to grant AMD and Nvidia export licenses to sell their China-specific AI accelerators to clients in the People's Republic, U.S. President Trump announced an unprecedented 15% sales tax on AMD's and Nvidia's hardware sold to China. China's government then made shipments of Nvidia's HGX H20 hardware strategic, and instructed leading cloud service providers to halt new purchases of Nvidia's H20 GPUs while it examines alleged security threats, a move that could potentially bolster demand for domestic hardware. This may be good news for companies like Biren Technology, Huawei, Enflame, and Moore Threads. There's a twist in this tale, though -- DeepSeek reportedly had to abandon training of its next-generation R2 model on domestically developed Huawei's Ascend platforms because of unstable performance, slower chip-to-chip connectivity, and limitations of Huawei's Compute Architecture for Neural Networks (CANN) software toolkit. This all begs the question: can China rely on its homegrown hardware for AI development? Nvidia has been supplying high-performance AI GPUs fully supported by a stable and versatile CUDA software stack for a decade, so it's not surprising that many, if not all, of the major Chinese AI hyperscalers -- Alibaba, Baidu, Tencent, and smaller players like DeepSeek currently use Nvidia's hardware and software. Although Alibaba and Baidu develop their own AI accelerators (primarily for inference), they still procure tons of Nvidia's HGX H20 processors. SemiAnalysis estimated that Nvidia produced around a million HGX H20 processors last year, and almost all of them were purchased by Chinese entities. No other company in China supplied a comparable number of AI accelerators in 2024. However, analyst Lennart Heim believes that Huawei had managed to illegally obtain around three million Ascend 910B dies in 2024 from TSMC, which is enough to build around 1.4 - 1.5 million Ascend 910C chips in 2024 - 2025. This is comparable to what Nvidia supplied to China in the same period. However, while Huawei may have enough Ascend processors to train its Pangu AI models, it appears that other companies have other preferences. DeepSeek trained the R1 model on a cluster of 50,000 Hopper-series GPUs. This consisted of 30,000 HGX H20s, 10,000 H800s, and 10,000 H100s. These chips were reportedly purchased by DeepSeek's investor, High-Flyer Capital Management. As a result, it's logical that the whole software stack of DeepSeek -- arguably China's most influential AI software developer -- is built around Nvidia's CUDA. However, when the time came to assemble a supercluster to train DeepSeek's upcoming R2 model, the company was reportedly persuaded by the authorities to switch to Huawei's Ascend 910-series processors. However, when it encountered unstable performance, slower chip-to-chip connectivity, and limitations of Huawei's CANN software toolkit, it decided to switch back to Nvidia's hardware for training, but use Ascend 910 AI accelerators for inference. Speaking of these exact accelerators, we do not know whether DeepSeek used Huawei's latest CloudMatrix 384, based on the latest Ascend 910C, or something else. Since DeepSeek has not disclosed these challenges officially, we can only rely on a report from the Financial Times. The publication claims that Huawei's Ascend platforms did not work well for DeepSeek. Why they were deemed to be unstable is another question. It's a distinct possibility that DeepSeek only began to work with CANN this Spring, so the company simply has not had enough time to port its programs from Nvidia's CUDA to Huawei's CANN toolkit. It is extremely complicated to analyze high-tech industries in China, as companies tend to keep secrets closely guarded and fly under the U.S. government's radar. However, two important factors that may have a drastic effect on the development of AI hardware in China occurred this summer. Firstly, the Model-Chip Ecosystem Innovation Alliance was formed, and secondly, Huawei made its CANN software stack open source. The Model Chip Ecosystem Innovation Alliance includes Huawei, Biren Technologies, Enflame, and Moore Threads and others. The group aims to build a fully localized AI stack, linking hardware, models, and infrastructure, which is a clear step away from Nvidia or any other foreign hardware. Its success depends on achieving interoperability among shared protocols and frameworks to reduce ecosystem fragmentation. While low-level software unification may be difficult due to varied architectures (e.g., Arm, PowerVR, custom ISAs), mid-level standardization is more realistic. By aligning around common APIs and model formats, the group hopes to make models portable across domestic platforms. Developers could write code once -- e.g., in PyTorch -- and run it on any Chinese accelerator. This would strengthen software cohesion, simplify innovation, and help China build a globally competitive AI industry using its own hardware. There is also an alliance called the Shanghai General Chamber of Commerce AI Committee, which focuses on applying AI in real-world industries; this also unites hardware and software makers. Either as part of the commitment to the new alliance, or as part of the general attempt to make its Ascend 910-series the platform of choice among China-based companies, Huawei open-sourced CANN in early August, which is specifically optimized for AI and its Ascend hardware. Until this summer, Huawei's AI toolkit for its Ascend NPUs was distributed in a restricted form. Developers had access to precompiled packages, runtime libraries, and bindings, which allowed TensorFlow, PyTorch, and MindSpore to run on the hardware. These pieces worked well enough to allow users to train and deploy models, but the underlying stack, such as compilers or libraries, remained closed. Now, this barrier has been removed. The company released the source code for the full CANN toolchain; however, it did not formally confirm what exactly it unseals, so we can only wonder or speculate. The list of opened up technologies likely includes compilers that convert model instructions into commands that Ascend NPUs understand, such as low-level APIs, libraries of AI operators that accelerate core math functions, and a system-level runtime. This will allow the management of memory, scheduling, and communication. This isn't officially confirmed, but merely an educated guess as to what CANN's open-sourcing might enable. By opening up CANN, Huawei can attract a broad community of developers from academia, startups, and other enterprises to its platform, and enable them to experiment with performance tuning or framework integration (beyond TensorFlow and PyTorch). This will inevitably speed up CANN's evolution and bug fixing. Eventually, these efforts could bring CANN closer to what CUDA offers, which will be a useful string in Huawei's bow. For Huawei, opening up CANN ahead of other Model-Chip Alliance members was beneficial, as it already had the most mature AI hardware platform in production, and needed to position its Ascend platform as the baseline software ecosystem others could rely on. This move makes CANN the default foundation for domestic models and hardware developers (at least for now). By taking this first step, Huawei set a reference point for interoperability and signalled a commitment to shared standards, which could help reduce fragmentation in China's AI software stack. But while unification of the software stack is a step in the right direction, there is an elephant in the room regarding China's AI hardware self-reliance. The People's Republic still cannot produce hardware that is on par with AMD or Nvidia in volume domestically. The hardware that can be made in China is years behind the processors developed on U.S. soil. All leading developers of AI accelerators in China, like Biren, Huawei, and Moore Threads, are in the U.S. Department of Commerce's Entity List. This means that they do not have access to the advanced fabrication capabilities of TSMC. To that end, they have to produce their chips at China-based SMIC, whose process technologies cannot match those offered by TSMC. While SMIC can produce chips on its 7nm-class fabrication process, Huawei had to obtain the vast majority of silicon for its Ascend 910B and Ascend 910C processors by deceiving TSMC. Companies like Biren or Moore's Threads do not disclose which foundry they use, but they do not have the luxury of choice. Of course, neither Huawei nor SMIC stands still. The two companies are working to advance China's semiconductor industry and build a local fab tools supply chain that will replace the leading-edge equipment that SMIC cannot acquire. Before this happens, SMIC is expected to start building chips on its 6nm-class process technology and even 5nm-class production node, so it may well build advanced AI processors for Huawei and other players. But the big question is whether volumes will manage to meet the demands of AI training and inference, especially if Nvidia hardware is largely unobtainable in China. The maturity of Huawei's CANN (and competing stacks) lags behind Nvidia's CUDA largely because there has not been a broad, stable installed base of Ascend processors outside Huawei's own projects. Developers follow scale, and CUDA became dominant because millions of Nvidia GPUs were shipped and widely available, which justified investment in tuning, libraries, and community support. In contrast, Huawei and other Chinese developers have their proprietary software stacks and cannot ship millions of Ascend NPUs or Biren GPUs due to sanctions from the U.S. government. On the other hand, even if Huawei and others managed to flood the market with Ascend NPUs or Moore Threads GPUs, a weak software stack makes them unattractive for developers. DeepSeek's attempt to train R2 on Ascend is a good example: performance instability, weaker interconnects, and CANN's immaturity reportedly made the project impractical, forcing a return to Nvidia hardware for training. Hardware volume alone will not change that. The new Model-Chip Ecosystem Innovation Alliance is attempting to address the issue by setting common mid-level standards -- things like shared model formats, operator definitions, and framework APIs. The idea is that developers could write code once in PyTorch or TensorFlow and then run it on any Chinese AI accelerator, whether it is from Huawei, Biren, or another vendor. However, until these standards are actually in place, fragmentation means every company will face several problems at once. The hardware and software face competition across multiple fronts in a saturated market. As a result, the low volume of China-developed AI accelerators, a lack of common standards, and competition on various fronts will make it very hard for Chinese companies to challenge Nvidia's already dominant ecosystem.
[4]
China Tells Tech Giants to Halt Nvidia H20 Orders After U.S. Official's 'Addiction' Remark -- Chinese leaders call Lutnick's comments 'insulting'
'You want to sell the Chinese enough that their developers get addicted to the American technology stack.' -- Sec. Lutnick Although China has been trying to move away from foreign chips for several years now, the Financial Times says that this effort has now intensified after being fueled by U.S. Commerce Secretary Howard Lutnick saying the aim was to make China "addicted" to US tech during an interview after Washington allowed Nvidia to resume the sales of its H20 to the country, using a term that holds signifcant cultural undertones in China. "We don't sell them our best stuff, not our second-best stuff, not even our third-best. I think fourth best is where we have come out that we're cool," Lutnick said during the July 15 CNBC interview. He also added, "So you want to sell the Chinese enough that their developers get addicted to the American technology stack. And that's the thinking." These remarks caught the ire of Chinese authorities, with some of the country's senior leadership calling it "insulting." Furthermore, the "addiction" connotation brings back memories of the Opium Wars, when Britain forced China to legalize the drug, leading to an addiction epidemic in the country and the loss of territory to foreign powers, including Hong Kong, the Kowloon Peninsula, and Outer Manchuria. This event is particularly raw among Chinese historians and is often marked as the start of its "Century of Shame." China has directed its regulatory bodies, including the Cyberspace Administration of China (CAC), the National Development and Reform Commission (NDRC), and the Ministry of Industry and Information Technology (MIIT), to find ways to force Chinese tech institutions to move away from buying Nvidia H20 chips. The Financial Times says that companies have been holding off on purchases and reducing their orders for Nvidia's AI GPUs because of this. This follows some of the broader initiatives already underway. Chinese regulators have recently been pushing companies away from Nvidia AI GPUs, especially the H20, with state media calling it unsafe and outdated, the government telling local companies to avoid it, and mandating that domestic data centers should purchase at least 50% of their AI chips from local manufacturers. This might be a problem for the company, especially as Nvidia initially reported seeing high demand for the H20 and ordering more chips from TSMC to satisfy this huge appetite from China. Nevertheless, it's unlikely that the American AI giant will be completely supplanted by its domestic rivals anytime soon. Although competitors like Huawei and Cambricon are capable of producing chips that, on paper, can compete with what Nvidia has to offer for specific applications, the latter's tech stack is still the primary reason why Chinese tech companies are hesitant to move away from it.
[5]
China mandates domestic firms source 50% of chips from Chinese producers -- Beijing continues to squeeze companies over reliance on foreign semiconductors
Questions remain about how native chips will work with existing AI infrastructure and software platforms like CUDA. China has mandated that all domestic data centers begin using more Chinese-produced processors as part of an increasing initiative to make China self-sufficient when it comes to silicon, as reported by SCMP. This comes at a time of enormous global investment in datacenters, AI, and processor production, with an increased drive towards self-sufficiency and nationalistic stances on cutting-edge technology. Moving forward, publicly-owned Chinese datacenter firms will reportedly be required to source more than 50% of their chips from domestic producers, according to people said to be familiar with the matter. This comes from guidelines initially published last March by the Shanghai municipality. This order appears to have now been extended to the wider country and could have a large impact on Chinese chip adoption and investment, as well as the country's overall interest in US processors. Major US companies like Nvidia and AMD have faced federal hurdles to selling their latest chips to Chinese firms for some time, and are currently only able to sell watered-down versions of their latest designs, all while giving the US government a 15% cut of any proceeds, per terms of the latest White House deal. Although it seems that the US government does want to maintain favorable trading relationships with China and its domestic companies, the process has been fraught with turbulence in recent months, raising concerns that companies unable to source the chips they needed would turn to other avenues to acquire them. That's led to smuggling, but also may be part of the drive towards more local chip production in China. Like Nvidia and other US company initiatives to spend hundreds of billions of dollars in datacenter construction and US-based chip fabrication, China has also announced a plethora of new datacenter projects in the past few years. Many of these were expected to be powered by some version of the latest GPUs from Nvidia and similar companies. However, amidst on-again, off-again trade blockades and Chinese concerns over tracking hardware built into the graphics cards, it's looking to drive more investment into locally produced chips moving forward. The news follows reports last week that China was discouraging the use of products like Nvidia's H20, hinting at security concerns. The result of this policy may be a slowdown in Chinese AI innovation, though. While locally produced chips are said to be capable enough for AI inference - the day-to-day operation of running the AIs - they aren't as capable as the latest Nvidia and AMD GPUs for training. Limited access to the latest designs - and any future designs going forward - could severely limit Chinese innovation in this space. Its domestically produced GPUs are considered years behind the likes of Nvidia's latest developments. Notably, DeepSeek R2 was recently reported to have run into delays after it was trained on Huawei chips rather than Nvidia hardware at the behest of Beijing. There's also the question of what software such Chinese chips might run on, and how that might work in conjunction with existing American processors. Nvidia chips can leverage the powerful and near-ubiquitous CUDA ecosystem, which Chinese-produced chips will not be able to take advantage of.
[6]
Nvidia could be readying B30A accelerator for Chinese market -- new Blackwell chip reportedly beats H20 and even H100 while complying with U.S. export controls
Nvidia is reportedly working on new AI accelerators based on the Blackwell architecture for the Chinese market -- the B20A for AI training and the RTX 6000D for AI inference -- that will beat existing HGX H20 and L20 PCIe products, but will still meet U.S. export controls, Reuters says, citing sources familiar with the matter. If the information is correct, then Chinese customers may get two rather formidable products with performance on par with previous-generation flagships. The Nvidia B30A is said to be based on the Blackwell Ultra microarchitecture but relies on only one compute chiplet and offers about half of B300's compute performance and 50% of its HBM3E memory capacity (i.e., 144 GB HBM3E). Previously, such a product was rumored to be called B300A and was intended for the general market, not just China. Since the B30A is said to be a China-oriented SKU, it is possible that Nvidia will introduce certain performance restrictions for this part. If the information is correct and the GPU will hit the reported performance targets, then it will not only outpace the HGX H20, but it will actually beat the previous-generation flagship, the H100. The unit will also feature NVLink for scale-up connectivity, though it is unclear whether Nvidia will restrict building rack-scale solutions or big clusters by cutting down the number of NVLinks for this product. Building B30A (or B300A) should be fairly easy for Nvidia as one compute chiplet and four HBM3E memory stacks can likely be packaged using TSMC's proven CoWoS-S technology (at least according to SemiAnalysis) that happens to be cheaper than the CoWoS-L used by the B200 and B300 processors, which contain two compute chiplets and eight HBM3E modules. In addition to the B30A, Nvidia reportedly plans to offer an RTX 6000D product, designed for AI inference and possibly for professional graphics applications. The unit will be cut down compared to the fully-fledged RTX 6000. Reuters suggests that its memory bandwidth will be around 1.398 TB/s, though details are unknown. Nvidia's customers are expected to get the first samples of B30A and RTX 6000D in September, so if the products are greenlit by the U.S. government, Nvidia will be able to ship commercial B30A modules and RTX 6000D cards by the end of 2025 or in early 2026. The rumor about new AI accelerators for China from Nvidia comes after President Donald Trump suggested he might permit sales of next-generation Nvidia parts in China as long as a new arrangement under which Nvidia and AMD would give the U.S. government 15% of revenue from their China revenue is met. Still, bipartisan opposition among legislators continues to question whether even cut-down versions of advanced AI hardware should be accessible to Chinese companies.
[7]
Nvidia responds to reports that its H20 GPU for China is ending production -- next-gen B30A green light "up to the United States government," according to CEO Jensen Huang
Following a turbulent summer that saw sales of Nvidia's China-only HGX H20 AI chip banned and then unbanned in China, the AI GPU now faces an uncertain future. A new report suggests that Nvidia is abruptly stopping production of the HGX H20, which Nvidia has responded to in a statement provided to Tom's Hardware. This comes soon after reports earlier this week that Nvidia is preparing a successor to the chip, the Blackwell Ultra-based B30A. A report from The Information claims that Nvidia has asked suppliers to wind down production of H20-related work. According to the report, Nvidia sent these communications to Amkor Technology, which handles packaging for the H20, and Samsung Electronics, which provides the High Bandwidth Memory (HBM) for the AI chip. This could potentially signal that production of the H20 is winding down, but that leaves the door open to more questions. When quizzed on whether or not the H20 would remain in production, an Nvidia spokesperson told Tom's Hardware, "We constantly manage our supply chain to address market conditions". While this isn't an exact response to our question about whether or not the H20 is still in production, it implies that some degree of the communications from Nvidia to Amkor and Samsung do indeed hold some weight. Nvidia also insists that the H20 is not a political tool or has security backdoors, with a spokesperson for the company stating, "As both governments recognize, the H20 is not a military product or for government infrastructure". Cybersecurity was also on the mind, with Nvidia once again insisting that the H20 has no security backdoors, despite the CAC's claims: "NVIDIA does not have 'backdoors' in our chips that would give anyone a remote way to access or control them. The market can use the H20 with confidence." This is a sentiment echoed by CEO Jensen Huang, who spoke to journalists in Taipei, Taiwan: "We have made very clear and put to rest that H20 has no security backdoors, there are no such things, there never has, and so hopefully the response that we've given to the Chinese government will be sufficient". In between rumors of the H20 seemingly winding down production, and the perceived highlighted risk of security backdoors, another factor has emerged: A potential successor to the H20 itself, a new high-powered chip, specifically for China. Earlier this week, reports emerged that the company was working on a de facto successor to the H20, which will comply with existing export control rules, and could be sold on the Chinese market. The Blackwell Ultra-based B30A is a substantially more powerful chip than the H20, which beats the full-fat H100 in many ways, but remains significantly slower than the B300. Samples are expected to land in September. But the product has yet to be greenlit by the U.S. Government. The earliest that commercial units could be shipped would be by the end of 2025, to early 2026. Huang further stressed: "Offering a new product to China for the data center, AI data centers, the follow-on to H20, that's not our decision to make. It's up to of course the United States government, and we're in dialogue with them, but it's too soon to know". Nvidia is caught between two global superpowers: The U.S. Government and the People's Republic of China. It's an unenviable situation, with complex factors at play, including the race to develop the most advanced AI models and local fabrication. China is making moves to reduce reliance on U.S. hardware with a new strategic alliance to reduce fragmentation, and the open-sourcing of CANN, in a bid to reduce reliance on Nvidia's CUDA. Nvidia's HGX H20 has long been a point of contention for the company, the U.S. Government, and the People's Republic of China. As the H20 is commonly used to develop AI applications in China, Nvidia's chips are likely being used for training leading AI models, such as DeepSeek R2. The HGX H20 itself is a significantly scaled-down version of Nvidia's H100 GPU, which is approximately 3.3 to 6.6 times faster than the H20. China has access to rapid hardware, such as Huawei's rack-scale CloudMatrix 384, but a reliance on Nvidia's CUDA-shaped moat has thrown a wrench into China's AI aspirations. Despite attempts to switch to Huawei platforms for training the new DeepSeek R2 model, the company has reportedly now switched back to using Nvidia hardware instead, highlighting the strategic importance of high-powered AI hardware. However, in April 2025, sales of the H20 to China were banned, causing Nvidia to write off $5.5 billion in inventory. This decision was later reversed in mid-July 2025. China's access to Nvidia's H20 chips was no longer hampered. With the floodgates opened, the inventory of existing H20 stock quickly became limited. So, Nvidia reportedly put in an order for 300,000 AI GPUs with TSMC in late July. This deal is believed to be worth around $3.9 billion, if each H20 GPU is valued between $12,000 - $14,000 per unit (via Reuters). Then, the U.S. wanted a slice of the pie, with the government requesting that 15% of each GPU sold to China be diverted into federal coffers. Days later, the Cyberspace Administration of China (CAC) asked major companies to pause new H20 purchases, as it investigated potential security risks, a claim which Nvidia has since rebuffed. Whether China's efforts will be successful remains to be seen. However, one thing is clear. The demand for high-powered chips remains strong, but whether or not Nvidia will be able to provide them for Chinese customers is a question only the U.S. Government can answer.
[8]
Nvidia Reportedly Asks Suppliers to Halt H20 Work
Nvidia has instructed component suppliers to stop production related to the H20, The Information reported, following news Beijing urged local companies to avoid using the AI chip designed specifically for the Chinese market. Nvidia chief Jensen Huang, meanwhile, told reporters in Taiwan that the company assured there were no "security backdoors" on those chips, after China voiced concerns. Ed Yardeni, president at Yardeni Research, explains what it could mean for the AI race and tech stocks. He speaks on Bloomberg Insight. (Source: Bloomberg)
[9]
Exclusive: Nvidia working on new AI chip for China that outperforms the H20, sources say
BEIJING/SINGAPORE, Aug 19 (Reuters) - Nvidia (NVDA.O), opens new tab is developing a new AI chip for China based on its latest Blackwell architecture that will be more powerful than the H20 model it is currently allowed to sell there, two people briefed on the matter said. U.S. President Donald Trump last week opened the door to the possibility of more advanced Nvidia chips being sold in China. But the sources noted U.S. regulatory approval is far from guaranteed amid deep-seated fears in Washington about giving China too much access to U.S. artificial intelligence technology. The new chip, tentatively known as the B30A, will use a single-die design that is likely to deliver half the raw computing power of the more sophisticated dual-die configuration in Nvidia's flagship B300 accelerator card, the sources said. A single-die design is when all the main parts of an integrated circuit are made on one continuous piece of silicon rather than split across multiple dies. The new chip would have high-bandwidth memory and Nvidia's NVLink technology for fast data transmission between processors, features that are also in the H20 - a chip based on the company's older Hopper architecture. The chip's specifications are not completely finalised but Nvidia hopes to deliver samples to Chinese clients for testing as early as next month, said the sources who were not authorised to speak to media and declined to be identified. Nvidia said in a statement: "We evaluate a variety of products for our roadmap, so that we can be prepared to compete to the extent that governments allow." "Everything we offer is with the full approval of the applicable authorities and designed solely for beneficial commercial use," it said. The U.S. Department of Commerce did not respond to a Reuters request for comment. FLASHPOINT The extent to which China, which generated 13% of Nvidia's revenue in the past financial year, can have access to cutting-edge AI chips is one of the biggest flashpoints in U.S.-Sino trade tensions. Nvidia only received permission in July to recommence sales of the H20. It was developed specifically for China after export restrictions were put in place in 2023, but company was abruptly ordered to stop sales in April. Trump said last week he might allow Nvidia to sell a scaled-down version of its next-generation chip in China after announcing an unprecedented deal that will see Nvidia and rival AMD (AMD.O), opens new tab give the U.S. government 15% of revenue from sales of some advanced chips in China. A new Nvidia chip for China might have "30% to 50% off", he suggested in an apparent reference to the chip's computing power, adding that the H20 was "obsolete". U.S. legislators, both Democratic and Republican, have worried that access to even scaled-down versions of flagship AI chips will impede U.S. efforts to maintain its lead in artificial intelligence. But Nvidia and others argue that it is important to retain Chinese interest in its chips - which work with Nvidia's software tools - so that developers do not completely switch over to offerings from rivals like Huawei. Huawei has made great strides in chip development, with its latest models said to be on par with Nvidia in some aspects like computing power, though analysts say it lags in key areas such as software ecosystem support and memory bandwidth capabilities. Complicating Nvidia's efforts to retain market share in China, Chinese state media have also in recent weeks alleged that the U.S firm's chips could pose security risks, and authorities have cautioned Chinese tech firms about purchasing the H20. Nvidia says its chips carry no backdoor risks. Nvidia is also preparing to start delivering a separate new China-specific chip based on its Blackwell architecture and designed primarily for AI inference tasks, according to two other people familiar with those plans. Reuters reported in May that this chip, currently dubbed the RTX6000D, will sell for less than the H20, reflecting weaker specifications and simpler manufacturing requirements. The chip is designed to fall under thresholds set by the U.S. government. It uses conventional GDDR memory and features memory bandwidth of 1,398 gigabytes per second, just below the 1.4 terabyte threshold established by restrictions introduced in April that led to the initial H20 ban. Nvidia is set to deliver small batches of RTX6000D to Chinese clients in September, said one of the people. Reporting by Liam Mo in Beijing and Fanny Potkin in Singapore; Additional reporting by Alexandra Alper in Washington and Wen-Yee Lee in Taipei; Editing by Brenda Goh and Edwina Gibbs Our Standards: The Thomson Reuters Trust Principles., opens new tab
[10]
Beijing turns against Nvidia's AI chip after 'insulting' Lutnick remarks
Beijing's move to restrict sales of Nvidia's China-specific artificial intelligence processor was prompted by remarks from US commerce secretary Howard Lutnick about chip exports that officials found "insulting". A group of Chinese regulators have mobilised in an effort to dissuade domestic tech companies from acquiring the H20 -- a watered-down processor widely used for artificial intelligence in China. According to people with knowledge of the regulatory action, the Cyberspace Administration of China (CAC), the National Development and Reform Commission (NDRC) and the Ministry of Industry and Information Technology (MIIT) moved in response to comments made by Lutnick last month. "We don't sell them our best stuff, not our second-best stuff, not even our third-best," Lutnick told CNBC on July 15, the day after the Trump administration lifted export controls, implemented in April, on H20 sales. "You want to sell the Chinese enough that their developers get addicted to the American technology stack, that's the thinking," he added. Some of China's senior leaders found the comments "insulting", leading to the policymakers to seek ways to restrict Chinese tech firms from buying the processors, according to two people with knowledge of the latest regulatory decision-making. As a result, Chinese tech groups held off or significantly downsized their H20 orders, according to those with knowledge of their plans. The moves have come as a blow to Nvidia, whose chief executive Jensen Huang last month visited Beijing and committed to stay competitive in the country despite growing geopolitical tensions with the US. Following Huang's warmly received trip, Nvidia received sufficient interest from Chinese clients that it told fabrication partner TSMC to reopen its H20 production lines, according to two people with knowledge of the matter. Chinese regulators have urged more use of domestic chips in recent years, but tech giants from Alibaba to ByteDance argued that their AI development would be impaired without Nvidia's chips, hurting China's chance to win the technology arms race with the US. However, some close to the tech companies said they have now become more accepting of a switch, especially for "inference", in which AI systems respond to requests from users. That shift came after testing and adopting chips from domestic producers led by Huawei and Cambricon at a larger scale, following Washington's initial April ban on exporting Nvidia's H20s. "Lutnick's speech gives the coalition [of regulators] one more reason to intensify its efforts to push tech firms to use China's own chips," said a person close to the policymakers. A week after his comments, China's internet regulator CAC issued so-called "window guidance" -- an informal notice -- to major tech firms such as ByteDance and Alibaba, citing security concerns and instructing them to halt new orders for Nvidia's H20 chips, according to people with knowledge of the regulator's effort. On July 31, the agency summoned Nvidia executives over alleged "serious security issues". In a statement, the CAC claimed that US AI experts had revealed Nvidia's chips have location tracking and can be shut down remotely -- a claim strongly disputed by Nvidia. MIIT, China's regulator of telecoms and software, also spoke with Chinese tech executives informally to echo CAC's stance, according to one of the people with knowledge of the meetings. The NDRC, China's state planner that is in charge of the country's drive for tech independence, then issued its own window guidance, requesting that tech firms refrain from purchasing all Nvidia chips, including the H20, said those with knowledge of the move. NDRC has been for years tasked with promoting chip independence and helping domestic players such as Huawei to capture market share from Nvidia. The new involvement of CAC in particular has created more pressure for Chinese tech giants to comply, even if the instructions remain informal, as any potential penalties from the watchdog would choke their daily operations. These actions are in sharp contrast to other Chinese departments, such as the commerce and foreign affairs ministries, which have been more open to Nvidia's business, according to industry insiders. The two ministries, which are also in charge of trade negotiations with the US, welcomed Huang in July as a positive signal for foreign businesses and to show Beijing's goodwill in the ongoing trade talks, the people said. "A lot of uncertainties remain depending on trade negotiations and Washington's next moves," said one of the people. "The fact that all current restrictive guidance from various regulators remain informal provides some room for future changes." Despite some Washington officials and lawmakers seeking stronger chip curbs, President Trump has said he could potentially allow a downgraded Nvidia Blackwell chip to be exported to China in addition to H20. Some Chinese tech companies have held off their H20 orders also because they want see if the China specific Blackwell chip, which potentially has better performance than H20, would become available, according to people with knowledge of their thinking. Some Beijing policymakers are pushing to ban foreign chips altogether for inference, which accounts for most AI demand, according to a person recently summoned for a meeting with them. That is unlikely to happen soon due to a shortage of domestic chip supplies, which Beijing hopes to significantly improve by next year when several advanced production lines are scheduled to launch. China's foreign ministry said: "As a matter of principle, science, technology, and economic and trade issues should not be politicised, instrumentalised, or weaponised. Containment and suppression will not hold back China's development." Other Chinese regulators and ministries did not respond to questions for comment. Alibaba and ByteDance did not answer emails requesting comment. The US Commerce department did not immediately respond to a request for comment. Nvidia declined to comment.
[11]
Nvidia asks Foxconn to suspend work for H20 chip, sources say
BEIJING, Aug 22 (Reuters) - Nvidia (NVDA.O), opens new tab has asked Foxconn (2317.TW), opens new tab to suspend work on the H20 AI chip, the most advanced product the U.S. company is currently permitted to sell to China, two people briefed on the matter said. Item 1 of 2 Nvidia logo is seen on graphic card package in this illustration taken August 19, 2025. REUTERS/Dado Ruvic/Illustration/File Photo [1/2]Nvidia logo is seen on graphic card package in this illustration taken August 19, 2025. REUTERS/Dado Ruvic/Illustration/File Photo Purchase Licensing Rights, opens new tab Foxconn, formally known as Hon Hai Precision Industry Co Ltd, handles backend processing work as one of Nvidia's component suppliers for the chip. "We constantly manage our supply chain to address market conditions," Nvidia said in a statement, declining to elaborate further. Reporting by Fanny Potkin in Singapore, Liam Mo in Beijing, Wen-Yee Lee in Taipei and Brenda Goh in Shanghai; Editing by Himani Sarkar Our Standards: The Thomson Reuters Trust Principles., opens new tab
[12]
Nvidia orders suppliers to halt work on China-focussed H20 AI chip, The Information says
Aug 21 (Reuters) - Nvidia (NVDA.O), opens new tab has told some component suppliers to suspend production of its H20 AI chip, designed specifically for the Chinese market, the Information reported on Thursday, citing two people with direct knowledge of the communications. According to the report, Nvidia instructed Arizona-based Amkor Technology (AMKR.O), opens new tab to stop production of the H20 chips this week and also notified South Korea's Samsung Electronics (005930.KS), opens new tab. Amkor handles advanced packaging for the chip, while Samsung Electronics supplies high-bandwidth memory chips for the model. Neither companies immediately responded to a Reuters request for comment. Meanwhile, Nvidia spokesperson said in a statement, "We constantly manage our supply chain to address market conditions." "As both governments recognise, the H20 is not a military product or for government infrastructure. China won't rely on American chips for government operations, just like the U.S. government would not rely on chips from China," it said. This comes as Chinese authorities last week summoned domestic companies, including major internet firms Tencent (0700.HK), opens new tab and ByteDance, over their H20 chip purchases, expressing concerns over information risks. Reporting by Yazhini MV in Bengaluru; Editing by Sumana Nandy Our Standards: The Thomson Reuters Trust Principles., opens new tab
[13]
Nvidia says it's evaluating a 'variety of products' after report of new AI China chip
Nvidia said Tuesday that it is evaluating several products following a report that the company is working on a new artificial intelligence chip for China that is more powerful than the currently available H20. The new product, tentatively called the B30A, is expected to be based on Nvidia's Blackwell chip architecture, Reuters reported, citing people familiar with the company's plans. Nvidia hopes to deliver sample units to Chinese clients for testing as soon as next month, according to Reuters. "We evaluate a variety of products for our roadmap, so that we can be prepared to compete to the extent that governments allow," the company told CNBC in a statement. "Everything we offer is with the full approval of the applicable authorities and designed solely for beneficial commercial use." Commerce Secretary Howard Lutnick on Tuesday praised Nvidia CEO Jensen Huang and said he wouldn't be surprised if Huang wants to sell a new chip to China. "I'm sure he's pitching the president all the time," Lutnick said in an interview with CNBC's "Squawk on the Street." "I've listened to him pitch the president, and the president listens to our great technology companies, and he'll decide how he wants to play it."
[14]
Nvidia working on new AI chip for China that outperforms the H20, Reuters reports
Nvidia is developing a new AI chip for China based on its latest Blackwell architecture. Nvidia is developing a new AI chip for China based on its latest Blackwell architecture that will be more powerful than the H20 model it is currently allowed to sell there, two people briefed on the matter said. U.S. President Donald Trump last week opened the door to the possibility of more advanced Nvidia chips being sold in China. But the sources noted U.S. regulatory approval is far from guaranteed amid deep-seated fears in Washington about giving China too much access to U.S. artificial intelligence technology. The new chip, tentatively known as the B30A, will use a single-die design that is likely to deliver half the raw computing power of the more sophisticated dual-die configuration in Nvidia's flagship B300 accelerator card, the sources said. A single-die design is when all the main parts of an integrated circuit are made on one continuous piece of silicon rather than split across multiple dies. The new chip would have high-bandwidth memory and Nvidia's NVLink technology for fast data transmission between processors, features that are also in the H20 - a chip based on the company's older Hopper architecture. The chip's specifications are not completely finalized but Nvidia hopes to deliver samples to Chinese clients for testing as early as next month, said the sources who were not authorized to speak to media and declined to be identified. Nvidia said in a statement: "We evaluate a variety of products for our roadmap, so that we can be prepared to compete to the extent that governments allow." "Everything we offer is with the full approval of the applicable authorities and designed solely for beneficial commercial use," it said. The U.S. Department of Commerce did not respond to a Reuters request for comment. The extent to which China, which generated 13% of Nvidia's revenue in the past financial year, can have access to cutting-edge AI chips is one of the biggest flashpoints in U.S.-Sino trade tensions. Nvidia only received permission in July to recommence sales of the H20. It was developed specifically for China after export restrictions were put in place in 2023, but the company was abruptly ordered to stop sales in April. Trump said last week he might allow Nvidia to sell a scaled-down version of its next-generation chip in China after announcing an unprecedented deal that will see Nvidia and rival AMD give the U.S. government 15% of revenue from sales of some advanced chips in China. A new Nvidia chip for China might have "30% to 50% off", he suggested in an apparent reference to the chip's computing power, adding that the H20 was "obsolete". U.S. legislators, both Democratic and Republican, have worried that access to even scaled-down versions of flagship AI chips will impede U.S. efforts to maintain its lead in artificial intelligence. But Nvidia and others argue that it is important to retain Chinese interest in its chips -- which work with Nvidia's software tools -- so that developers do not completely switch over to offerings from rivals like Huawei. Huawei has made great strides in chip development, with its latest models said to be on par with Nvidia in some aspects like computing power, though analysts say it lags in key areas such as software ecosystem support and memory bandwidth capabilities. Complicating Nvidia's efforts to retain market share in China, Chinese state media have also in recent weeks alleged that the U.S firm's chips could pose security risks, and authorities have cautioned Chinese tech firms about purchasing the H20. Nvidia says its chips carry no backdoor risks. Nvidia is also preparing to start delivering a separate new China-specific chip based on its Blackwell architecture and designed primarily for AI inference tasks, according to two other people familiar with those plans. Reuters reported in May that this chip, currently dubbed the RTX6000D, will sell for less than the H20, reflecting weaker specifications and simpler manufacturing requirements. The chip is designed to fall under thresholds set by the U.S. government. It uses conventional GDDR memory and features memory bandwidth of 1,398 gigabytes per second, just below the 1.4 terabyte threshold established by restrictions introduced in April that led to the initial H20 ban. Nvidia is set to deliver small batches of RTX6000D to Chinese clients in September, said one of the people.
[15]
NVIDIA is reportedly developing an AI chip for China more powerful than the H20
NVIDIA is working on a new AI chip meant for the Chinese market that's more powerful than the H20, according to Reuters. It will reportedly be based on the company's latest Blackwell architecture, which can produce chips between seven and 30 times faster than its previous AI platform. Reuters says the product is tentatively named B30A and will have a single-die design, putting all its main components on a single piece of silicon. It will apparently be capable of half the computing power of NVIDIA's Blackwell Ultra GPUs, which have dual-die configuration. The product will also come with high-bandwidth memory and the company's NVLink technology for speedier data transmission between processors. It's possible that NVIDIA is developing the chip after the Chinese government discouraged local companies from using the H20, especially for government and national security purposes. Chinese regulators even reportedly ordered big tech corporations, including Alibaba, Bytedance and Tencent, to suspend their purchases from NVIDIA until the government is done with a national security review. China's warning to local companies came after the US government lifted its export restriction on the company's H20 chips. If you'll recall, the US blocked the company from selling its H20 chips to China back in April over concerns that the Chinese military could use the chips to develop AI technology. In July, NVIDIA announced that the government has assured the company that it will approve licenses to ship and export H20 chips to China. The Financial Times then reported in August that the government had agreed to grant NVIDIA (and AMP) export licenses in exchange for 15 percent of their profits. Reuters says NVIDIA is still finalizing the specs of B30A, but it's hoping to deliver samples to Chinese clients for testing as soon as September this year. Trump seems to already be aware that NVIDIA is working on a chip based on Blackwell for the Chinese market, but whether it gets regulatory and export approval remains to be seen: The president told reporters that NVIDIA CEO Jensen Huan is "coming to see [him] again about that."
[16]
China reportedly discouraged purchase of NVIDIA AI chips due to 'insulting' Lutnick statements
Chinese regulators reportedly dissuaded local companies from purchasing NVIDIA's H20 chips, because they found certain statements by US commerce secretary Howard Lutnick "insulting." According to the Financial Times, the Cyberspace Administration of China (CAC), the National Development and Reform Commission (NDRC) and the Ministry of Industry and Information Technology (MIIT) teamed up to intensify their efforts to push the use of homegrown chips following Lutnick's remarks in an interview with CNBC. The US, if you'll recall, blocked NVIDIA from selling its H20 chips to China back in April out of concern that the Chinese military would use them to develop AI technology. When the US government reversed its decision in July and allowed the company to start shipping its chips to China, Lutnick told CNBC: "We don't sell them our best stuff, not our second best stuff, not even our third best. The fourth one down, we want to keep China using it... The idea is the Chinese are more than capable of building their own. You want to keep one step ahead of what they can build, so they keep buying our chips. You want to sell the Chinese enough that their developers get addicted to the American technology stack. That's the thinking." To note, a previous Times report stated that the government allowed NVIDIA to ship its products to China again after agreeing to hand over 15 percent of its profits. As a response to Lutnick's remarks, the Times says Chinese authorities sought ways to prevent local companies from buying H20 chips. CAC issued an informal notice instructing China's biggest tech firms, such as ByteDance and Alibaba, to stop new orders for H20 chips until the government is done conducting a national security review. The companies are compelled to comply, because they could face substantial fines from the CAC if they don't. Meanwhile, NDRC also issued an informal notice, asking local tech companies not to purchase any NVIDIA chip. Reuters recently reported that NVIDIA is developing a new chip for the Chinese market that's more powerful than the H20, perhaps driven in part by China's move to discourage its purchase. It will be based on the company's Blackwell architecture, but will only be capable of half the computing power of NVIDIA's Blackwell Ultra GPUs. Their regulatory and export approval aren't guaranteed, but the president previously implied that he was aware of the project and said he expects NVIDIA CEO Jensen Huang to talk to him about it.
[17]
Nvidia looking to halt H20 chip production after China cracks down on purchases, reports say
An Nvidia chip is seen through a magnifying glass in Beijing, China, on August 1, 2025. Nvidia has asked some of its component suppliers to stop production related to its made-for-China H20 general processing units, as Beijing cracks down on the American chip darling, The Information reported Friday. The directive comes weeks after the Chinese government told local tech companies to stop buying the chips due to alleged security concerns, the report said, citing people with knowledge of the matter. Nvidia reportedly as asked Arizona-based Amkor Technology, which handles the advanced packaging of Nvidia's H20 chips, and South Korea's Samsung Electronics, which supplies memory for them to halt production. Samsung and Amkor did not immediately respond to CNBC's request for comment. A separate report from Reuters, citing sources, said that Nvidia had asked Foxconn to suspend work related to the H20s. Foxconn did not immediately respond to a request for comment. In response to an inquiry from CNBC, an Nvidia spokesperson said "We constantly manage our supply chain to address market conditions." The news further throws the return of the H20s to the China market in doubt, after Washington said it would issue export licenses, allowing the chip's exports to China -- whose shipment had effectively been banned in April. Last month, the Cyberspace Administration of China had summoned Nvidia regarding national security concerns with the H20s and had asked the company to provide information on the chips. Beijing has raised concerns that the chips could be have certain tracking technology or "backdoors," allowing them to be operated remotely. U.S. lawmakers have proposed legislation that would require AI chips under export regulations to be equipped with location-tracking systems to avoid their illegal shipments. Speaking to reporters in Taiwan on Friday, Nvidia CEO Jensen Huang acknowledged that China had asked questions about security "backdoors," and that the company had made it clear they do not exist. "Hopefully the response that we've given to the Chinese government will be sufficient. We're in discussions with them," he said, adding that Nvidia had been "surprised" by the queries. "As you know, [Beijing] requested and urged us to secure licenses for the H20s, for some time and I've worked quite hard to help them secure the licenses, and so hopefully this will be resolved," he said. Nvidia in a statement on Friday said "The market can use the H20 with confidence." It added: "As both governments recognize, the H20 is not a military product or for government infrastructure. China won't rely on American chips for government operations, just like the U.S. government would not rely on chips from China. However, allowing U.S. chips for beneficial commercial business use is good for everyone." Last month, Nvidia had reportedly sent notices to major tech companies and AI developers urging them against the use of the H20s, in what first had appeared as a soft mandate. The Information later reported that Beijing had told some firms, including ByteDance, Alibaba and Tencent, to halt orders of the chips altogether, until the completion of a national security review. It had been seen as a major win for Nvidia when Huang announced last month that the U.S. government would allow sales of the company's H20 chips to China. However, the national security scrutiny the H20s are now facing from the Chinese side, highlights the difficulties of navigating Nvidia's business through increasing tensions and shifting trade policy between Washington and Beijing. Chip industry analysts have also said Beijing's actions appear to reinforce its commitment to its own chip self-sufficiency campaigns and its intention to resist the Trump administration's plan to keep American AI hardware dominant in China.
[18]
NVIDIA reportedly stops production of H20 AI chips
NVIDIA has reportedly asked its suppliers to halt production related to its H20 AI chips for the Chinese market. According to The Information, the company told Arizona-based Amkor Technology and Samsung Electronics to put a pause on their work for the H20. Amkor produces advanced packaging for the H20 chips, while Samsung supplies memory for NVIDIA. Reuters has also reported that NVIDIA asked Foxconn, which is in charge of backend processing for the chip, to suspend its work. "We constantly manage our supply chain to address market condition," the company told CNBC in a statement when asked to comment about the supposed production pause. The US government had blocked NVIDIA from selling the H20 in China back in April, out of concerns that the country could use it to develop AI tech for its military. It allowed the company to resume selling the chip in China by July, reportedly after closing a deal that would give it 15 percent of the sales. But China didn't welcome the H20 with open arms. Local regulators instructed the biggest Chinese tech companies, including ByteDance and Alibaba to stop new orders for H20 chips, citing security concerns. The Cyberspace Administration of China talked to NVIDIA, claiming that AI experts had revealed that the chips could be tracked and controlled remotely. NVIDIA CEO Jensen Huang had admitted that Chinese regulators asked him about the supposed "backdoor" and said that he made it clear it didn't exist. "Hopefully the response that we've given to the Chinese government will be sufficient," Huang said. A recent report by the Financial Times, however, claimed that Chinese authorities didn't issue warnings against using NVIDIA chips just because of security concerns. Apparently, they found certain remarks by US commerce secretary Howard Lutnick "insulting." When the US allowed shipments of the H20 to China again, Lutnick said during an interview: "We don't sell them our best stuff, not our second best stuff, not even our third best. The fourth one down, we want to keep China using it... You want to sell the Chinese enough that their developers get addicted to the American technology stack." The H20 is currently the most advanced AI chip NVIDIA can sell in the Chinese market, but the company is reportedly developing a more powerful product. It will be based on the company's Blackwell architecture, Reuters previously reported, and will be capable of half the computing power of NVIDIA's Blackwell Ultra GPUs.
[19]
Nvidia reportedly designing powerful new AI chip for China
Nvidia is reportedly working on a new AI chip for China that is more powerful than the H20 chip it currently sells to the country. The new product is tentatively called the B30A, Reuters reported, citing people familiar, and it is thought to be based on Nvidia's Blackwell chip architecture. When asked about the new chip in a CNBC interview, Commerce Secretary Howard Lutnick did not rule out allowing Nvidia's chief executive Jensen Huang to sell the more powerful product to China. "Of course, he would like to sell a new chip to China," Lutnick said. "I've listened to him pitch the president, and the president listens to our great technology companies, and he'll decide how he wants to play." In April, the U.S. Commerce Department banned Nvidia from selling its H20 chips to China, citing the need to safeguard "national and economic security." Last year, Nvidia booked roughly $17 billion from the Chinese market -- about 13% of its total revenue. But the export ban imposed by the Trump administration in April put it on the defensive, with Nvidia taking a $5.5 billion write-down on stranded H20 inventory and warning of billions in lost sales. The Trump administration reversed its ban in July. The reversal was part of a broader de-escalation between the two countries ahead of a potential fall summit between Trump and Chinese President Xi Jinping. Lutnick has previously said the administration changed course so that China would still be "addicted" to using U.S. technology. Earlier in August, the U.S. government announced a deal that would have Nvidia and another chipmaker, Advanced Micro Devices, pay 15% of their revenues from sales in China, in return for being able to sell chips there. "We evaluate a variety of products for our roadmap, so that we can be prepared to compete to the extent that governments allow," Nvidia said in a statement. "Everything we offer is with the full approval of the applicable authorities and designed solely for beneficial commercial use."
[20]
Nvidia reportedly paused production on H20 chip for China
Nvidia has already suspended production of its H20 AI chip meant for China, the Information reported Thursday. The chip designer told some suppliers -- including Amkor Technology and Samsung Electronics -- to halt production on its H20 AI chip this week, the outlet said, citing people familiar with the matter. The apparent pause comes one month after China's Cyberspace Administration raised security concerns over alleged "backdoors" in Nvidia's H20 chips, per CNBC. Nvidia's stock has dropped 1.9% over the last five days. The reported pause also follows a Tuesday report from Reuters that said Nvidia is working on a new AI chip for China that's more powerful than the H20 chip. The new product is tentatively called the B30A, Reuters reported. It is thought to be based on Nvidia's Blackwell chip architecture. When asked about the new chip in a CNBC interview, Commerce Secretary Howard Lutnick did not rule out allowing Nvidia's chief executive Jensen Huang to sell the more powerful product to China. "Of course, he would like to sell a new chip to China," Lutnick said. "I've listened to him pitch the president, and the president listens to our great technology companies, and he'll decide how he wants to play." In April, the U.S. Commerce Department banned Nvidia from selling its H20 chips to China, citing the need to safeguard "national and economic security." The Trump administration reversed its ban in July. Last year, the U.S. tightened chip-related export rules. In response, Nvidia launched its H20 chip, a lower-spec version of its flagship GPU designed to fall just under those limits. -- Alex Daniel contributed to this article.
[21]
Why is China turning its back on AI chip giant Nvidia?
China reportedly told local tech companies to stop buying Nvidia's chips due to alleged security concerns, as the global AI race intensifies. Chip giant Nvidia has told some of its component suppliers to stop production of its H20 artificial intelligence chip, designed specifically for China, according to a report. The move comes after Chinese authorities told local tech companies, including Tencent and ByteDance, to stop buying the semiconductors several weeks ago due to Beijing having alleged security concerns. Nvidia told Arizona-based Amkor Technology to halt production of the H20 chips this week and also notified South Korea's Samsung Electronics, according to a report by The Information, which cited two people familiar with the matter. The United States banned Nvidia from selling its most powerful chips, the Blackwell chip, to China in April, arguing it is necessary to safeguard US national and economic security as the AI global race gains pace. In July, the Trump Administration reversed its decision on H20 chips, which is not Nvidia's most powerful chip, largely due to trade negotiations between Washington and Beijing. China's Cyberspace Administration last month summoned Nvidia over national security concerns related to the H20s and asked the company to give information about the chips. China's internet regulator CAC claimed that US AI experts had said Nvidia's chips have location tracking and can be shut down remotely. Nvidia denies the allegations. Nvidia CEO Jensen Huang told reporters on Friday that China had asked about security "backdoors," and that the firm had said they do not exist. "Hopefully the response that we've given to the Chinese government will be sufficient. We're in discussions with them," CNBC quoted him as saying, adding that Nvidia had been "surprised" by the queries. An Nvidia spokesperson said in a statement quoted by media that: "We constantly manage our supply chain to address market conditions". "As both governments recognise, the H20 is not a military product or for government infrastructure. China won't rely on American chips for government operations, just like the US government would not rely on chips from China," it said. Huang also said on Friday that Nvidia is discussing a potential new computer chip designed for China with the Trump administration. Huang was asked about a possible "B30A" semiconductor for AI data centres for China. "I'm offering a new product to China for ... AI data centres, the follow-on to H20," Huang said in Taiwan. But he added that "That's not our decision to make. It's up to, of course, the United States government. And we're in dialogue with them, but it's too soon to know". The Financial Times also reported this week that Chinese regulators turned against Nvidia due to "insulting" remarks made by the US Commerce Secretary Howard Lutnick. "We don't sell them our best stuff, not our second-best stuff, not even our third-best," Lutnick told CNBC in July. However, it may take a few years for China to develop its own domestic AI chip that rivals Nvidia. Nvidia is the world's leading AI chip supplier. Chinese firms such as Alibaba and ByteDance have said their AI development would be hindered without Nvidia's chips. However, China is trying to promote chip sovereignty, and companies such as Huawei are trying to rival Nvidia.
[22]
China has reportedly told its data center operators to source more than 50% of their chips from domestic manufacturers in an effort to break away from US tech
The Chinese government has reportedly mandated that its domestic data center operators nationwide should source more than 50 per cent of their chips from domestic producers. The mandate is said to have originated in guidelines proposed last March for the Shanghai municipality, which stipulated that "adoption of domestic computing and storage chips at the city's intelligent computing centres should be above 50 per cent by 2025" (via the South China Morning Post). According to the SCMP, a source working as an adviser in the data center industry told the outlet that the Shanghai chip quotas for the city's data centers had since become a mandatory nationwide policy. In the great AI race, China is often viewed to be somewhat behind when it comes to computing power, despite efforts in recent years to build more than 500 new data centers across the country. Sources have told the SCMP that while Chinese chips are considered "usable" in inference training for AI models within these facilities, Nvidia chips are still the go-to choice. While the US government has recently decided to grant licenses to Nvidia to sell its H20 GPUs to the country once more, a requirement of 50% or more all-Chinese chips would likely affect potential sales -- although given the rate of expansion and the apparent popularity of Nvidia hardware in the country, I'd say it was still likely to shift a fair few units to Chinese shores. The SCMP also reports that data centers are facing adaptation challenges in integrating Nvidia's hardware with domestic solutions. Nvidia's AI GPUs make use of Nvidia's CUDA software ecosystem, while Chinese models often use Huawei CANN or similar. Integrating the two together is apparently quite the technical challenge, particularly if firms wish to min/max the number of faster Nvidia chips they're allowed to use. Despite these issues, it should be noted that Chinese developers were still able to create DeepSeek's AI models under previous chip restrictions, an open-source alternative that shook the industry considerably at the start of this year. While China may be behind in the hardware game, it doesn't appear to have held it back as much as advocates for the previous US sanctions may have hoped, with Nvidia CEO Jensen Huang openly praising the Chinese AI industry last month. In reference to fears of the Chinese military advancing its tech with US AI hardware, Huang said: "There's plenty of computing capacity in China already. If you just think about the number of supercomputers in China, built by amazing Chinese engineers, that are already in operation." "They don't need Nvidia's chips, certainly, or American tech stacks, in order to build their military." How much of China's current AI output is being trained and run on Chinese hardware, compared to US equivalents, is currently unclear. Still, this mandate looks to be an attempt by the Chinese government to put curbs on a potential future US tech dominance within its AI industry, and ensure that Chinese AI hardware remains at the forefront of its AI development.
[23]
After months of back and forth, Nvidia has reportedly paused making its H20 chip due to security concerns, this time from China
Nvidia's H20, an AI and compute chip specifically modified for use in the Chinese market, has reportedly stopped production in the midst of security concerns from Chinese officials. As reported by the Information, and citing "two people with direct knowledge of the communications", Nvidia has reportedly "told some of its component suppliers to suspend production work." Amkor Technology and Samsung are two companies which were reportedly told about this new development. But first, a quick rundown on Nvidia's H20 chip. It is effectively a less powerful, modified version of the H100 GPU, a large-scale processor that is used for AI workloads, data analysis, and more. In April, Nvidia was set to lose $5.5 billion due to trade restrictions on chips to China. A month after this, export controls made Nvidia write the chip off entirely. Then, by July, the US government had signed off on Nvidia selling its chips in the Chinese market. In return, the US would see 15% of the revenue made from these sales. This was promptly hit by concerns from the Chinese government about backdoors in chips (this was suggested by some US politicians). Nvidia was adamant that there are no backdoors in its chips, but still, the Cyberspace Administration of China halted the sale of H20 chips due to security concerns. Just last week, China reportedly told data center operators to source the majority of its chips from domestic manufacturers instead. This is part of a broader plan to move away from relying on US tech. According to the Information, ByteDance, Alibaba Group, and Tencent Holdings were among the firms pushed just last week to suspend any purchases of Nvidia chips. It's been a wild couple of months for Nvidia when it comes to this specific chip. It reportedly saw demand accounting for 700,000 H20 chips after declaring it could go on sale, which is why the halting of production feels like a particularly strong blow. Nvidia's Jensen Huang has made it clear his desire to provide chips to China. In response to the original chip export controls, he said, "If they don't have enough Nvidia, they will use their own." He has also said that Chinese competitors are "quite formidable", whilst also arguing that the Chinese military wouldn't use US chips as "they simply can't rely on it." Effectively, he doesn't seem to think that Nvidia is giving China an edge when it comes to the AI land grab, and the potential of extra cash from the region certainly couldn't hurt Nvidia or its stockholders' view of the company.
[24]
U.S. Sanctions Push Nvidia to Suspend H20 GPU Supply to China
Nvidia's plans for its H20 graphics chips in China are facing another major disruption. According to reports from The Information and Reuters, the company has asked several of its partners to stop work on the chip. Amkor Technology, which is responsible for packaging the processors, Samsung, which supplies the memory modules, and Foxconn, which handles back-end assembly, have all reportedly been told to suspend their involvement. This means the H20, Nvidia's latest China-focused chip, may be delayed even further -- or potentially shelved altogether. The timing of this move is tied closely to U.S. trade policy. Back in April, the U.S. government added new export rules that targeted the H20 GPU. Nvidia was required to apply for a license before sending the chips to Chinese customers. In July, the Biden administration agreed to loosen those restrictions, but only in exchange for a deal that would allow the U.S. to capture part of Nvidia's revenue from Chinese sales. Nvidia has not publicly confirmed the production halt but told Reuters that it is actively managing its supply chain to adjust to changing market conditions. For Nvidia, the H20 was meant to be a workaround to continue selling chips in China despite earlier bans on high-performance GPUs like the A100 and H100. The H20 was designed with some reduced specifications to comply with U.S. export thresholds but still deliver solid performance for AI and data center use in China. However, these ongoing policy shifts make it increasingly difficult for Nvidia to maintain a stable strategy in the Chinese market. At the same time, political tensions between Washington and Beijing are creating additional challenges. Earlier this month, the Chinese government reportedly advised state agencies and companies tied to national security not to adopt the H20. The advice was reportedly motivated in part by comments from U.S. Commerce Secretary Howard Lutnick. In an interview with CNBC, Lutnick stated that the U.S. does not sell "even its third-best products" to China, suggesting that the goal is to limit China's access to advanced technology while keeping its developers dependent on U.S. hardware. These remarks reportedly frustrated officials in Beijing and may have played a role in discouraging adoption of the chip. For Nvidia, this leaves the H20 in a precarious position. On one hand, the company faces U.S. regulations that complicate exports and require ongoing negotiation with Washington. On the other hand, China is signaling a reluctance to rely on a product that is both restricted and politically sensitive. Even if restrictions were lifted tomorrow, Chinese customers may already be looking for alternatives, whether in domestic chipmakers or other suppliers willing to sidestep U.S. controls. This situation highlights a broader trend: the growing entanglement of advanced semiconductor technology with geopolitics. Chips like Nvidia's H20 are not just pieces of hardware -- they are tied directly to national strategies, export policies, and the global balance of technological power. Nvidia is caught in the middle, trying to maintain one of its largest markets while navigating the political expectations of Washington and Beijing. Company Supply‑Chain Role Tasks Reportedly Halted Reported By Notes / Context Amkor Technology OSAT / packaging provider Chip packaging (and related assembly steps) for H20 The Information Asked to suspend work on H20 units intended for China. Samsung Memory supplier Memory supply and integration for H20 The Information Reportedly told to halt H20-related memory work. Foxconn Back‑end processing / final assembly & test Back‑end processing for H20 Reuters Sources say Nvidia asked for suspension of H20 back‑end work. Nvidia H20 Supply-Chain Halt Summary The future of the H20 remains uncertain. Nvidia may continue to look for compromises that allow limited sales in China, but the delays caused by supplier suspensions could seriously undermine the chip's rollout. Meanwhile, Chinese tech companies may double down on efforts to design their own accelerators and reduce dependency on U.S. suppliers. For now, the H20 illustrates how quickly the market for advanced chips can shift when government policy collides with global supply chains.
[25]
China pushing to ban foreign chips from being used for AI inference, but need NVIDIA for now
TL;DR: China is reportedly pushing to ban foreign AI inferencing chips, especially from NVIDIA, amid a domestic supply shortage. Beijing aims to boost local production by 2026 while responding to U.S. tech restrictions and equity moves. This shift impacts Chinese tech firms' AI chip orders and reflects rising geopolitical tech tensions. China is reportedly pushing for the ban of foreign chips made by any company -- namely NVIDIA -- for use in inferencing, which accounts for the most AI demand, according to new reports. In an article from the well-trusted outlet the Financial Times, it's being reported that according to someone who was "recently summoned" to meet with Beijing policymakers in China are pushing for a ban on foreign-made chips used for AI inferencing, but it won't happen any time soon. There is a major shortage of domestic AI chip supplies in China, but Beijing hopes to "significantly improve" that in 2026 with multiple advanced production lines coming online. China's foreign ministry said: "As a matter of principle, science, technology, and economic and trade issues should not be politicized, instrumentalized, or weaponized. Containment and suppression will not hold back China's development". The Trump administration and US Commerce Secretary Howard Lutnick is currently mulling equity stakes in big overseas technology and semiconductor companies like TSMC, Samsung, SK hynix, and more, so appropriate countermoves are happening in China as a reaction to these changes. The US government also just took a 10% stake in US-based chipmaker Intel, with hopes to see it return to form in the semiconductor industry on American soil. Some of the comments that Lutnick made at the White House recently have ruffled some feathers in China, with the Financial Times reporting from people with "knowledge of the regulatory action" that the Cyberspace Administration of China (CAC), the National Development and Reform Commission (NDRC) and the Ministry of Industry and Information Technology (MIIT) "moved in response to comments made by Lutnick last month". Lutnick told CNBC during a July 15 interview: "we don't sell them our best stuff, not our second-best stuff, not even our third-best. You want to sell the Chinese enough that their developers get addicted to the American technology stack, that's the thinking". The FT reports that some of China's senior leaders found the comments "insulting", something that saw policymakers in Beijing look for ways to restrict Chinese technology companies from buying the AI chips, according to two people with knowledge of the latest regulatory decision-making. As a result of this, it's reported that Chinese tech giants have "held off" or "significantly downsized" their H20 orders, according to FT's two sources that have knowledge of the plans.
[26]
NVIDIA suspends production on H20 chip over security concerns
TL;DR: NVIDIA has halted production of its H20 AI chip for the Chinese market amid ongoing US-China trade restrictions and security concerns. Despite a temporary export license, China's Cyberspace Administration blocked sales, prompting NVIDIA to instruct suppliers Amkor Technology and Samsung to stop manufacturing, impacting major Chinese tech firms. NVIDIA has reportedly halted production of its H20 AI chip that is designed specifically for the Chinese market, with NVIDIA sending notifications to Arizona-based Amkor Technology and Samsung Electronics to stop manufacturing. NVIDIA's AI chips, specifically the H20, are yet again up in the air. The sale of AI chips to China has been nothing short of a fiasco, with NVIDIA initially set to lose a staggering $5.5 billion over trade restrictions implemented on China for high-powered AI chips. These trade restrictions were set to ensure the US remains at the top of the global AI race, and at the time cited national security risks as China would use NVIDIA's GPUs to develop a sophisticated AI that the Chinese military could then utilize. Only a month later, the H20, which is a heavily modified version of the H100 GPU, was banned from entering China. However, by July, the US government had changed its stance on NVIDIA chips sold to China. NVIDIA was able to secure a license, which would approve exporting the H20 to China, provided that 15% of the revenue from these sales would be directly allocated to the US government. The flip-flopping of these decisions has now raised security concerns about the chips from Chinese government officials. NVIDIA has assured all parties that there aren't any backdoors in any of its chips, but those statements didn't convince the Cyberspace Administration of China, which has now prevented the sale of H20 chips over possible security threats. While this back and forth continues between the US government, China, and NVIDIA, the Chinese government is pushing its biggest chip-hungry companies to source the hardware they are after domestically, with The Information reporting companies such as ByteDance, Alibaba Group, and Tencent Holdings have been told to suspend purchases of NVIDIA chips. The Information report cites "two people with direct knowledge of the communications," who said NVIDIA has "told some of its component suppliers to suspend production work." Those two companies were Amkor Technology, which handles the advanced packaging for the chip, and Samsung Electronics, which supplies the high-bandwidth memory (HBM) modules for the GPU.
[27]
Nvidia Considering 'Variety of Products' as New China AI Chip Details Emerge
Kara Greenberg is a senior news editor for Investopedia, where she does work writing, editing, and assigning daily markets and investing news. Prior to joining Investopedia, Kara was a researcher and editor at The Wire. Earlier in her career, she worked in financial compliance and due diligence at Loomis, Sayles & Company, and The Bank of New York Mellon. Nvidia (NVDA) said it's considering a "variety of products" for its roadmap as more details emerge about a new AI chip for the Chinese market. This follows recent signs of inroads in winning approvals from the Trump administration to expand sales to the country. Nvidia could look to deliver samples of the new chip to Chinese clients for testing as early as next month, Reuters reported Tuesday. The new chip, tentatively known as the B30A, is based on Nvidia's Blackwell architecture and more advanced than the H20 chip the company recently won approval to resume selling in China after striking a deal with President Trump to pay the U.S. government 15% of its China chip revenue. A Nvidia spokesperson told Investopedia Tuesday that the company wants to be "prepared to compete to the extent that governments allow," adding "everything we offer is with the full approval of the applicable authorities and designed solely for beneficial commercial use." Shares of Nvidia fell nearly 3% in recent trading. Still, they've added nearly a third of their value this year, after their recent rally amid optimism about the company's sales in China and strong demand for AI hardware. While many on Wall Street saw the recent 15% revenue-sharing agreements for Nvidia and Advanced Micro Devices (AMD) as a positive development for the companies -- which took a hit on export curbs earlier in the year -- some have raised concerns about what they could mean for American tech companies navigating rapidly changing China trade policies. "It feels like a slippery slope to us," Bernstein analysts wrote last week of the deals, which could open the door to others and pointed to a lack of clarity on whether and how much companies could be pushed to pay to sell such products in China. "Sure it might raise some money, but doesn't seem to address any strategic issues beyond a grab for dollars," the analysts said. Last week, Trump suggested he could be open to expanding Nvidia's licenses to cover more powerful chips, after further negotiations. Nvidia CEO Jensen Huang could be asked for more details on the company's plans when the chipmaker reports its latest quarterly financial results after the bell Aug. 27.
[28]
Nvidia Halts China Chip Production as Beijing Warns Over Security, Reports Say
Earlier this month, Nvidia and rival Advanced Micro Devices were given the approval to resume sales of key AI chips to China, with the condition that they pay 15% of their chip revenue generated there to the U.S. government in exchange for the export licenses. Nvidia (NVDA) reportedly has told suppliers to suspend production of its H20 chip, after Beijing asked local firms to avoid using the chip tailored for the Chinese market due to security concerns. Citing unidentified sources, The Information reported that Nvidia has instructed Samsung Electronics and Amkor Technology to halt production of the H20 chip, which are less powerful than its latest semiconductors. Reuters separately reported that Nvidia had asked Foxconn to suspend work related to the H20 chips. Foxconn, Samsung, and Amkor didn't immediately respond to requests for comment. "We constantly manage our supply chain to address market conditions," an Nvidia spokesperson told Investopedia. Last month, Nvidia and rival Advanced Micro Devices (AMD) were given approval from the Trump administration to resume sales of key AI chips to China, with the condition that they pay 15% of their chip revenue generated there to the U.S. government in exchange for the export licenses. Beijing reportedly has raised concerns that the Nvidia chips contained "backdoors," allowing remote access to or control of the chips, a charge the tech firm has denied. "As both governments recognize, the H20 is not a military product or for government infrastructure. China won't rely on American chips for government operations, just like the U.S. government would not rely on chips from China," the Nvidia spokesperson said. "However, allowing U.S. chips for beneficial commercial business use is good for everyone." The spokesperson added, "Cybersecurity is critically important to us. NVIDIA does not have 'backdoors' in our chips that would give anyone a remote way to access or control them. The market can use the H20 with confidence." China is a key market for Nvidia. The chipmaker said in May that it took a $4.5 billion charge in the first quarter due to export curbs on H20 chips to the Asian country. Nvidia shares, which entered Friday up 30% this year, are falling 1% in premarket trading.
[29]
Nvidia stocks in focus as Jensen Huang asks suppliers to halt H20 chip work
Nvidia has asked suppliers to halt production of its China-focused H20 AI chip. The move comes as Beijing urges local firms to avoid the product, highlighting how geopolitics keeps colliding with tech. Nvidia shares dipped in pre-market trading Friday, adding pressure just days before its August 27 earnings report. The H20 chip is Nvidia's most advanced product currently allowed for sale in China, and it plays a critical role in powering artificial intelligence applications for both enterprises and government-related projects. The halt raises important questions about the future of AI hardware exports, global supply chains, and the semiconductor industry as a whole. The H20 AI chip represents Nvidia's cutting-edge technology for AI and machine learning applications. It is designed to handle large-scale computations required for data centers, research, and enterprise AI projects. Chinese companies, which are rapidly expanding their AI capabilities, rely on such high-performance chips to compete in areas like cloud computing, autonomous systems, and scientific research. ALSO READ: U.S. stocks moved higher Friday as Dow, S&P 500, and Nasdaq futures jump; optimism remains cautious The strategic significance of the H20 chip is not just technical but geopolitical. As one of the few advanced chips approved for export to China, it bridges the gap between U.S. technological innovation and Chinese AI ambitions. This unique position makes it a sensitive asset in international trade and policy discussions. Nvidia Corp (NVDA) is under pressure Friday as pre-market trading shows shares at $172.79, down 1.25%, after closing Thursday at $174.98 (-0.23%). The decline comes just days before its highly anticipated August 27 earnings report, which Wall Street expects to show a 48% revenue surge to $45.9 billion. ALSO READ: Crypto Crash Alert: BTC, ETH, XRP slide -- will the Fed trigger the next crypto boom? The immediate drag stems from reports that Nvidia has told key suppliers -- including TSMC, Samsung, and Amkor -- to pause production of its China-specific H20 AI chip. The decision follows tighter U.S. export controls and growing security concerns, sparking fresh questions about Nvidia's China strategy. That news shaved about 1-1.3% off pre-market trading, setting a cautious tone for the day. Beyond the headline, it underscores a bigger risk: AI chipmakers are no longer driven solely by demand, but also by geopolitics and regulation. Investors worry this could ripple across the tech sector, particularly for companies tied closely to Nvidia's supply chain. For investors, Nvidia remains the bellwether of the AI trade. Its stock not only drives the Nasdaq and S&P 500 tech components, but also sets sentiment across the AI megacap sector. That's why every development -- whether regulatory, geopolitical, or earnings-driven -- matters beyond just one company. At the moment, Nvidia's futures are reflecting geopolitical risk premiums, suggesting cautious trading ahead of the weekend. With the earnings countdown ticking, Nvidia isn't just another stock in play -- it's the pulse check for the entire AI-driven market cycle. Beijing recently advised domestic firms to refrain from using Nvidia's H20 processors, particularly in projects linked to government or critical infrastructure. The guidance appears to be part of China's broader strategy to strengthen its technological independence and assert control over advanced computing resources. Nvidia responded by temporarily halting H20 chip production, a precautionary measure that ensures the company remains compliant with evolving regulatory frameworks. By pausing manufacturing, Nvidia avoids potential conflicts with both Chinese policies and U.S. export regulations, navigating a delicate landscape where commercial interests and national security concerns intersect. To implement the halt, Nvidia has communicated directly with key suppliers, including Foxconn and Amkor Technology, requesting a suspension of H20 chip production. This step underscores Nvidia's commitment to regulatory compliance while protecting its market position. Company leadership has noted that shipments of the H20 chip to China do not pose a security threat. From a business perspective, maintaining good relations with both the U.S. government and Chinese clients is crucial. Any misstep could disrupt supply chains, damage partnerships, and affect Nvidia's standing in one of the world's largest AI markets. For companies in China and beyond, the halt could create short-term disruptions in AI hardware availability. Enterprises depending on H20 chips for data processing, AI training, and high-performance computing may face delays or need to seek alternative solutions. At the same time, this situation highlights a broader challenge for AI-focused companies globally: relying on a limited number of suppliers or geographic regions for critical technology can introduce risk. Diversifying supply chains and considering alternative chip vendors may become increasingly important strategies. The Nvidia H20 pause is emblematic of the challenges faced by technology companies operating in an increasingly fragmented global market. Semiconductors are at the heart of modern technology -- from AI and cloud computing to defense and telecommunications. However, international regulations, export controls, and geopolitical tensions are increasingly influencing where and how these chips can be produced and sold. Companies must balance innovation with compliance, ensuring that technological progress does not come into conflict with national policies or international diplomacy. Q1: Why did Nvidia halt H20 AI chip production in China? Nvidia paused H20 production to comply with Beijing's guidance and navigate regulatory and geopolitical concerns. Q2: What impact does the H20 AI chip halt have on global AI supply? The halt affects Chinese AI projects and may create short-term supply delays for enterprises relying on Nvidia's advanced AI chips.
[30]
Nvidia orders suppliers to halt work on China-focussed H20 AI chip: Report - The Economic Times
According to the report, Nvidia instructed Arizona-based Amkor Technology to stop production of the H20 chips this week and also notified South Korea's Samsung Electronics .Nvidia has told some component suppliers to suspend production of its H20 AI chip, designed specifically for the Chinese market, the Information reported on Thursday, citing two people with direct knowledge of the communications. According to the report, Nvidia instructed Arizona-based Amkor Technology to stop production of the H20 chips this week and also notified South Korea's Samsung Electronics . Amkor handles advanced packaging for the chip, while Samsung Electronics supplies high-bandwidth memory chips for the model. Neither companies immediately responded to a Reuters request for comment. Meanwhile, Nvidia spokesperson said in a statement, "We constantly manage our supply chain to address market conditions." "As both governments recognise, the H20 is not a military product or for government infrastructure. China won't rely on American chips for government operations, just like the US government would not rely on chips from China," it said. This comes as Chinese authorities last week summoned domestic companies, including major internet firms Tencent and ByteDance, over their H20 chip purchases, expressing concerns over information risks.
[31]
Nvidia orders suppliers to halt work on China-focussed H20 AI chip, The Information says
Nvidia halted production of its H20 AI chip for China. The company instructed Amkor Technology and Samsung Electronics to suspend work. This followed concerns from Chinese authorities regarding information risks. Tencent and ByteDance were among the firms summoned. Nvidia maintains the H20 is not for military or government use. The company manages its supply chain to address market conditions. Nvidia has told some component suppliers to suspend production of its H20 AI chip, designed specifically for the Chinese market, the Information reported on Thursday, citing two people with direct knowledge of the communications. According to the report, Nvidia instructed Arizona-based Amkor Technology to stop production of the H20 chips this week and also notified South Korea's Samsung Electronics . Amkor handles advanced packaging for the chip, while Samsung Electronics supplies high-bandwidth memory chips for the model. Neither companies immediately responded to a Reuters request for comment. Meanwhile, Nvidia spokesperson said in a statement, "We constantly manage our supply chain to address market conditions." "As both governments recognise, the H20 is not a military product or for government infrastructure. China won't rely on American chips for government operations, just like the U.S. government would not rely on chips from China," it said. This comes as Chinese authorities last week summoned domestic companies, including major internet firms Tencent and ByteDance, over their H20 chip purchases, expressing concerns over information risks.
[32]
Why Is Nvidia Stock Gaining Tuesday? - NVIDIA (NASDAQ:NVDA)
Nvidia NVDA is pressing ahead with a China strategy that balances U.S. export restrictions with the need to defend market share in one of its most important regions. The company is developing a new AI chip, tentatively called the B30A, built on its latest Blackwell architecture, Reuters reported on Tuesday. According to people briefed on the matter, the B30A will use a single-die design, delivering about half the raw computing power of the company's flagship dual-die B300 accelerator, but more potent than the H20 model Nvidia is currently allowed to sell in China. Also Read: Beijing Asks Alibaba, ByteDance Why They Need Nvidia H20 Chips Instead Of Local Alternatives The chip will include high-bandwidth memory and NVLink connectivity, both of which appear in the H20, a Hopper-based processor customized for the Chinese market. Nvidia hopes to ship samples to Chinese customers for testing by September. China accounted for 13% of Nvidia's revenue in fiscal 2024. After an abrupt ban in April, Nvidia only regained permission to sell H20 in July. Last week, President Donald Trump signaled that his administration might allow Nvidia to sell scaled-down versions of its most advanced chips in China with 30% to 50% reductions in computing power. NVIDIA's stock gained 36% year-to-date as Big Tech giants like Microsoft MSFT and Alphabet GOOGL keep splurging on their AI ambitions, fueling demand for Nvidia processors. Nvidia and rival Advanced Micro Devices AMD have already agreed to a deal requiring them to hand over 15% of revenue from advanced chip sales in China to the U.S. government. In parallel, Nvidia is preparing to roll out another China-specific processor, the RTX6000D, which is also based on Blackwell but tuned for AI inference. This chip will feature conventional GDDR memory and bandwidth capped at 1,398 GB/s, just under the 1.4 TB/s limit imposed by new U.S. rules. The RTX6000D will also be priced below the H20 and is scheduled for initial deliveries in September. Analysts remain divided on Nvidia's outlook in China. Bernstein warned that U.S. export controls and the rise of local players like Huawei could shrink Nvidia's market share from 66% in 2024 to 54% in 2025, with China's AI chip localization ratio climbing to 55% by 2027. In contrast, Needham's N. Quinn Bolton struck a bullish tone, projecting $3 billion in quarterly H20 shipments with near-100% margins on written-down inventory and forecasting $20 billion in China data center sales by fiscal 2028, supported by strong demand for upcoming Blackwell GPU variants. Price Action: NVDA stock is trading higher by 0.33% to $182.61 premarket at last check Tuesday. Read Next: Semiconductor Stocks Stay In Focus As Massive AI Spending, and China Deal Fuel Growth Hopes Image by El editorial via Shutterstock NVDANVIDIA Corp$182.280.15%Stock Score Locked: Want to See it? Benzinga Rankings give you vital metrics on any stock - anytime. Reveal Full ScoreEdge RankingsMomentum86.47Growth99.23QualityN/AValue6.53Price TrendShortMediumLongOverviewAMDAdvanced Micro Devices Inc$174.33-1.03%GOOGLAlphabet Inc$202.98-0.26%MSFTMicrosoft Corp$516.65-0.09%Market News and Data brought to you by Benzinga APIs
[33]
China's Policymakers Are Reportedly Pushing to Completely Ban NVIDIA's AI Chips in the Country, But Weak Domestic Alternatives Are Holding Them Back
It seems that Beijing is now concerned with the domestic industry's reliance on American AI technology and is exploring newer options. Well, there's no doubt that China heavily relies on AI chips and equipment from the US, mainly because domestic options cannot serve the nation's needs. However, recent events have forced Beijing to counter the influence of NVIDIA's chips in the region, which is why China's Cyberspace Administration opened up a regulatory investigation to determine the presence of security backdoors in the H20 AI accelerators. In a report by Financial Times, it is claimed that China's lawmakers are proposing an 'outright ban' of the H20 AI chip, but it isn't as easy as it might sound. It is claimed that Chinese officials are voicing opposition to NVIDIA's AI chips after the recent remarks by US Commerce Secretary Howard Lutnick, where he stated that the US should make Chinese developers "addicted" to American technology, and that the nation isn't getting the "best stuff". China's senior leaders have found these comments insulting, and they are now convincing local tech giants to slow down their pursuit of NVIDIA's H20 AI chips. Many Chinese firms claim to have significantly downsized their H20 orders or completely abandoned them, which shows that there is a growing resistance to adopting the American AI tech stack. Moreover, it is also reported that domestic firms favor switching to Chinese platforms such as those from Huawei and Cambricon, given that the solutions from these companies are said to be decent enough for inferencing workloads. China hopes to address its AI computing needs through domestic offerings by next year once production lines scale up, but for now, the region has no other option than NVIDIA. We recently reported on how DeepSeek's next AI model, the R2, was delayed due to the use of domestic AI chips, and the firm later switched to NVIDIA's tech stack, which shows that China's AI advancements are in dire need of American chips for now.
[34]
NVIDIA's Emerging AI Chip Rival in China, Cambricon, Plans to Raise $560 Million to Boost Competition as Beijing Moves to Mandate Homegrown AI Chips For Datacenters
China's AI industry has seen massive developments recently, as the government starts to make efforts to ensure reliance on domestic AI chips. The recent trade situation, particularly between the US and China, has prompted both nations to tighten their grip on their respective AI technologies due to their importance. We recently saw how Beijing is moving against NVIDIA's H20 AI chip by reportedly advising local tech giants not to buy foreign chips as they could contain security flaws. And now, China is pushing domestic AI chip efforts, as the nation is now imposing requirements on data center buildouts to use homegrown solutions, heavily benefiting the likes of Huawei and Cambricon. It is claimed that the government is demanding that data centers be equipped with more than fifty percent of AI chips coming from domestic AI companies. The nation wants to reduce its reliance on NVIDIA, and considering that the US government has plans to impose security backdoors into chips flowing into China, the switch towards chips like Huawei's Ascend might be much more widespread. It is revealed that domestic AI chips cannot deliver the performance required to train top-tier AI models, which is why it is rumored that DeepSeek's next R2 model is delayed. With the growing demand for Chinese AI chips, firms like Cambricon are capitalizing on the hype by raising capital to fund their optimistic projects. Cambricon is expected to raise around 4 billion yuan, as the firm pushes its AI chips to replace the likes of AMD and NVIDIA. The Chinese AI firm offers options like the Siyuan Series for data centers and cloud computing, and is currently developing advanced options for LLMs to allow domestic AI firms to train next-gen AI models. But for now, the company has yet to make a solid breakthrough. There are a few options available to Chinese AI customers, mainly coming from Huawei. The firm offers its Ascend AI chip lineup, which includes models like the Ascend 910B and 910C, with the latter one rumored to beat NVIDIA's H100 chip in training performance. Similarly, Huawei also has a rack-scale solution called the CloudMatrix 384, which is said to rival NVIDIA's Blackwell NVL72 system. However, switching to domestic AI solutions isn't easy for Chinese firms, especially when no software matches the likes of NVIDIA's CUDA for now. China is looking for an alternative to NVIDIA's AI chips for now, but at least in the near term, the nation has to rely on American technology, given that domestic counterparts are
[35]
Howard Lutnick's 'Insulting' Words Trigger Backlash In Beijing As Regulators Pressure Alibaba, ByteDance To Shun Nvidia Chips: Report - NVIDIA (NASDAQ:NVDA)
Beijing has reportedly moved to curb sales of Nvidia Corporation's NVDA China-specific AI chip after U.S. Commerce Secretary Howard Lutnick made remarks about export policy that Chinese officials reportedly found "insulting." Chinese Regulators Push Alibaba, ByteDance to Slash H20 Chip Orders Multiple Chinese regulators -- including the Cyberspace Administration of China, the National Development and Reform Commission and the Ministry of Industry and Information Technology -- mobilized after Lutnick's comments in July, reported the Financial Times, citing people familiar with the matter. Last month, Lutnick told CNBC, "We don't sell them our best stuff, not our second-best stuff, not even our third-best," adding, "You want to sell the Chinese enough that their developers get addicted to the American technology stack, that's the thinking." Sources told FT that senior Chinese leaders considered the statement "insulting," prompting regulators to pressure domestic tech companies such as Alibaba Group BABA and ByteDance to cut back or cancel orders of Nvidia's H20 processor, a downgraded chip approved for China. See Also: Dan Ives Says Apple's AI Strategy Has Been A 'Disaster' And No One On The Street Believes Innovation Will Come From Within Beijing's Push For Domestic Alternatives The H20 chip was created after U.S. restrictions barred Nvidia from selling its most advanced GPUs in China. While Chinese tech giants argued their AI projects could not progress without Nvidia's processors, regulators have been urging firms to adopt local alternatives from Huawei Technologies and Cambricon. Some companies have begun using domestic chips for inference tasks -- the process of running AI models -- but still rely on Nvidia hardware for training due to performance gaps. Previously, it was reported that Chinese startups like DeepSeek still rely on Nvidia. DeepSeek even had to delay its new AI model after Huawei's Ascend processors failed in training. One insider told the publication, "Lutnick's speech gives the coalition [of regulators] one more reason to intensify its efforts to push tech firms to use China's own chips." Nvidia's Position And Chinese Reaction Following Lutnick's remarks, the CAC reportedly issued informal "window guidance" urging tech groups to halt new Nvidia orders. The agency also summoned Nvidia executives over alleged "serious security issues," including claims the chips could be remotely shut down -- an accusation Nvidia strongly disputed. Despite the regulatory pressure, other Chinese ministries, such as commerce and foreign affairs, had welcomed Nvidia CEO Jensen Huang during his July visit. Meanwhile, Nvidia responded to these concerns saying, "As both governments recognise, the H20 is not a military product or for government infrastructure ... allowing U.S. chips for beneficial commercial business use is good for everyone." What's Next For Nvidia In China China accounted for about 13% of Nvidia's revenue in fiscal 2024, making it a critical market. The company is now preparing to ship samples of a new "China Ready" Blackwell chip, the B30A, by September, which could offer better performance than the H20 while still complying with U.S. rules. Price Action: Nvidia shares dipped 0.14% in Wednesday's regular session but recovered slightly, rising 0.23% in after-hours trading to $175.81 at the time of writing, according to Benzinga Pro. Benzinga's Edge Stock Rankings show NVDA continues to exhibit strong upward momentum across short, medium and long-term trends. Further performance insights are available here. Read Next: AMD CEO Lisa Su Says China Strategy Rebounding As MI308 AI Chips Await US License: 'Better Position Than We Were Ninety Days Ago' Disclaimer: This content was partially produced with the help of AI tools and was reviewed and published by Benzinga editors. Photo courtesy: Saulo Ferreira Angelo on Shutterstock.com NVDANVIDIA Corp$175.810.10%Stock Score Locked: Edge Members Only Benzinga Rankings give you vital metrics on any stock - anytime. Unlock RankingsEdge RankingsMomentum86.53Growth99.28QualityN/AValue6.57Price TrendShortMediumLongOverviewBABAAlibaba Group Holding Ltd$119.80-0.16%Market News and Data brought to you by Benzinga APIs
[36]
Nvidia's China Trouble Could Spoil Q3 Outlook, Analyst Warns -- 'Might Exclude Direct Revenue From...' - NVIDIA (NASDAQ:NVDA)
Nvidia Corporation NVDA could potentially guide below Wall Street expectations for the Q3 quarter, despite the anticipation of strong earnings for the July quarter, stated KeyBanc Capital Markets. China Revenue Risks Loom Amid Export Curbs KeyBanc Capital Markets, in a recent note, predicted that Nvidia's Q3 guidance could be lower than the consensus due to uncertainties in China, reported Investing.com. The brokerage noted that Nvidia's outlook might "exclude direct revenue from China given pending license approvals and uncertainty on timing." If sales from China were included, it could potentially boost revenues by an additional $2-3 billion, stated KeyBanc. Check out the current price of NVDA stock here. Despite the U.S. easing some restrictions on AI chips, KeyBanc expects Nvidia to remain cautious. The note also pointed out potential hurdles, including a possible 15% tax on AI exports and Chinese government pressure on AI providers to adopt domestically produced AI chips. That being said, KeyBanc pointed to solid fundamental drivers. "GPU supply grew 40% in F2Q and [is] projected to increase another 20% in F3Q" with the ramp of Nvidia's Blackwell (B200). The upcoming Blackwell Ultra (B300), scheduled to start shipping in the October quarter, could account for half of total Blackwell deliveries. KeyBanc reaffirmed its Overweight rating on Nvidia and increased the price target to $215 from $190. SEE ALSO: Kanye West Launches Solana-Based 'YZY Money' -- Token Surges Over 800% Despite Rapper's February Claim That 'Coins Prey On Fans With Hype' - Grayscale Bitcoin Mini Trust (BTC) Common units of fractiona Nvidia Charts Its Way Through Geopolitics However, Daniel Newman, CEO of Futurum Group, dismissed the idea that China can move on without Nvidia Corporation, calling it "next level nonsense." Newman stated, "$NVDA looks to be advancing efforts to bring the B30, a 'China Ready' Blackwell to market." Despite the challenges in China, Nvidia has been navigating the geopolitical maze, with its GPUs at the center of the Silicon Silk Road, trading in silicon and threading its GPUs through the geopolitical maze of China's firewalls and Saudi Arabia's data palaces. Nvidia is scheduled to report its results on August 27. Benzinga's Edge Rankings place Nvidia in the 86th percentile for momentum and the 99th percentile for growth, reflecting its strong performance. Check the detailed report here. READ MORE: Nvidia To Rally More Than 22%? Here Are 10 Top Analyst Forecasts For Wednesday Image via Shutterstock Disclaimer: This content was partially produced with the help of AI tools and was reviewed and published by Benzinga editors. NVDANVIDIA Corp$176.190.45%Stock Score Locked: Edge Members Only Benzinga Rankings give you vital metrics on any stock - anytime. Unlock RankingsEdge RankingsMomentum85.52Growth99.28QualityN/AValue6.51Price TrendShortMediumLongOverviewMarket News and Data brought to you by Benzinga APIs
[37]
China Doesn't Need Nvidia? Futurum CEO Calls That 'Next-Level Nonsense' -- Dan Ives Reacts With A Fiery Response - NVIDIA (NASDAQ:NVDA)
On Tuesday, Daniel Newman, CEO of Futurum Group, dismissed the idea that China can move on without Nvidia Corporation NVDA. Trending Investment OpportunitiesAdvertisementArrivedBuy shares of homes and vacation rentals for as little as $100. Get StartedInteractive BrokersTrade global markets with low costs and pro-level tools at Interactive Brokers.Get StartedRangeRange delivers AI-powered wealth management at a fraction of the cost. Get StartedRocket HELOCGet a HELOC with mid-600s credit -- borrow and repay on repeat. Get StartedPacasoJoin 10,000+ investors betting on Pacaso's global expansion at $2.90 per share.Get StartedWorthy BondsEarn 7% fixed interest with Worthy Bonds -- start investing with just $10.Get StartedOptimus FuturesNinjaTrader gives you futures access with low day-trading margins.Get StartedIRA FinancialInvest your IRA or 401(k) in real estate, crypto, and more with IRA Financial. Get StartedAcornsGrow wealth effortlessly -- Acorns invests your spare change automatically. Get StartedSmartAssetFind a vetted financial advisor near you in minutes with SmartAsset's free tool. Get StartedDan Ives Backs Futurum CEO On Nvidia's China Push Taking to X, formerly Twitter, Newman said, "$NVDA looks to be advancing efforts to bring the B30, a 'China Ready' Blackwell to market. The whole 'China doesn't want/need NVIDIA' is next level nonsense. Don't believe it." Wedbush analyst Dan Ives quickly chimed in with a fiery emoji reply, signaling agreement that Nvidia's role in China remains critical. See Also: Dan Ives Says Apple's AI Strategy Has Been A 'Disaster' And No One On The Street Believes Innovation Will Come From Within What Is Nvidia's China Strategy Newman's statement came after it was reported that Nvidia is developing the B30A, a new AI chip tailored for China. Built on the company's advanced Blackwell architecture, the chip will feature a single-die design with about half the computing power of the flagship B300. It could also offer more performance than the H20 chip, which U.S. regulators currently allow for sale in China. Moreover, it also includes high-bandwidth memory and NVLink connectivity, core features from previous models. The company aims to ship test samples to Chinese customers by September 2025. China accounted for 13% of Nvidia's fiscal 2024 revenue, making it one of the company's most important markets. US Restrictions And Trump's Export Deal Nvidia's expansion comes against a tense political backdrop. In April, the company faced a ban but regained approval in July to sell the H20 model. President Donald Trump has since indicated his administration will permit scaled-down AI chips in China, provided Nvidia and AMD agree to give back 15% of their Chinese chip revenues to the U.S. government. Both companies reportedly accepted the deal, sparking criticism from national security experts who warn the chips could still aid China's military. Chinese Rivals And DeepSeek's Setback Nvidia also faces rising competition from companies like Huawei Technologies and Cambricon, as Beijing pushes companies to adopt domestic chips. However, China's local hardware continues to lag. Chinese AI startup DeepSeek reportedly delayed the launch of its R2 model after persistent technical failures with Huawei's Ascend processors. The firm was forced to rely on Nvidia GPUs for training while using Huawei chips only for inference tasks. Price Action: On Tuesday, Nvidia shares slipped 3.50% during regular trading and edged down another 0.34% in after-hours, settling at $175.04 at the time of writing, according to Benzinga Pro. Benzinga's Edge Stock Rankings indicate that NVDA maintains a strong upward trajectory across short, medium and long-term horizons. Additional performance details are available here. Read Next: AMD CEO Lisa Su Says China Strategy Rebounding As MI308 AI Chips Await US License: 'Better Position Than We Were Ninety Days Ago' Photo Courtesy: Saulo Ferreira Angelo on Shutterstock.com Disclaimer: This content was partially produced with the help of AI tools and was reviewed and published by Benzinga editors. NVDANVIDIA Corp$175.04-3.83%Stock Score Locked: Edge Members Only Benzinga Rankings give you vital metrics on any stock - anytime. Unlock RankingsEdge RankingsMomentum86.47Growth99.23QualityN/AValue6.53Price TrendShortMediumLongOverviewMarket News and Data brought to you by Benzinga APIs
[38]
NVIDIA's GPU Software Makes Chinese Efforts To Switch To Domestic AI Chip Use Full Of Headaches, Says Report
This is not investment advice. The author has no position in any of the stocks mentioned. Wccftech.com has a disclosure and ethics policy. Chinese AI data centers are finding it difficult to make the switch from NVIDIA's AI GPUs to Huawei's products due to software constraints, reports the South China Morning Post. The report outlines that the Chinese government has mandated all publicly funded AI data centers to use at least 50% domestic chips in order to reduce reliance on foreign chips. The chip mandate stemmed from the Shanghai municipality's guidelines last year, which required the city's computing centers to use 50% domestic chips, and these quotas were made mandatory nationwide across China this year, according to the SCMP's sources. After the Trump administration allowed NVIDIA to sell its H20 GPUs to China, the chips were caught in a controversy which hinted that they contained backdoors, tracking software or other vulnerabilities. While NVIDIA denied these reports, additional reports claimed that the Chinese government was wary of the hardware. However, soon, sources speculated that China was also wary of relying too much on foreign chips for its AI computing needs - a fact that NVIDIA CEO Jensen Huang had relied on when arguing for the export control restriction removal on his firm's products. The latest report from the SCMP suggests that China has now made it mandatory for state-run or owned computing infrastructures to rely on domestic chips as well. The details suggest that these centers will now have to use at least 50% of domestically procured chips, a requirement that stems from the Shanghai municipality's rules, which were introduced in 2024. The latest Chinese chips, which can act as a substitute for NVIDIA's hardware, are those designed by Huawei and manufactured by SMIC. SMIC is the only firm Huawei can turn to for making its chips since US sanctions restrict it from relying on TSMC. As a result, the latest Chinese indigenous chips are restricted to the 7-nanometer process, as shifting to advanced technologies requires EUV equipment, which has also been sanctioned for sale to SMIC by the US. NVIDIA's chips are also key for training new AI models, according to SCMP's sources. As a result, while Huawei's chips can be used to run new AI models, the government's requirements have created headaches for cluster operators whose AI applications have been developed with NVIDIA's chips. The troubles stem from the complementary software models required to run the chips. NVIDIA's GPUs run on the CUDA platform while Huawei's chips rely on the CANN platform instead. As a result, data centers that are now forced to use at least 50% of Huawei's chips are struggling to make their AI models that were trained with NVIDIA's chips compatible with Huawei's chips.
[39]
Nvidia Suspends China-Focused H20 Chip As Beijing Pushes For Domestic Alternatives, Analysts Warn Of Weaker Q3 Outlook: Report - NVIDIA (NASDAQ:NVDA), Amkor Tech (NASDAQ:AMKR)
Nvidia Corporation NVDA has reportedly ordered suppliers to halt work on its H20 artificial intelligence chip for China, as regulators in Beijing push local firms toward domestic alternatives and Wall Street analysts caution that ongoing uncertainty could weigh on the company's next quarter outlook. Nvidia Pulls Back On H20 Production Nvidia told suppliers, including Arizona-based Amkor Technology AMKR and South Korea's Samsung Electronics Co. SSNLF, to suspend work on the H20 chip, reported Reuters (via The Information). Amkor provides advanced packaging for the chip, while Samsung supplies high-bandwidth memory. Nvidia declined to address the specific report but told Benzinga in an emailed statement, "We constantly manage our supply chain to address market conditions." The company also reiterated that the H20 is not a military product and added, "NVIDIA does not have 'backdoors' in our chips that would give anyone a remote way to access or control them. The market can use the H20 with confidence." See Also: China Doesn't Need Nvidia? Futurum CEO Calls That 'Next-Level Nonsense' -- Dan Ives Reacts With A Fiery Response Beijing Pressures Companies To Shift To Local Chips The suspension comes after Chinese regulators, including the Cyberspace Administration of China, reportedly summoned internet firms such as Tencent Holdings TCEHY and ByteDance over their H20 chip purchases. Officials expressed concerns about information risks and have been urging companies to adopt domestic processors from Huawei Technologies and Cambricon. Some firms have started using Huawei's Ascend chips for inference tasks, though many still depend on Nvidia for training more complex AI models. Reports suggest startups like DeepSeek have struggled with delays when relying solely on local alternatives. US Remarks Spark Diplomatic Fallout The regulatory push intensified after U.S. Commerce Secretary Howard Lutnick told CNBC, "We don't sell them our best stuff, not our second-best stuff, not even our third-best ... the thinking is to sell the Chinese enough that their developers get addicted to the American technology stack." Chinese officials reportedly considered the remarks "insulting" and moved quickly to curb Nvidia's sales. Analysts See Risk To Q3 Outlook KeyBanc Capital Markets warned that Nvidia may guide below Wall Street expectations for the third quarter, citing uncertainties in China. The brokerage estimated that excluding China-related revenue could reduce sales by $2 billion to $3 billion. Still, analysts pointed to strong demand elsewhere, noting GPU supply grew 40% in the second quarter and is projected to rise another 20% in the third with the rollout of Nvidia's Blackwell B200 chips. China represented about 13% of Nvidia's fiscal 2024 revenue, making it a critical but increasingly volatile market. Price Movement: Nvidia shares fell 0.24% in Thursday's regular session and slipped an additional 0.10% in after-hours trading, finishing at $174.80, according to Benzinga Pro. Benzinga's Edge Stock Rankings indicate that NVDA maintains a robust upward price trend across short, medium and long-term timeframes. Further performance details are available here. Read Next: AMD CEO Lisa Su Says China Strategy Rebounding As MI308 AI Chips Await US License: 'Better Position Than We Were Ninety Days Ago' Photo: Hepha1st0s On Shutterstock.com Disclaimer: This content was partially produced with the help of AI tools and was reviewed and published by Benzinga editors. AMKRAmkor Technology Inc$23.391.70%Stock Score Locked: Want to See it? Benzinga Rankings give you vital metrics on any stock - anytime. Reveal Full ScoreEdge RankingsMomentum24.17Growth62.09Quality22.07Value83.51Price TrendShortMediumLongOverviewNVDANVIDIA Corp$174.80-0.34%SSNLFSamsung Electronics Co Ltd$42.480.34%TCEHYTencent Holdings Ltd$76.460.98%Market News and Data brought to you by Benzinga APIs
[40]
Nvidia's Silicon Silk Road: From China's Firewalls To Saudi's Data Palaces - NVIDIA (NASDAQ:NVDA)
If the original Silk Road moved silk and spices, the 2025 version is trading in silicon. At the center sits NVIDIA Corp NVDA, the $4 trillion chip colossus, threading its GPUs through the geopolitical maze of China's firewalls and Saudi Arabia's gilded data palaces. The stakes? Not just who builds the fastest AI, but who controls the digital infrastructure of the next decade. Track NVDA stock here. China's Workarounds and Walling Off For Nvidia, China has been both a goldmine and a minefield. The U.S. slapped export controls on advanced AI chips in 2022, forcing Nvidia into compromises. The RTX 4090D and now the newly unveiled B30A chip -- a watered-down version of its Blackwell B300 -- are designed specifically to pass Washington's scrutiny while keeping Chinese buyers hooked. Despite such efforts, the black market roared ahead: more than $1 billion worth of Nvidia chips were smuggled into China, underscoring both demand and enforcement gaps. China still represented 13% of Nvidia's revenue last year, but every new U.S. policy tweak threatens to shrink that figure. Read Also: Nvidia Vs. AMD: Who Gets Hit Harder By The 15% China Revenue Tax? Saudi Arabia's AI Palaces Meanwhile, Riyadh is rolling out a red carpet. In May 2025, during President Donald Trump's visit, the U.S. greenlit exports of up to 500,000 high-performance Nvidia and Advanced Micro Devices Inc AMD chips annually to Saudi Arabia. The Kingdom's sovereign wealth fund-backed startup Humain is already building AI "factories," beginning with 18,000 Blackwell GPUs powering a supercomputer in phase one of a $5 billion initiative. The message: Saudi Arabia isn't dabbling -- it's vaulting straight into AI nationhood, with Nvidia as chief architect. Add in the UAE's push for one of the world's largest AI data centers, and the Middle East looks less like a follower and more like an alternative AI axis. Investor Takeaway Nvidia is effectively running two AI experiments: in China, it's survival-by-compromise; in Saudi Arabia, it's growth-by-indulgence. Wall Street is watching closely -- analysts estimate Middle Eastern AI projects alone could add $1 trillion to the global AI market, a tailwind that sent Nvidia shares climbing after its May deals. From contraband GPUs in Shenzhen to supercomputers in Riyadh, Nvidia's silicon silk road is redefining not just tech markets, but geopolitics. Read Next: Nvidia Remains Hedge Fund King, But Lam Research Climbs The Ranks Photo: Hepha1st0s via Shutterstock NVDANVIDIA Corp$172.91-1.55%Stock Score Locked: Edge Members Only Benzinga Rankings give you vital metrics on any stock - anytime. Unlock RankingsEdge RankingsMomentum86.53Growth99.28QualityN/AValue6.57Price TrendShortMediumLongOverviewAMDAdvanced Micro Devices Inc$163.13-2.05% This content was partially produced with the help of AI tools and was reviewed and published by Benzinga editors. Market News and Data brought to you by Benzinga APIs
[41]
NVIDIA's H20 China AI GPU Is Now Completely Banned In The Country, Says Report
This is not investment advice. The author has no position in any of the stocks mentioned. Wccftech.com has a disclosure and ethics policy. AI GPU giant NVIDIA's H20 chips have been banned in China, suggests a report which quotes investment bank Jefferies. NVIDIA's H20 GPUs have been at the center of US trade tensions with China, after the Trump administration granted the firm licenses to sell them, multiple reports surfaced about Chinese discomfort with using the GPUs due to their US origin. However, one snippet of an investment note purportedly from Jefferies shared on social media claims that the H20 GPUs have been completely banned in China, which prevents Chinese firms from placing orders with NVIDIA. In a note purportedly belonging to Jefferies and shared on X, the investment bank outlines that Chinese authorities have completely banned NVIDIA H20 AI GPU sales in the country. The note follows several recent reports that have claimed that NVIDIA has asked some supply chain partners to halt H20 production. Jefferies also quotes these reports, as it outlines that media reports indicate that the H20 can no longer be sold in China. However, the bank also adds that "our industry checks indicate China has officially banned the procurement of H20 and other 'downgraded US AI chips' until further notice. As a result, Chinese entities are now forbidden from buying the NVIDIA GPUs, which appears to be a step up over earlier reports that have indicated that the Chinese government had simply required state-linked computing clusters to use at least 50% of domestic AI chips. As a result, Jefferies notes that "no Chinese companies will place orders with NVDA." The bank also adds that its channel checks in Taiwan have yielded the same results as media reports. "Our industry checks in Taiwan also indicate NVDA has suspended all packaging and server production work related to H20, also 'until further notice,' says the bank. As a result, Jefferies concludes that NVIDIA will be unable to generate any AI chip revenue from China, "likely until the US and China come to an agreement on trade." If the note bears fruit, then NVIDIA could see more headaches in the Asian country just as it is about to report its earnings. The shares are flat in premarket trading and enter the week after having lost 1.5% over the past five trading days. Since NVIDIA is in quiet period ahead of its earnings, it cannot comment on news, which creates an ample opportunity for short sellers to target the stock on any expected weakness.
[42]
What's Going On With Nvidia Stock Thursday? - NVIDIA (NASDAQ:NVDA)
Nvidia's NVDA AI momentum is gaining fresh traction as surging global demand, a rebound in China sales, and accelerating data center investments position the chipmaker for another strong earnings season. Nvidia isn't just growing, it's a profit machine. Investors paying 40x forward earnings know exactly what they're getting: a dominant player in AI with unmatched pricing power, scale, and ecosystem control. Ahead of the company's August 27 earnings report, analysts at UBS and Wedbush expressed confidence in its AI-driven growth. They cited a powerful combination of surging AI infrastructure spending and renewed traction in key markets. Also Read: Nvidia Prepares New China-Specific AI Chip To Defend Market Share UBS analyst Timothy Arcuri reaffirmed a Buy rating with a $205 price target, pointing to robust tailwinds supporting Nvidia's long-term growth. Wedbush's Matt Bryson went further, lifting his price target from $175 to $210 and reiterating an Outperform rating. Bryson also raised his revenue and earnings forecasts, underscoring stronger-than-expected demand trends, a recovery in China beginning in the fiscal third quarter, and expanding hyperscale and neocloud spending through 2026. Bryson emphasized that hyperscale capital expenditures rose 67% year over year in the second quarter, evidence that AI buildouts are accelerating and disproportionately benefiting Nvidia, which commands a dominant share of AI server value. He noted that neocloud operators and AI model builders are set to sustain data center investments well into 2025, with further tailwinds expected in late 2025 and 2026 as long-term infrastructure projects come online. On the product side, Bryson pointed to robust demand for Nvidia's B200 and GB200 GPUs, with supply occasionally unable to keep pace. He highlighted growing expectations for NVL72 rack shipments in the second half of 2025, supported by bullish commentary from manufacturing partner Foxconn. In China, server builds have rebounded sharply following U.S. licensing approvals for Nvidia's H20 chips, though policymakers in Beijing are simultaneously pushing domestic alternatives. Looking ahead, Bryson said GB300 server shipments remain on track for late third or early fourth quarter, while the transition to Blackwell Ultra chips is expected to proceed without major disruptions. He also reaffirmed that early 2026 production of the Rubin architecture appears on schedule despite speculation of delays, underscoring confidence in Nvidia's execution. Based on these dynamics, Bryson lifted his fiscal 2027 EPS estimate to $6.10, arguing Nvidia's valuation is supported by its role at the center of global AI infrastructure expansion, with major projects underway in Saudi Arabia, the UAE, Europe, and Asia. Price Action: NVDA stock is trading lower by 0.03% to $175.37 at last check Thursday. Read Next: Alibaba Unites Its Empire To Fight Back Against Meituan Photo by Below the Sky via Shutterstock NVDANVIDIA Corp$175.20-0.11%Stock Score Locked: Edge Members Only Benzinga Rankings give you vital metrics on any stock - anytime. Unlock RankingsEdge RankingsMomentum85.52Growth99.28QualityN/AValue6.51Price TrendShortMediumLongOverviewAMDAdvanced Micro Devices Inc$164.24-0.58%Market News and Data brought to you by Benzinga APIs
[43]
What Is Going On With Nvidia Stock Friday? - NVIDIA (NASDAQ:NVDA)
Nvidia Corp. NVDA shares slipped on Friday after reports surfaced that the company has halted work on its H20 artificial intelligence chip for China, just as domestic rival DeepSeek unveiled a new model optimized for local processors. The developments underscore mounting headwinds for Nvidia in a critical market where Beijing is accelerating efforts to reduce reliance on U.S. technology. DeepSeek, one of China's most closely watched AI startups, launched its upgraded V3.1 model on Thursday. The system is designed to run on domestic chips, reflecting Beijing's push for technological self-sufficiency. Also Read: Nvidia's AI Expansion Could Push Revenue To $300 Billion By 2026, Analyst Says The update introduces faster processing capabilities, a new FP8 precision format, and a hybrid inference structure capable of toggling between reasoning and non-reasoning tasks. The company also added a "deep thinking" feature to its platforms and plans to implement new API pricing beginning September 6. The move comes after DeepSeek encountered performance hurdles with Huawei's Ascend processors, forcing the startup to rely on Nvidia hardware for earlier models. Its R1 release in January rattled global markets, erasing roughly $600 billion in Nvidia's market value amid fears of a serious new competitor. At the same time, Nvidia is facing heightened regulatory pressure in China. The company has reportedly instructed suppliers, including Amkor Technology AMKR and Samsung Electronics SSNLF, to suspend work on its China-focused H20 chip. The pause follows meetings between Chinese regulators and tech giants such as Tencent Holdings TCEHY and ByteDance, in which officials urged companies to pivot away from Nvidia hardware in favor of domestic alternatives from Huawei and Cambricon. Analysts caution that the disruption could weigh on Nvidia's near-term results. KeyBanc Capital Markets estimates that excluding China sales could cut quarterly revenue by $2 billion to $3 billion. However, strong global demand and the company's rollout of its next-generation Blackwell B200 chips may help offset some of the pressure. China contributed 13% of Nvidia's revenue in fiscal 2024, highlighting the strategic importance of the market. Despite the latest setback, Nvidia stock has gained more than 30% this year, cementing its dominance in the graphics processing unit segment as Big Tech players including Microsoft MSFT and Meta Platforms META pour billions into AI infrastructure. Price Action: NVDA stock is trading lower by 1.14% to $172.98 premarket at last check Friday. Read Next: Oracle Veteran Security Chief Exits As Company Restructures For AI Growth Image via Shutterstock NVDANVIDIA Corp$172.99-1.14%Stock Score Locked: Want to See it? Benzinga Rankings give you vital metrics on any stock - anytime. Reveal Full ScoreEdge RankingsMomentum85.30Growth99.29QualityN/AValue6.44Price TrendShortMediumLongOverviewAMKRAmkor Technology Inc$23.360.65%METAMeta Platforms Inc$740.000.12%MSFTMicrosoft Corp$504.700.09%SSNLFSamsung Electronics Co Ltd$42.480.34%TCEHYTencent Holdings Ltd$76.46-%Market News and Data brought to you by Benzinga APIs
[44]
Nvidia working on new AI chip for China that outperforms the H20, sources say
BEIJING/SINGAPORE -- Nvidia is developing a new artificial intelligence chip for China based on its latest Blackwell architecture that will be more powerful than the H20 model it is currently allowed to sell there, two people briefed on the matter said. U.S. President Donald Trump last week opened the door to the possibility of more advanced Nvidia chips being sold in China. But the sources noted U.S. regulatory approval is far from guaranteed amid deep-seated fears in Washington about giving China too much access to U.S. AI technology. The new chip, tentatively known as the B30A, will use a single-die design likely to deliver half the raw computing power of the more sophisticated dual-die configuration in Nvidia's flagship B300 accelerator card, the sources said. A single-die design has all the main parts of an integrated circuit on one continuous piece of silicon rather than split across multiple dies. The new chip would have high-bandwidth memory and Nvidia's NVLink technology for fast data transmission between processors, features that are also in the H20, a chip based on the company's older Hopper architecture. The new chip's specifications are not completely finalized but Nvidia hopes to deliver samples to Chinese clients for testing as early as next month, said the sources who were not authorized to speak to media and declined to be identified. "We evaluate a variety of products for our roadmap, so that we can be prepared to compete to the extent that governments allow," Nvidia said in a statement. "Everything we offer is with the full approval of the applicable authorities and designed solely for beneficial commercial use." "Of course (CEO Jensen Huang) would like to sell a new chip to China," U.S. Commerce Secretary Howard Lutnick said when asked about the Reuters story in a CNBC interview, repeatedly praising Huang. "I'm sure he's pitching the president all the time." "I've listened to him pitch the president, and the president listens to our great technology companies, and he'll decide how he wants to play it. But the fact Jensen is pitching a new chip shouldn't surprise anybody." The U.S. Department of Commerce did not respond to a request for comment. The extent to which China, which generated 13 per cent of Nvidia's revenue in the past financial year, can have access to cutting-edge AI chips is one of the biggest flashpoints in U.S.-Sino trade tensions. Nvidia only received permission in July to recommence sales of the H20. It was developed specifically for China after export restrictions were put in place in 2023, but the company was abruptly ordered to stop sales in April. Trump said last week he might allow Nvidia to sell a scaled-down version of its next-generation chip in China after announcing an unprecedented deal that will see Nvidia and rival AMD AMD.O give the U.S. government 15 per cent of revenue from sales of some advanced chips in China. A new Nvidia chip for China might have "30 per cent to 50 per cent off," he suggested in an apparent reference to the chip's computing power, adding that the H20 was "obsolete." U.S. legislators, both Democratic and Republican, have worried that access to even scaled-down versions of flagship AI chips will impede U.S. efforts to maintain its lead in artificial intelligence. But Nvidia and others argue that it is important to retain Chinese interest in its chips - which work with Nvidia's software tools - so that developers do not completely switch over to offerings from rivals like Huawei. Huawei has made great strides in chip development, with its latest models said to be on par with Nvidia in some aspects like computing power, though analysts say it lags in key areas such as software ecosystem support and memory bandwidth capabilities. Complicating Nvidia's efforts to retain market share in China, Chinese state media have also in recent weeks alleged that the U.S. company's chips could pose security risks, and authorities have cautioned Chinese tech firms about purchasing the H20. Nvidia says its chips carry no backdoor risks. Nvidia is also preparing to start delivering a separate new China-specific chip based on its Blackwell architecture and designed primarily for AI inference tasks, according to two other people familiar with those plans. Reuters reported in May that this chip, dubbed the RTX6000D, will sell for less than the H20, reflecting weaker specifications and simpler manufacturing requirements. The chip is designed to fall under U.S. government thresholds. It uses conventional GDDR memory and features memory bandwidth of 1,398 gigabytes per second, just below the 1.4 terabyte threshold established under restrictions introduced in April that led to the initial H20 ban. Nvidia is set to deliver small batches of RTX6000D to Chinese clients in September, said one of the people.
[45]
Nvidia developing more powerful AI chip to sell in China: report
Nvidia is developing an AI chip for China that is more powerful than the H20 model it is currently allowed to sell in the region, according to a report. The new chip, tentatively named the B30A, is a less advanced version of its Blackwell B300 design, which is blocked from sale in China by export restrictions, two sources familiar with the company's plans told Reuters. It's likely to deliver about half of the raw computing power of the B300 - after President Trump signaled he'd be open to allowing a new chip into China that's "30 to 50% off" in terms of power, adding that the H20 has grown "obsolete." Commerce Secretary Howard Lutnick on Tuesday said he wouldn't be surprised if CEO Jensen Huang wants to sell a new chip to China. "I'm sure he's pitching the president all the time," Lutnick told CNBC's "Squawk on the Street." "I've listened to him pitch the president, and the president listens to our great technology companies, and he'll decide how he wants to play it." Nvidia is aiming to send samples of the new chip to Chinese clients for testing as early as next month, the sources told Reuters. "We evaluate a variety of products for our roadmap, so that we can be prepared to compete to the extent that governments allow," an Nvidia spokesperson told The Post in a statement. "Everything we offer is with the full approval of the applicable authorities and designed solely for beneficial commercial use." Trump recently announced an unprecedented deal for Nvidia and AMD to give the US government 15% of revenue from their sales in China - in exchange for permission to recommence sales of the H20 and AMD's MI308. Nvidia built the H20 model specifically to bypass Biden-era export controls, but the company was ordered to halt sales in April. Trump said last week that he asked for a 20% cut of Nvidia's China sales, but that Huang negotiated it down to 15%. It's unknown whether Nvidia would receive regulatory approval for the sale of these chips, as there are still strong national security concerns around handing China a lead in the AI race. And there's growing distrust of Nvidia in China as state media claim their chips pose security risks, though Nvidia has maintained this is false. But Nvidia doesn't want to lose its share of the Chinese market to domestic rivals like Huawei. About 13% of Nvidia's revenue in the past financial year came from China.
[46]
Howard Lutnick's 'insulting' remarks about Nvidia's H20 chips irk...
Chinese officials are miffed over Commerce Secretary Howard Lutnick's "insulting" remarks about Nvidia's H20 computer chip -- and are now pressuring domestic firms not to buy it, according to a report Thursday. During a July 15 appearance on CNBC, Lutnick downplayed the Trump administration's decision to lift an additional set of export controls that would have barred sales of the H20 - a less-powerful version of Nvidia's chips that was made specifically to comply with US export controls on the sensitive technology. "We don't sell them our best stuff, not our second-best stuff, not even our third-best," Lutnick said at the time. "You want to sell the Chinese enough that their developers get addicted to the American technology stack, that's the thinking," he added. Shortly after Lutnick's CNBC appearance, Beijing began its regulatory crackdown, the Financial Times reported. Top Chinese officials viewed Lutnick's comments about the H20 chip as "insulting," two sources with knowledge of the crackdown told the FT in a story published Thursday. Shares of Nvidia, led by CEO Jensen Huang, were down less than 1% in early trading on Thursday. Some Chinese tech firms have scaled back their H20 orders in response to the pressure from Beijing, the report added. "Lutnick's speech gives the coalition [of regulators] one more reason to intensify its efforts to push tech firms to use China's own chips," one of the FT's sources said. Agencies involved in the pressure campaign reportedly include the Cyberspace Administration of China, the National Development and Reform Commission and the Ministry of Industry and Information Technology. "As both governments recognize, the H20 is not a military product or for government infrastructure," an Nvidia spokesperson said in a statement. "China won't rely on American chips for government operations, just like the U.S. government would not rely on chips from China. However, allowing U.S. chips for beneficial commercial business use is good for everyone." The Commerce Department did not immediately return a request for comment. Nvidia's chips have become a hot commodity because they are considered the most advanced in the market -- and are crucial for powering artificial intelligence models. Some critics fear that allowing sales of even the H20 chip to China will hurt the US's competitiveness in the AI race. The situation has unfolded during a period of ongoing trade discussions between the US and China, which have yielded some progress but have yet to result in a full-fledged agreement. In order to secure the necessary export licenses, Nvidia and fellow chipmaker AMD have agreed to pay the US government 15% of the revenue from their chip sales in China.
[47]
NVIDIA Stock Dips 1.15% as China Pushback Halts H20 AI Chip Production
NVIDIA (NASDAQ: NVDA) stock is under heat on August 22, 2025. The company has instructed its suppliers, such as Samsung Electronics and Amkor Technology, to cease development on its. This news has concerned investors over a possible sharp decline in NVIDIA share price. The stoppage comes after Beijing instructed its domestic companies to steer clear of the H20 due to security reasons. The chip was specifically built as a weaker variation of NVIDIA's top-of-the-line AI accelerators. However, Chinese customers are increasingly turning away from the company and opting for local options from Huawei and Cambricon.
[48]
Nvidia halts H20 chip production as China clamps down on orders,...
Nvidia has slammed the brakes on production of its controversial H20 AI chip after Beijing urged Chinese firms to dump the US hardware on alleged security risks -- a move that rattled investors and sent shockwaves through the global chip industry. The chip giant ordered suppliers Samsung Electronics and Amkor Technology to halt manufacturing this week following China's crackdown on the scaled-down processor designed for its market, according to The Information. Nvidia shares slipped 1.1% in early trading Friday as Wall Street digested the latest blow to its China business, which pulled in $17 billion last year. The freeze raises fresh doubts about demand for the H20, a watered-down version of Nvidia's flagship accelerators created to skirt US export bans while still tapping China's lucrative market. Rivals Huawei Technologies and Cambricon Technologies are now poised to seize ground. Cambricon's stock soared 20% Friday, fueling a rally among domestic chipmakers. The timing couldn't be worse for Nvidia, which already wrote off $5.5 billion in H20 inventory after the Trump administration initially banned the product. In recent weeks, Chinese regulators have warned firms against using American chips, citing alleged security risks. Nvidia CEO Jensen Huang, caught off guard by the move, insisted the H20 contains no backdoors. "We're in dialogue with them but it's too soon to know," he told reporters during an impromptu airport briefing in Taiwan, where he was meeting with TSMC about his upcoming Rubin chip. Both Nvidia and rival AMD recently won approval from Washington to resume limited AI chip sales to China under controversial terms requiring them to hand over 15% of related revenue to the US government. But Beijing is accelerating a push to wean itself off American tech. That push gained momentum Thursday when Chinese AI phenom DeepSeek said its latest model was built to run on next-generation homegrown chips, though it gave no details. Meanwhile, semi-finished Nvidia chips are "piling up" at Amkor, The Information reported. The company acknowledged it still holds large H20 stockpiles but said market conditions remain "highly uncertain." "We constantly manage our supply chain to address market conditions," an Nvidia spokesperson told Bloomberg, adding the H20 was designed strictly for commercial use. "As both governments recognize, the H20 is not a military product or for government infrastructure." Nvidia hopes to roll out a successor to the H20, but Huang cautioned any launch hinges on approval from the incoming Trump administration. "Offering a new product to China for AI data centers -- the follow-on to the H20 -- that's not our decision to make," he said. Bloomberg Intelligence analysts said the halt creates "fresh uncertainty" over when Nvidia's China business can recover, though they expect strong US demand to soften the blow. Nvidia reports earnings next week, giving investors their first detailed look at how escalating trade tensions are hitting the world's most valuable chipmaker. "We constantly manage our supply chain to address market conditions," an Nvidia spokesperson told The Post. "As both governments recognize, the H20 is not a military product or for government infrastructure. China won't rely on American chips for government operations, just like the U.S. government would not rely on chips from China. However, allowing U.S. chips for beneficial commercial business use is good for everyone." The Nvidia rep added that "cybersecurity is critically important to us. NVIDIA does not have 'backdoors' in our chips that would give anyone a remote way to access or control them. The market can use the H20 with confidence." The Post has sought comment from Samsung, Amkor, TSMC, Huawei, Cambricon, DeepSeek and the Chinese government.
[49]
Nvidia Faces a Delicate Balancing Act Ahead of Q2 Earnings | Investing.com UK
Wherever AI hype has risen, warnings of an AI bubble have followed close behind. The comparison to the dot-com era is not without merit: real value was often buried under layers of overpromising. In the case of AI, the exaggeration shows up in claims like "AGI next year" or "AI will replace all jobs." Yet beneath the noise, tangible productivity gains are already evident -- from imaging and video to software development pipelines. Governments, too, are driving the momentum because AI functions as a force multiplier. Properly deployed, AI offers unprecedented leverage over information flows and cultural narratives - the very foundations of political legitimacy. Palantir (NASDAQ:PLTR) and Oracle (NYSE:ORCL) are emerging as central players in this state-level adoption. And at the junction of AI hype and bubble concerns sits Nvidia (NASDAQ:NVDA), the dominant AI chipmaker, set to release its Q2 earnings on August 27. What can investors expect? According to Zacks Investment Research covering 14 analyst inputs, the expected Nvidia earnings per share (EPS) consensus for Q2 sits at $0.94, which is 44.6% higher than the year-ago quarter's $0.65. In the last reported quarter ending April, Nvidia tracked a growth of 69% year-over-year. During this quarter, the company suffered a $4.5 billion charge owing to new export requirements for H20 AI inference GPUs, specifically for the Chinese market. In Q1, Nvidia expected the upcoming Q2 report to bring $45 billion in revenue, plus or minus 2%. For comparison, Q2 2024 brought in $30 billion revenue, which was up 122% from the year-ago quarter. In other words, despite Nvidia being caught between US-China geopolitics, the growth rate is dwindling but still impressive. Overall, the aggregated revenue expectation for Q2 across 43 analysts sits at $45.92 billion, with a high estimate of $52.62 billion and a low one at $45 billion, which is Nvidia's baseline. This would put Nvidia's average YoY sales growth just under 53%. All of Nvidia's chips are manufactured by TSMC in Taiwan and at foundries in the U.S. Taiwan is China's neighboring island whose status remains disputed. Presently, the island is effectively a U.S. military protectorate, which gives the U.S. leverage over China's AI development. From 2023 to 2025, Nvidia's China revenue share shrunk from 21.45% to 13.11% ($17B). At the same time, China has ample leverage as the dominant exporter of rare earth elements (REEs) that are so critical in the advanced semiconductor sector for military applications. On top of this dynamic, the USG doesn't necessarily want to harm Nvidia's bottom line as the carrier of U.S. leadership in AI. For this reason, President Trump recently adopted a novel approach. Namely, both Nvidia and AMD (NASDAQ:AMD) agreed to pay fixed 15% of sales to USG in exchange for export licenses to China, which was purportedly negotiated down from 20%. This is related to Nvidia's H20 chips and AMD's equivalent MI308 chips. "While we haven't shipped H20 to China for months, we hope export control rules will let America compete in China and worldwide." Logically, it wouldn't make sense to tie purported national security to the curtailment of China's AI advancement, and then put a price tag on that concern. However, President Trump later clarified that the H20 series is "an old chip that China already has." At this point in time, H20 sales in China seem to be irrelevant. According to Qingyuan Lin, Bernstein's senior semiconductor analyst, the Chinese government apparently issued a "hard mandate" to stop domestic companies from buying H20s. This aligns with our previous reporting, wherein Huawei and SMIC are closing in on Nvidia's tech. Therefore, China is now more than ever incentivized to spur domestic AI chip ecosystem, despite the loosened export licensing from the US/Taiwan. Nonetheless, Nvidia has an opening with a new lineup. Presently, Nvidia's most advanced AI workload architecture is Blackwell. Last week, President Trump described Blackwell as "super-duper advanced." At the same time, he hinted that even this platform may be on the negotiation table for export licensing. "It's possible I'd make a deal...on somewhat enhanced - in a negative way - Blackwell. In other words, take 30% to 50% off of it." According to Deloitte research, the global TAM (total addressable market) for semiconductors should reach $2 trillion by 2040, at a compound annual growth rate (CAGR) of 19%. For Nvidia, China would present a $50 billion opportunity. This puts Nvidia in a tough but manageable spot. The company has to offer something more enticing than H20 to China, while avoiding the sales of its very best - Blackwell. This is where B30A comes in according to insider sources to Reuters. Although its specifications are not yet finalized, the new B30A chip would have half the power of Blackwell B300 on a single-die. Potentially, this would allow Nvidia to commit to China sales even under President Trump's steep revenue reduction between 30% to 50%. After all, the H20 has been negotiated down from 20%, and Trump is an openly transactional president. President Trump seems open to widening the doors for US domestic semiconductor companies into China. Moreover, the recent EU-US trade deal greatly favors these companies with expanded market access and asymmetrical tariff conditions, further entrenching Europe into the U.S. tech stack. *** Looking to start your trading day ahead of the curve?
[50]
Nvidia working on new AI chip for China: sources
STORY: Nvidia is developing a new AI chip for China. That's according to two sources, who said the U.S. tech giant will base the chips on its latest Blackwell architecture. They said it will also be more powerful than the H20 model Nvidia's currently allowed to sell there. U.S. President Donald Trump last week implied the firm might be permitted to sell more advanced chips in China. But the sources noted U.S. regulatory approval isn't guaranteed. They say Washington has strong concerns about giving China too much access to U.S. AI technology. The sources said the new chip is currently known as the B30A, and its specifications aren't completely finalized. They added Nvidia hopes to deliver samples to Chinese clients for testing as early as next month. Nvidia said in a statement: "we evaluate a variety of products for our roadmap, so that we can be prepared to compete to the extent that governments allow." The U.S. Department of Commerce did not respond to a Reuters request for comment. China generated 13% of Nvidia's revenue in the past financial year. The company only received permission in July to resume sales of the H20 after it was ordered to stop in April. The chip was developed specifically for China after export restrictions were put in place two years ago. Trump said last week he might allow Nvidia to sell a scaled-down version of its next-generation chip in China. That's after he announced an unprecedented deal that will see Nvidia and rival AMD give the U.S. government 15% of revenue from sales of some advanced chips in China.
[51]
Nvidia Working on New AI Chip for China That Tops H20, Returns Reports, Citing Sources
--Nvidia is developing a new AI chip for China based on its latest Blackwell architecture that will be more powerful than the H20 model it currently sells in the country, Reuters reports, citing people familiar with the matter. --President Trump suggested last week that he might allow more advanced Nvidia chips to be sold in China, though regulatory approval of such an arrangement is far from guaranteed, the people told Reuters. --The new chip is tentatively being called the B30A and will use a single-die design that could deliver half the raw computing power of the more sophisticated dual-die configuration used in Nvidia's flagship B300 accelerator card, according to the report.
[52]
Exclusive-Nvidia working on new AI chip for China that outperforms the H20, sources say
BEIJING/SINGAPORE (Reuters) -Nvidia is developing a new AI chip for China based on its latest Blackwell architecture that will be more powerful than the H20 model it is currently allowed to sell there, two people briefed on the matter said. U.S. President Donald Trump last week opened the door to the possibility of more advanced Nvidia chips being sold in China. But the sources noted U.S. regulatory approval is far from guaranteed amid deep-seated fears in Washington about giving China too much access to U.S. artificial intelligence technology. The new chip, tentatively known as the B30A, will use a single-die design that is likely to deliver half the raw computing power of the more sophisticated dual-die configuration in Nvidia's flagship B300 accelerator card, the sources said. A single-die design is when all the main parts of an integrated circuit are made on one continuous piece of silicon rather than split across multiple dies. The new chip would have high-bandwidth memory and Nvidia's NVLink technology for fast data transmission between processors, features that are also in the H20 - a chip based on the company's older Hopper architecture. The chip's specifications are not completely finalised but Nvidia hopes to deliver samples to Chinese clients for testing as early as next month, said the sources who were not authorised to speak to media and declined to be identified. Nvidia said in a statement: "We evaluate a variety of products for our roadmap, so that we can be prepared to compete to the extent that governments allow." "Everything we offer is with the full approval of the applicable authorities and designed solely for beneficial commercial use," it said. The U.S. Department of Commerce did not respond to a Reuters request for comment. FLASHPOINT The extent to which China, which generated 13% of Nvidia's revenue in the past financial year, can have access to cutting-edge AI chips is one of the biggest flashpoints in U.S.-Sino trade tensions. Nvidia only received permission in July to recommence sales of the H20. It was developed specifically for China after export restrictions were put in place in 2023, but company was abruptly ordered to stop sales in April. Trump said last week he might allow Nvidia to sell a scaled-down version of its next-generation chip in China after announcing an unprecedented deal that will see Nvidia and rival AMD give the U.S. government 15% of revenue from sales of some advanced chips in China. A new Nvidia chip for China might have "30% to 50% off", he suggested in an apparent reference to the chip's computing power, adding that the H20 was "obsolete". U.S. legislators, both Democratic and Republican, have worried that access to even scaled-down versions of flagship AI chips will impede U.S. efforts to maintain its lead in artificial intelligence. But Nvidia and others argue that it is important to retain Chinese interest in its chips - which work with Nvidia's software tools - so that developers do not completely switch over to offerings from rivals like Huawei. Huawei has made great strides in chip development, with its latest models said to be on par with Nvidia in some aspects like computing power, though analysts say it lags in key areas such as software ecosystem support and memory bandwidth capabilities. Complicating Nvidia's efforts to retain market share in China, Chinese state media have also in recent weeks alleged that the U.S firm's chips could pose security risks, and authorities have cautioned Chinese tech firms about purchasing the H20. Nvidia says its chips carry no backdoor risks. Nvidia is also preparing to start delivering a separate new China-specific chip based on its Blackwell architecture and designed primarily for AI inference tasks, according to two other people familiar with those plans. Reuters reported in May that this chip, currently dubbed the RTX6000D, will sell for less than the H20, reflecting weaker specifications and simpler manufacturing requirements. The chip is designed to fall under thresholds set by the U.S. government. It uses conventional GDDR memory and features memory bandwidth of 1,398 gigabytes per second, just below the 1.4 terabyte threshold established by restrictions introduced in April that led to the initial H20 ban. Nvidia is set to deliver small batches of RTX6000D to Chinese clients in September, said one of the people. (Reporting by Liam Mo in Beijing and Fanny Potkin in Singapore; Additional reporting by Alexandra Alper in Washington and Wen-Yee Lee in Taipei; Editing by Brenda Goh and Edwina Gibbs)
[53]
Nvidia asks Foxconn to suspend work for H20 chip, sources say
BEIJING (Reuters) -Nvidia has asked Foxconn to suspend work on the H20 AI chip, the most advanced product the U.S. company is currently permitted to sell to China, two people briefed on the matter said. Foxconn, formally known as Hon Hai Precision Industry Co Ltd, handles backend processing work as one of Nvidia's component suppliers for the chip. "We constantly manage our supply chain to address market conditions," Nvidia said in a statement, declining to elaborate further. (Reporting by Fanny Potkin in Singapore, Liam Mo in Beijing, Wen-Yee Lee in Taipei and Brenda Goh in Shanghai; Editing by Himani Sarkar)
[54]
Nvidia Reportedly Asks Suppliers to Halt H20 Work
Nvidia (NVDA) has instructed component suppliers to stop production related to the H20, The Information reported, following news that Beijing urged local companies to avoid using the AI chip designed specifically for the Chinese market. Nvidia Chief Jensen Huang, meanwhile, told reporters in Taiwan that the company assured there were no "security backdoors" on those chips, after China voiced concerns. Ed Yardeni, president at Yardeni Research, explains what it could mean for the AI race and tech stocks. He speaks on Bloomberg Insight. Bloomberg videos, provided by MT Newswires
Share
Share
Copy Link
Nvidia's efforts to maintain its presence in China's AI chip market face challenges from U.S. export restrictions and China's push for domestic chip production, highlighting the complex interplay of technology, politics, and market dynamics in the global AI industry.
Nvidia, the world's most valuable chipmaker, is reportedly developing a new AI chip codenamed B30A for sale in China
1
. This chip is said to be half as powerful as Nvidia's flagship B300 Blackwell GPU but more powerful than the H20 GPUs currently allowed for sale in China. The B30A features a single-die design, fast data transmission, NVLink support, and high-bandwidth memory.Source: Tom's Hardware
However, recent developments have complicated Nvidia's plans. The company has reportedly halted production of its H20 AI chips after Beijing warned Chinese companies against using them due to potential security issues
2
. This comes just a month after companies like Nvidia were given the green light to sell AI chips designed for China's market.The AI chip market has become a focal point of geopolitical tensions between the United States and China. The Trump administration recently imposed a 15% sales tax on AMD and Nvidia hardware sold to China
3
. This move follows earlier restrictions on high-performance AI chip exports to China.Adding to the complexity, U.S. Commerce Secretary Howard Lutnick's comments about making China "addicted" to U.S. technology have sparked controversy
4
. Chinese authorities found these remarks insulting, leading to directives for Chinese tech institutions to move away from purchasing Nvidia H20 chips.Source: Wccftech
In response to these challenges, China has intensified its efforts to develop domestic AI chip capabilities. The country has mandated that all domestic data centers source more than 50% of their chips from Chinese producers
5
. This policy aims to reduce reliance on foreign semiconductors and boost China's technological self-sufficiency.However, China's domestic chip industry faces significant hurdles. While companies like Huawei and Cambricon are producing competitive chips for specific applications, they still lag behind Nvidia in terms of overall performance and software ecosystem
3
. The lack of an effective alternative to Nvidia's CUDA software stack remains a major challenge for Chinese AI developers.Related Stories
The ongoing tensions and restrictions are impacting AI development in China. DeepSeek, a prominent Chinese AI company, reportedly abandoned training its next-generation R2 model on domestically developed Huawei's Ascend platforms due to performance issues and software limitations
3
. This highlights the continued dependence on Nvidia's technology stack in China's AI sector.The situation poses a dilemma for Chinese tech companies. While there's pressure to adopt domestic chips, concerns about performance, stability, and software compatibility make many hesitant to fully transition away from Nvidia's solutions
4
. This reluctance could potentially slow down AI innovation in China, especially in areas requiring high-performance computing for AI training.Source: TweakTown
The ongoing chip dispute between the U.S. and China has broader implications for the global AI industry. As both countries invest heavily in AI and semiconductor technologies, the fragmentation of the global supply chain and technology ecosystems could lead to divergent development paths in AI.
For Nvidia and other U.S. chip manufacturers, the Chinese market remains crucial despite the challenges. The company continues to navigate the complex regulatory landscape while developing new products tailored for the Chinese market
1
. However, the long-term sustainability of this approach remains uncertain as geopolitical tensions persist and China pushes for greater technological independence.Summarized by
Navi
[1]
[3]
[4]
25 Jul 2025•Technology
20 Jul 2025•Business and Economy
29 Apr 2025•Technology
1
Business and Economy
2
Technology
3
Business and Economy