Acquiring “Illegal” GPUs in China<\/p>\n
In China, it’s not an illegal market — it’s just a market.<\/p>\n
<\/p>\n
We spoke to everyone about this NVIDIA AI GPU black market: We found middlemen who connect buyers and suppliers, users who can understand the demand and explain it, using the most dystopian definition of wealth — how many GPUs one has.<\/p>\n
<\/p>\n
We also found independent repair shops who, simply doing their jobs, salvage valuable silicon components from dead PCBs of banned GPUs, innovatively hand-modifying them to be better than and have more VRAM than NVIDIA\u2019s own official product SKUs. These shops are not explicitly a part of any “black market,” they’re just repair shops that happen to sometimes work on export-controlled GPUs.<\/p>\n
We met multiple people who, when asked the same question, gave the same passcode-like Chinese idiomatic expression, or chengyu (\u6210\u8bed), which translates to \u201copen one eye, close one eye.\u201d In other words, it means, \u201cto turn a blind eye.\u201d We heard this saying so much that we made a T-shirt based on it to help fund this investigative report<\/a>.<\/p>\n “There’s a new kind of black market, and it’s high-end AI GPUs. This particular black market is worth billions of dollars a year”<\/p>\n And among others, we even spoke with a US-based Chinese national buying video cards to strip them and ship the GPUs to Chinese companies, which violates US export control law.<\/p>\n One of our viewers was able to meet with a GPU smuggler, whom we\u2019ll call \u201cThe Plug.\u201d The smuggler spoke limited English, but they both understood one universal truth: money. He operated a GPU testing rig inside of a Prius that he drives around multiple states in the Western United States. The least suspicious thing in his car was a spare license plate in the trunk, but we\u2019ll come back to \u201cThe Plug\u201d towards the end.<\/p>\n “Black market” is normally a phrase associated with drugs or guns, but there’s a new kind of black market: high-end, AI GPUs. This particular black market is worth billions of dollars a year and is hiding in plain sight.<\/p>\n Hong Kong is our first stop along our journey.\u00a0<\/p>\n By skyscraper count, Hong Kong would be the tallest city on earth. The density is unbelievable. We spent a few days here for this story, wandering markets and meeting sources.<\/p>\n We went to Hong Kong to learn what the demand drivers are for the banned GPUs, how they get into China, and more about the illicit side of the GPU smuggling business.\u00a0<\/p>\n Big Adventure\u00a0<\/p>\n To provide a big overview of our big adventure, it began when we booked a 24-hour plane ticket to Hong Kong and 20 days of hotels across Hong Kong, Shenzhen, Zhengzhou, Huizhou, Taipei, and more. At least one of us got detained by at least one of the governments involved in this story, but we can’t talk about these indecipherably singular or plural instances or instance of any aforementioned detainment.<\/p>\n Then we talked to a lawyer about the previous sentence, then investigated whether or not GPUs are actually smuggled with lobsters<\/a> into Hong Kong.<\/p>\n We eventually found a pile of GPUs that are now export controlled and met a professor who really wants to make sure you know that they were \u201clegally obtained.\u201d<\/p>\n After leaving Hong Kong, we booked a boat to Shekou, China. Then headed to Shenzhen, got kicked out of a warehouse, and used high-speed rail to go deep into China, where we met a guy who thinks that desoldering a GPU and reballing it is no big deal.\u00a0<\/p>\n We got a ton of information about GPU smuggling from a guy named \u201c5.\u201d\u00a0<\/p>\n After that, we went back to Shenzhen for the third time in two weeks, went back to Hong Kong, flew to Taiwan, before finally getting back on the plane to the US.<\/p>\n Perspective<\/p>\n We shot over 12 hours of interviews<\/p>\n We spoke to a lot of people on the record for this story and learned about the dystopian world of high-tech GPU smuggling. We spoke to people ranging from owner-operator trading companies to professors of economics building datacenters. We even tried to talk to the US Department of Commerce, but they didn\u2019t reply, and every person on the chain from the US Department of State had out of office auto responders because we coincidentally emailed them the same week of an especially problematic story that enigmatically involved namedropping.\u00a0<\/p>\n We wanted to connect as many pieces of this puzzle together as we can today, and that\u2019s why we shot over 12 hours of interviews that we just spent weeks cutting down. The point is finding people who know people, and each person in our lineup today led us to at least one other person in this video, eventually building the full pipeline of smuggler-to-user. Although this story isn\u2019t about drugs, it is about a different thing that billionaire executives get a high from: AI<\/a>.\u00a0<\/p>\n But before we get to any of those interviews, we need to establish the basics of this geopolitical mess.<\/p>\n The story is complicated, so we’ll start with defining these key facts: Why these GPUs are banned, the new 15% license<\/a> (which only applies to two GPU models), who buying and selling is legal or illegal for, and then the timeline.\u00a0<\/p>\n We have over 400 pages of research that went into finding sources and understanding the laws.<\/p>\n Why They\u2019re Banned<\/p>\n AI GPUs have been in the mainstream news constantly.<\/p>\n This story has been a complete mess to follow. It has spanned years and two US administrations. We have over 400 pages of research that went into finding sources and understanding the laws.\u00a0<\/p>\n Reports warned of NVIDIA product use in nuclear weapons research<\/p>\n Here\u2019s why governments care about AI GPUs:<\/p>\n NVIDIA functionally holds a GPU monopoly in our sector of the industry<\/a>, which is building computers to play video games. That now feels relatively innocent by comparison to AI. The company leveraged decades of gaming domination to build a foundation for what it now focuses on, which is making the most powerful GPUs for AI in the world. Reports warned of NVIDIA product use in nuclear weapons research<\/a>, facial recognition technology allegedly used in Russia to suppress dissent<\/a> and growing concerns of AI facial recognition use in the US for similar deployments<\/a>, alongside reports of use in international spying<\/a> and in drone warfare<\/a>.\u00a0<\/p>\n NVIDIA finds itself in the middle of all of this. Even though NVIDIA disputes selling to some of these entities — for example, it says that it doesn\u2019t sell GPUs to Russia<\/a> — the products still find their way there. NVIDIA is making money one way or the other. Someone is buying it<\/a>, maybe from someone else who bought it from someone else, and it may be transacted through smuggling. But NVIDIA does end up selling the device ultimately to somebody. Regardless of who they sell to, NVIDIA plays a big part in this worldwide obsession of AI.<\/p>\n And we think it\u2019s playing all sides, but we\u2019ll talk about that more at the end. Besides, when there\u2019s a gold rush, it\u2019s better to sell the pickaxe than swing it.<\/p>\n Although the US doesn\u2019t talk too much about its own use of AI, it spends a lot of time talking about China\u2019s.<\/p>\n The US restricts NVIDIA\u2019s GPUs through export control rules<\/a> that ban the sale of certain GPUs into China. The restriction is for the sale of GPUs by American companies or companies that want to transact business in America to companies that are in China or the Chinese government itself. Some examples of export-controlled GPUs include gaming GPUs like the RTX 5090<\/a> and 4090<\/a>, which are useful in AI applications (mostly for their high VRAM capacity), and data center\/AI GPUs like the A100, H100, H200, and B100, as well as the others shown in the image above. This list is constantly in flux. There are some new and incoming exceptions for the NVIDIA H20 specifically<\/a>, which has faced Schrodinger\u2019s GPU ban depending on whether CEO Jensen Huang had a one-million-dollar dinner at Mar-a-Lago<\/a> with Donald Trump on a given week.\u00a0<\/p>\n GPU Export Control Timeline<\/p>\n Now, we\u2019ll get into the timeline of GPU export controls across multiple administrations. We have a separate article with the full, bulleted timeline here<\/a>.<\/p>\n Obama Administration and Early AI Talk<\/p>\n Even at the end of the Obama Administration in 2016, the US Government was just starting to talk about AI in relation to national security in a Wired interview. Then-President Barack Obama said<\/a>:<\/p>\n \u201cDeveloping international norms, rules, protocols, verification mechanisms around cyber security generally and AI, in particular, is in its infancy. You got a lot of non-state actors who are the biggest players. Part of the problem is that identifying who’s doing what is much more difficult. If you’re building a bunch of ICBMs, we see them. If somebody’s sitting at a keyboard, we don’t. And so, we’ve begun this conversation. A lot of the conversation right now is not at the level of dealing with real sophisticated AI but has more to do with essentially states establishing norms about how they use their cyber capability. Who are you more afraid of: big brother and the state or the guy who’s trying to empty out your bank account? Part of the reason that’s so difficult is that if we’re going to police this wild west, whether it’s the internet, or AI, or any of these other areas, then by definition, the government’s got to have capabilities. If it’s got capabilities, then they’re subject to abuse. And, at a time when there’s been a lot of mistrust built up about government, that makes it difficult.\u201d<\/p>\n First Trump Administration and Biden Administration<\/p>\n Those were the early days.<\/p>\n the Biden Administration took major action in 2022 by restricting <\/a>exports to China, Hong Kong, and Macau<\/p>\n The first Trump Administration started a commission<\/a> advising Congress on maintaining AI leadership, including simply banning<\/a> the sale of some advanced semiconductor equipment and chips to China. Years of back-and-forth, a pandemic, an election, and Chat GPT\u2019s launch<\/a> later, then the Biden Administration took major action in 2022 by restricting <\/a>exports to China, Hong Kong, and Macau.\u00a0<\/p>\n NVIDIA\u2019s Ampere architecture A100 and newer codename \u201cHopper\u201d H100 GPUs and systems were all restricted. NVIDIA shed tears for $400 million worth of lost sales as a result and was especially sad when many of its export-compliant alternatives to these also got banned, like its newly-created A800, H800, and L40S, in addition to NVIDIA\u2019s RTX 4090 gaming card. NVIDIA said<\/a> it didn\u2019t expect \u201cnear-term meaningful impact\u201d on its financials<\/a>.<\/p>\n Second Trump Administration<\/p>\n NVIDIA responded<\/a> by designing a workaround to the previously<\/a> worked-around workaround, leading to the H20<\/a>. Then the government added <\/a>new rules for high memory bandwidth cards and the Biden admin tried to come up with an AI chip diffusion rule<\/a> that would limit the quantity <\/a>of GPUs being sold into different countries rather than only by processing power metrics, because the government really didn\u2019t know how the fuck to measure these things and NVIDIA, knowing more about GPUs, could tweak <\/a>any dial it wanted to just barely be compliant. Then DeepSeek <\/a>came <\/a>out and everyone panicked<\/a>, stocks <\/a>plummeted, and the government scrutinized<\/a> the role of NVIDIA GPUs in it. In February 2025, Fiscal Year 2025 results were posted and NVIDIA\u2019s Singapore revenue skyrocketed <\/a>to 18% of total revenue based on customer billing location despite shipments to Singapore being claimed to be less than 2% of Fiscal Year 2025 revenue, which caused people to say \u201cwait a minute.\u201d Unrelated: Several GPU smugglers<\/a> were arrested in Singapore one day after the fiscal year report was posted, which caused people <\/a>to say \u201cthat makes more sense.\u201d<\/p>\n In May, 2025, Trump implemented wide-sweeping tariffs <\/a>and rescinded<\/a> the<\/a> Biden chip diffusion rule that would have limited how many AI GPUs Jensen could sell to other countries<\/p>\n AMD spawned<\/a> out of nowhere to say it wrote-down $800MM of inventory due to export controls. NVIDIA one-upped it with<\/a> a write-off of $5.5 billion.<\/p>\n In May, 2025, Trump implemented wide-sweeping tariffs <\/a>and rescinded<\/a> the<\/a> Biden chip diffusion rule that would have limited how many AI GPUs Jensen could sell to other countries, Jensen then said<\/a>, \u201cIt\u2019s just an incredible vision. I think this is going to be a transformative idea for the next century for us. These 2 initiatives are completely visionary and it\u2019s going to be transformative for America.\u201d<\/p>\n The H20 GPU that was created<\/a> to comply with rules was still okay, then Jensen had dinner<\/a> with Trump at Mar-a-Lago for $1,000,000, then the H20 got banned<\/a> — he must have chosen a bad restaurant. In July<\/a>, Jensen Huang met with Trump and was permitted<\/a> to sell H20s again, Huang went<\/a> to China, China said Jensen\u2019s GPUs have tracking devices and backdoors, NVIDIA denied<\/a> that, Jensen went<\/a> to Washington, Trump then spoke very highly of Huang<\/a> (and Lisa Su), and more importantly, he ragged on Intel\u2019s CEO. Intel\u2019s CEO then went<\/a> to Washington, so Trump likes<\/a> him now. Tim Cook materialized from the infernal plane to give Trump a 24-karat gold \u201cgift,\u201d and \u201cgift\u201d is in quotes because that\u2019s not what that\u2019s called. Then we get to this past week, when Trump asked<\/a> NVIDIA and AMD to pay 20% to the US government for sale<\/a> SPECIFICALLY of the H20 and AMD Instinct MI308 sales, not all GPUs as some erroneously reported. Cousins<\/a> Jensen Huang and Lisa Su negotiated<\/a> Trump down to 15%, and now they\u2019re allowed to sell two specific cards that were originally created to comply with the laws before they changed and somehow everyone walks away a winner. Except now China doesn\u2019t<\/a> want them anyway<\/a>.<\/p>\n That about sums it up.<\/p>\n Except one last thing that happened as we were writing this: The Department of Commerce doesn\u2019t yet know the legality <\/a>of the deal, with Tom\u2019s Hardware highlighting <\/a>legal expert arguments over Article I, Section 9 of the Constitution.<\/p>\n If you want the full details with all of the in-between, make sure to check out our massive timeline article<\/a>.\u00a0<\/p>\n A New Law<\/p>\n There was also recent news about a 15% revenue share between NVIDIA and AMD with the United States Government<\/a> for sale of some AI GPUs. Trump stated<\/a>, \u201cThis is an old chip that China already has and I deal with Jensen who is a great guy and NVIDIA. The chip that we\u2019re talking about, the H20, is an old chip. China already has it in a different form, different name, but they have it. Or they have a combination of 2 will make up for it and even then some [\u2026] but the H20 is obsolete. You know, it\u2019s one of those things, but it still has a market. So I said, listen, I want 20% if I’m going to prove this for you, for the country, for our country, for the US. I don’t want it myself, you know, every time I say like 747, I want. Yeah, for the Air Force. So when I say I want 20, I want for the country. I only care about the country. I don’t care about myself and he said \u2018Would you make it 15?\u2019 So we negotiated a little deal. So he’s selling an essentially old chip that Huawei has a similar chip, a chip that does the same thing and I said ‘good, if I’m going to give it to you’ because they have a, you know, they have a stopper, what we call a stopper. Not allowed to do it. A restricted is really known as a restrictive covenant.\u201d<\/p>\n If you were to read <\/a>only <\/a>the headlines, you\u2019d think that this applies to all GPUs and that the ban is over, and then you might also think that a market in China, or in the US\u2019 eyes, an illegal market in China, would cease to be so illegal. That\u2019s not the case.\u00a0<\/p>\n This new 15% revenue share would, if it\u2019s legal <\/a>(and they\u2019re<\/a> not sure yet), allow NVIDIA to sell specifically the NVIDIA H20 GPU to approved Chinese entities. They likely can\u2019t be on the entity list<\/a>. It would allow AMD to sell specifically the Instinct MI308 GPU to approved Chinese entities. The proposed license would not affect any other GPU that currently does not have a license. The government hasn\u2019t made clear yet if NVIDIA\u2019s partners would also be permitted to make these sales.\u00a0<\/p>\n That means that other banned GPUs, including the RTX 4090, RTX 5090, H100, B100, B200, and so forth, remain banned.\u00a0<\/p>\n The H20\u2019s extremely high 96GB memory capacity would enable large models to fit in memory and actually run, especially with multiple GPUs in a single rack, even if it\u2019s slower. That means companies can achieve performance targets by stacking GPUs which are lower clock and core count but higher capacity.<\/p>\n As for the newer Blackwell architecture GPUs, Trump stated<\/a>, \u201cNow Jensen also has, Jensen\u2019s a very brilliant guy, and Jensen also has a new chip, the Blackwell. Do you know what the Blackwell is? The Blackwell is super duper advanced.\u201d<\/p>\n Let\u2019s not give NVIDIA any ideas on new GPU names.<\/p>\n Trump added, \u201cI wouldn\u2019t make a deal with that. Although it\u2019s possible I\u2019d make a deal, a somewhat enhanced in a negative way Blackwell. In other words, take 30% to 50% off of it. But that\u2019s the latest and the greatest in the world. Nobody has it. They won\u2019t have it for 5 years. On the Blackwell, I think he\u2019s coming to see me again about that. But that will be an unenhanced version of the big one.\u201d<\/p>\n For now, these \u201cSuper Duper\u201d GPUs are not licensed for sale.\u00a0<\/p>\n In short: The H20 and MI308 were compliant with the US Government\u2019s original rules<\/a>, then the rules changed during design & production and they were banned, then AMD<\/a> and NVIDIA collectively declared<\/a> over $6B in financial impact as a result.\u00a0<\/p>\n Then the US Government said, \u2018Wait a minute, we can help with that if you cut us in.\u2019<\/p>\n It reminds us of the unpredictability that we highlighted in our tariffs documentary<\/a>.<\/p>\n The H20 most certainly is not \u201cobsolete,\u201d though. It\u2019s still very desirable in China<\/a>, and with a lot of them, they become particularly potent.\u00a0\u00a0<\/p>\n How the US Determines GPU Bans<\/p>\n Setting a threshold for banned hardware should be objective since it can be tested.<\/p>\n This is a graph from the Department of Commerce that visualizes the original threshold at which a computing product became automatically banned for export to parts of the Middle-East and China without a granted license. There have been some changes since, but back when this was made, the \u201cTotal Processing Performance\u201d score on the Y-axis was used to determine cards in need of a license. The government needed a metric to calculate against, so it created its own.<\/p>\n Accelerators and video cards have a lot of metrics in their spec sheets, including memory capacity (which is critical and as simple as a pass\/fail for certain training and AI uses), memory bandwidth, GPU clock speed, GPU SM or CU count, TPCs, Tensor Cores, ROPs that are sometimes randomly missing on NVIDIA devices<\/a>, TFLOPS, TOPS, PFLOPS, GFLOPS, gigabits, power, and more.<\/p>\n So then, banning a product could probably be based on some sort of benchmark rather than a random metric from a spec sheet, but the US government, illustrating what an absolute clusterfuck this situation was and now remains, decided<\/a> to instead multiply one random metric from a specsheet against the bit length of the operation being executed. FLOPS, or Floating Point Operations Per Second, and TOPS, or Tera Operations Per Second, are calculated by the company making the spec sheet and aren\u2019t a great measurement of actual performance. These numbers are based on both marketing and whether we\u2019re talking about FP8, half-precision FP16, single-precision FP32, double-precision FP64, or Tensor performance, so the government said<\/a>, \u201cthe rate of MacTOPS is to be calculated at its maximum value theoretically possible\u201d and \u201cthe rate of MacTOPS is assumed to be the highest value the manufacturer claims in brochure[s] for the integrated circuit.\u201d So it\u2019s not based on a bunch of real-world benchmarks of applications or something useful. The government also references MacTOPS as the theoretical peak of TeraOPS in multiply-accumulate computations.<\/p>\n The Biden administration used this<\/a> \u201cTotal Processing Performance\u201d (or TPP) score in 2023, with the government later adding a \u201cPerformance Density\u201d metric dividing the TPP by the die area in square millimeters. In other words, the government didn\u2019t want NVIDIA to be able to sell more of a lower-performance GPU to make-up for the loss of high-performing parts with multi-GPU solutions.\u00a0<\/p>\n There are a lot of reasons this doesn\u2019t capture the full picture, like sparsity, APIs, differing methods to calculate FLOPS, and different performance for different applications, but the government needed a way to define a threshold, so this is what it made. The limit was<\/a> a TPP score of 4800, exceeded by even the RTX 4090<\/a> when calculated using Tensor performance.<\/p>\n Now, if this doesn\u2019t mean anything to you, that\u2019s OK, because it probably doesn\u2019t mean anything to people signing the laws either. Or maybe that\u2019s not OK, but you get the idea.\u00a0\u00a0<\/p>\n Something like a higher memory capacity, lower FLOPS performance GPU or even series of GPUs like RTX 3060 12GB cards might be able to get the work done more effectively if it only needs memory. Memory wasn\u2019t factored into the TPP calculation.<\/p>\n That was the point of the H20<\/a>, but then the absurdity of the situation expanded by introducing an opaque memory bandwidth requirement. The Register<\/a> wrote, \u201cUnlike with previous export controls, the Commerce Department’s Bureau of Industry and Security (BIS) hasn’t issued specific guidance on how much I\/O or memory bandwidth is too much.\u201d<\/p>\n So it seems that, across now two administrations, the United States is creating formulas based upon numerical calculations and then, when NVIDIA and AMD tweak numbers to fit within that box, it is retconning those rules in a guess-and-check process.<\/p>\n Experts & Roles<\/p>\n We spoke to a lot of people in this story across different languages, which made it complicated. The discussions come from:<\/p>\n Dr. Vinci Chow, Senior Lecturer in the Department of Economics at the Chinese University of Hong Kong, responsible for building his department\u2019s machine learning servers and sourcing GPUs from middlemen suppliers.<\/p>\n Dr. Z\u01d0 H\u00e1o F\u00f9, a Research Assistant Professor at the Chinese University of Hong Kong and Affiliated Lecturer at the University of Cambridge. Zihao specializes in both computer science and linguistics.<\/p>\n Various retail workers at the computer markets in Hong Kong to get a ground-level understanding.<\/p>\n Zhou, Creative Director of Product at video card manufacturer Yeston.\u00a0Yeston is not involved in any “black market” of GPUs and mostly deals with AMD; however, the company gave us a tour of its GPU factories to better understand production processes.<\/p>\n An anonymous seller who goes by the pseudonym of \u201cSILVER,\u201d based in Shenzhen Bao\u2019an and manages a warehouse in Hong Kong that receives and processes smuggled GPUs.<\/p>\n Vincent, a resourceful fence who lives next to the warehousing and markets filled with GPUs and accelerators. He buys and sells these devices in Shenzhen Huaqiangbei. His job is to know people.\u00a0<\/p>\n Vincent\u2019s Cousin is the fixer\u2019s fixer. If someone needs a component to fix a broken video card or to build a new one, they go to people like him to get the integrated circuits.<\/p>\n \u201cMr. 5,\u201d a Bilibili hardware reviewer with a specialized focus on thermal solutions. Like us, Mr. 5 has had run-ins with NVIDIA that have ended in a soured relationship over disputes regarding independent reviews and editorial independence.<\/p>\n Brother Zhang, a renowned Bilibili uploader (basically a YouTuber in China) who runs a video card repair shop in Zhengzhou. In addition to repairs, he regularly gets large orders from customers asking him to build them custom, unofficial, higher VRAM capacity NVIDIA GPUs for large language model tasks. Brother Zhang is not a direct part of any “black market,” as, again, his business operates legitimately within China as a repair shop. It does, however, come into contact with GPUs the USA considers to be “illegal” for sale into China.<\/p>\n Companies in Singapore and Taiwan who act as intermediaries between NVIDIA, NVIDIA\u2019s partners, and companies in China. The Singaporean and Taiwanese companies are able to bring banned GPUs and servers in and re-sell them to Chinese companies, skirting export controls. We are unable to disclose their identities as there would likely be punishment from multiple governments and NVIDIA.<\/p>\n The \u201cPlug,\u201d a US-based Chinese citizen who drives around the country buying hardware from American end-users and resells it to companies in China and Hong Kong.<\/p>\n And a special thanks to our translator Raymen Wu of the BLK SODA agency in Taiwan.<\/p>\n Here’s the pipeline.<\/p>\n If a GPU doesn\u2019t fall off the back of a truck in China after it was assembled, or if it isn\u2019t a “QC defect” that disappears from the scrap pile, it may instead be moved by \u201cants\u201d to get to China. There\u2019s a saying in Chinese that we heard a few times (mayi banjia – \u8682\u8681\u642c\u5bb6), which translates to \u201cants moving\u201d that represents a linear pipeline. Each ant in the fireline serves a specific role. It\u2019ll help to name those to keep everything straight. Here\u2019s what we came up with:<\/p>\n The Source has access to GPUs. This could be as innocent as you unknowingly selling your card on Facebook Marketplace to The Plug.\u00a0<\/p>\n The Plug is responsible for acquiring from the original source and re-selling the hardware to the China-based distributors.\u00a0<\/p>\n In between, there\u2019s a Mule. Sometimes this is the Plug himself doing a trip home. We learned that overseas students also regularly return with what are feasibly defended as personal GPUs, that they bought at a retailer like Best Buy, that may then get resold for markup and profit. In either case, the Mule gets the GPU into the country either by shipping it without interception or by hand carrying it.<\/p>\n Next is the Middleman, receiving the GPUs and often interfacing with or managing the warehouses that store the cards. The Middleman buys from multiple Plugs, including factories that get rid of rejects with fixable or unimportant QC defects, then sells those devices to more localized distributors.<\/p>\n That\u2019s when we get to the Fence, who buys and sells GPUs between middlemen and warehouses to end users in China.<\/p>\n Then, we have who we\u2019re calling the Fixer, except this time, it\u2019s literal. The Fixer is an optional step that may involve soldering and modifying a GPU to improve it beyond its original specification to make it more marketable for domestic AI uses. They might also just fix QC defects from the factories.<\/p>\n Finally, we have the User. This is self-explanatory: The user is the demand driver, and oftentimes, large enterprises want dozens or hundreds of GPUs or more, while smaller users like the university may just want individual units or small batches.<\/p>\n With everyone\u2019s role named, let\u2019s continue.<\/p>\n Demand Drivers<\/p>\n Our journey began in Hong Kong.\u00a0<\/p>\n Upon arriving in Hong Kong, one of our GPU dealer informants slipped us a price sheet for a mix of GPUs that are both export controlled and not.<\/p>\n Some of the banned ones on the list include the 5090, 4090, 4090 (48GB), A6000, A100, and A100 (80GB). The person who provided the price sheet works with smugglers. The problem is, we don\u2019t know what the going rate for these GPUs is.\u00a0\u00a0<\/p>\n So we hopped in a Hong Kong cab and headed across Victoria Harbor and over to the Chinese University of Hong Kong to meet up with Dr. Vinci Chow.<\/p>\n We found him through a Reuters story from a couple years ago. He works in the economics department at the university and he\u2019s responsible for having built many of the servers and machine learning systems that are in use daily.\u00a0<\/p>\n “Upon meeting Chow, he showed us some A100 GPUs, which he emphasized were, ‘legally obtained.'”<\/p>\n He has a whole blog<\/a> detailing his process, including the difficulty of sourcing components and the ease with fixing problems when so close to Shenzhen, such as having custom PCIe riser cables built and basically done the same day.\u00a0<\/p>\n Dr. Vinci Chow is the right person to start us off with this story and educate us on where to go next. We\u2019re here to understand the user\u2019s perspective because before there could be any market for it, there has to be demand. And that is what the university and organizations like it generate.\u00a0<\/p>\n it’s illegal to export to China without [a] permit but it’s completely legal on our side, right? There’s nothing that says that you cannot buy a high-end GPU.<\/p>\n – Dr. Vinci Chow<\/p>\n Upon meeting Chow, he showed us some A100 GPUs, which he emphasized were \u201clegally obtained.\u201d He elaborated that he obtained them around when ChatGPT was released and got them for \u201cquite cheap.\u201d He expressed that, during this time, even the A100, which became export-banned in 2022, \u201cwas actually quite cheap.\u201d He shared that he got the A100s for \u201cless than 10,000 US Dollars.\u201d Following ChatGPT\u2019s release, however, Chow said that the prices went \u201ccrazy.\u201d<\/p>\n Pointing out the oddity of the export ban, Chow stated, \u201cIt’s interesting because, you know, it’s the export ban, right? So it’s illegal to export to China without [a] permit but it’s completely legal on our side, right? There’s nothing that says that you cannot buy a high-end GPU. So, from our end, as long as we follow all relevant procedures, there’s absolutely nothing illegal about, for us over here, to buy these GPUs. So that makes it for a very interesting environment.\u201d He added, \u201cThese universities have been purchasing GPUs. It\u2019s completely legal on our side, right? Yet this is clearly not supposed to happen from the perspective of the US government.\u201d<\/p>\n Regarding pricing and availability of export-banned GPUs, the professor stated, \u201cIf you pay enough, supply is there. Maybe not if you want to build a super cluster, right? For research, most researchers are talking about 1 or 2 GPUs. And if they have the funding, then it\u2019s possible to obtain that. It\u2019s just everything is more expensive.\u201d<\/p>\n Having built Chinese University of Hong Kong\u2019s cluster of GPUs during COVID, he has first-hand knowledge of how much banned GPUs cost to get in Hong Kong and estimated that an H200 would cost around \u201c$30K.\u201d When we looked at our GPU price sheet, the H200 was going for 213,000 HKD, which amounted to roughly $29,700 USD. This corroborates his estimate.\u00a0<\/p>\n When we asked Chow how he thinks the export-banned GPUs get into China, he stated, \u201cSo these GPUs are almost certainly moved one at a time, right? It\u2019s very hard to get a full HGX system.\u201d\u00a0\u00a0<\/p>\n Later, Chow introduced us to his friend and associate Dr. Z\u01d0 H\u00e1o F\u00f9, who works in the university\u2019s linguistics department and has a background in computer science. He uses the university\u2019s mainline datacenter to train large-language models. He shared that they currently use A40<\/a> GPUs, gaming GPUs, and sometimes have an RTX 6000 available (NVIDIA has multiple generations of RTX 6000, including Ada Lovelace and Blackwell). Z\u01d0 H\u00e1o elaborated that memory is the \u201cmost important part for researchers.\u201d<\/p>\n When we asked Chow (hypothetically) if they could get whatever GPUs they wanted with limitless money, he explained that if they had millions of dollars before the GPU ban, they could get them. He also stated that, \u201cWith the export ban, [university] departments are not necessarily willing to publicly list their computing capabilities.\u201d<\/p>\n Chow showed off the GPU cluster he constructed. He explained that a single 8-GPU system can put out around 4,000 watts. For reference, this would require 2 standard US residential circuits to support at 20A per circuit or, as Dr. Chow says, would be about \u201ctwo hair dryers.\u201d<\/p>\n He also shared that you can\u2019t plug the GPUs into a standard wall socket as that would immediately blow the fuse and that they had to install 2 three-phase power within their GPU cluster room.\u00a0<\/p>\n Chow shared that they have roughly 30 A100-class GPUs and a bunch of 3090s and 3060s. This amounts to roughly 50-60 GPUs.<\/p>\n Discussing the logistics of how the US determines which GPUs should be banned, Chow stated, \u201cIf you go back like five years ago, everyone probably would think that like, yeah, FLOPs is a very important metric to consider. But turns out when it comes to loading large models, what we all care about is how much memory, how much VRAM you have, right? You simply cannot load a model if you don’t have the VRAM. So now H20 becomes a very attractive option.\u201d He added, \u201cthe H20 has drastically lower FLOPs, but then it actually has more memory than the original H100.\u201d For reference, the H20 has 96GB of memory<\/a>, whereas the H100 has 80GB<\/a>. Chow added, \u201cSo it\u2019s actually in some sense more attractive.\u201d<\/p>\n While the idea of black markets might convey back-alley deals, Chow says that the export-banned GPUs are simply shipped to customers.\u00a0<\/p>\n When we asked if the GPU bans were effective, Chow responded, \u201cIt is effective in preventing the building of a very big cluster. It\u2019s just not possible to get 100,000 of the GPUs.\u201d\u00a0<\/p>\n He mentioned that the H200s were also assembled in China, which raises an interesting question of what prevents these GPUs from ‘falling off the line.’ Chow elaborated on that, stating,\u00a0 ‘I don\u2019t understand how it works at all. How is the ban even working?’<\/p>\n When we asked them why they think the US government cares so much, Z\u01d0 H\u00e1o stated, \u201cI guess they just want to delay the speed of other countries of training the model, but their ban is very weird.\u201d When we asked them if the ban seemed targeted at military uses or Chinese companies in general, Z\u01d0 H\u00e1o said, \u201cI guess it\u2019s in general.\u201d Chow chimed and said, \u201cI don’t think, from the US government’s perspective, there’s this distinction between like Chinese academia versus Chinese military versus Chinese commercial. I don’t think they really consider there’s a distinction.\u201d<\/p>\n When we asked to see if Huawei\u2019s hardware seemed compelling, Z\u01d0 H\u00e1o shared that he doesn\u2019t know people who use Huawei hardware, but it may be a fallback option in the future if consumers in China can\u2019t get NVIDIA GPUs.\u00a0\u00a0<\/p>\n All these high-end GPUs are still manufactured in China, but they’re not allowed to be sold in China.<\/p>\n – Dr. Vinci Chow<\/p>\n Chow explained to us that the A100 GPUs he acquired were \u201cassembled in China.\u201d He mentioned that the H200s were also assembled in China, which raises an interesting question of what prevents these GPUs from \u201cfalling off the line.\u201d Chow elaborated on that, stating,\u00a0 \u201cI don\u2019t understand how it works at all. How is the ban even working?\u201d He points out that the box for the banned GPUs even \u201cclearly state that they are manufactured in China.\u201d Chow shared a theory on how banned GPUs are able to be sold in China, \u201cMy guess is there must be spares. I don’t know. Spare SXM modules, spare casings, and then somehow these spare parts just get assembled into a complete GPU and get sold.\u201d<\/p>\n I would be surprised if they don\u2019t, right? I would be surprised. I would be really surprised if they don’t. These are very expensive items. I would imagine you would keep track of everything<\/p>\n Dr. Vinci Chow<\/p>\n Recapping the ridiculousness of the situation, Chow stated, \u201cAll these high-end GPUs are still manufactured in China, but they’re not allowed to be sold in China. And somehow the US government thinks that’s going to work and somehow the Chinese government also allows that to happen. I have no idea how actually that whole thing works.\u201d<\/p>\n We asked both professors if NVIDIA knows whether all of this is happening and Chow stated, \u201cI would be surprised if they don\u2019t, right? I would be surprised. I would be really surprised if they don’t. These are very expensive items. I would imagine you would keep track of everything.\u201d\u00a0<\/p>\n Hunting for Banned GPUs (Golden Computer Center)<\/p>\n After this discussion, Zihao parted ways and Vinci brought us to one of his favorite tech spots in Hong Kong: The Golden Computer Center and the outdoor Apliu Street tech flea market. We asked him if we could go find some supposedly banned GPUs, like the RTX 5090, just available out in public. The hope was to find a shopkeeper with some ground-level or consumer knowledge.<\/p>\n When we asked Chow if the export controls are having their desired effect for the US, he responded, \u201cWell, yes, in terms of stopping China from building a comparable GPU cluster to the US.\u201d<\/p>\n While we were visiting Golden Computer Center, Chow pointed out some purchasable \u201cparallel import\u201d GPUs, which means they were smuggled into the country. We saw banned cards for sale here, including the 5090 Founders Edition. Despite the sometimes inflated prices, we were surprised at how easy these GPUs were to purchase. It seems like one of the main impacts of the GPU bans are the prices of the cards. Chow theorizes that the cards are moved into the country one by one, often by traveling students.\u00a0<\/p>\n We asked a shopkeeper there, \u201cWhat\u2019s the most popular card?\u201d He said it was the 5090 and that they come from Australia and Taiwan, neither of which has an export ban on these GPUs to China. He also said they can \u201cbuy a lot\u201d of 5090s from mainland China.\u00a0<\/p>\n Hong Kong was impressive and filled with character and culture — and GPUs, apparently, because one of our next sources sent a message the night we were planning to hitch a ferry to Shekou Port in Shenzhen. He told us that we could buy a GPU as soon as tomorrow if we wired him money immediately.<\/p>\n That seems like the responsible thing to do, and because it\u2019s fully above board as a buyer, and because he texted us a photo of an RTX Pro 6000, which was very intriguing since it\u2019s Blackwell, we decided the best way to learn more about buying export-controlled GPUs in China would be to just do it. So far, all the contacts were aware that some form of factory repurposing, theft from the line, QC rejects, and actual by-hand smuggling are involved — but none knew for sure how the GPUs move. To get closer to the sources, we decided to wire the funds and get an address.<\/p>\n We ended up sending out $3,289 via wire on a tight deadline, with our boat bringing us to the seller the next morning.\u00a0<\/p>\n Turn a Blind Eye: How GPU Smuggling WorksHUIZHOU: Mr. 5<\/p>\n We then packed up and boarded a ferry to Shekou Port, but we decided to take a quick detour: Rather than go straight to Shenzhen to meet our GPU plug, we first went to meet up with someone else — a source who told us he has more information on how GPUs get into China. We traveled to HuiZhou to meet a guy known as \u201cMr. 5.\u201d\u00a0<\/p>\n HuiZhou is a city with some serious grit.<\/p>\n The city has grown to take some of Shenzhen\u2019s factory industry as it\u2019s been pushed out over the years with Shenzhen\u2019s expansion and as it\u2019s turned into more of a metropolis. We spent a good amount of time in HuiZhou over the last decade, mostly visiting case, painting, tempered glass, and tooling factories.<\/p>\n “we asked him how 5090s get into China. He responded, ‘It\u2019s like this. China already produces the heatsinks and components. China makes a lot of 5090s. That\u2019s the first way in.'”<\/p>\n Mr. 5 is a cooling hardware reviewer and has specialist knowledge in factories that make video card cooling solutions, including NVIDIA\u2019s. This experience allows him more access to information about the peculiar relationship between factories making cards and the companies that technically can\u2019t sell them to the country where they\u2019re made. We instantly related to Mr. 5 for his own editorial disputes with NVIDIA\u2019s review sampling process, where we have felt the company seeks to control review direction.\u00a0<\/p>\n They can buy it and use a \u2018human flesh backpack.\u2019 They carry that back and it\u2019s not illegal in China<\/p>\n – Mr. 5<\/p>\n His username is \u201c51972\u201d on Bilibili<\/a>.<\/p>\n \u201cHuman Flesh Backpack\u201d<\/p>\n Speaking to him in Chinese, we asked him how 5090s get into China. He responded:<\/p>\n \u201cIt\u2019s like this: China already produces the heatsinks and components. China makes a lot of 5090s. That\u2019s the first way in. The second is […] The US has a ban on sales to China, but there\u2019s no ban in China. Because Shenzen and Hong Kong are close, there\u2019s only one customs check. Many people can get it from Hong Kong or other countries. For example, America, Japan, Singapore, etc. They can buy it and use a \u2018human flesh backpack.\u2019 They carry that back and it\u2019s not illegal in China. They can purchase it that way. There\u2019s one other method: for some people, it\u2019s just for money and they choose to smuggle. So they use a \u2018special channel\u2019 (smuggling) to get it back.\u201d<\/p>\n A few months ago, if you carried an RTX 5090 in from outside, you could earn about 2,000 to 5,000 RMB. You could make so much money.<\/p>\n – Mr. 5<\/p>\n We asked Mr. 5 if he thought most of the banned GPUs come in one at a time, and he replied:<\/p>\n \u201cThere are many ways. In China, we have a special type of \u2018job,\u2019 scalper. He can go back-and-forth many times in one day to bring them back, and every time he does, he brings some back. He can also organize people or a group of people to go to Hong Kong together to buy. Then he gets the difference in price. Each card gets from 500 to a few thousand RMB. That\u2019s the method he uses. Strictly speaking, this kind of action is illegal, but it\u2019s a gray area. They can organize people from Shenzhen, Hong Kong, or other countries. Of course, there are many other ways, like international students bringing them back. In China, the 5090 is not banned.\u201d He added, \u201cA few months ago, if you carried an RTX 5090 in from outside, you could earn about 2,000 to 5,000 RMB (about $280 to $700 USD). You could make so much money. The craziest is when the 5090 released during Chinese New Year, some people who brought one back from overseas made 10,000 RMB (about $1,400 USD). So a lot of people thought, \u2018If I go abroad, I\u2019ll buy an extra one.\u2019\u201d\u00a0<\/p>\n Of course [NVIDIA] know. For this matter, of course they know, but, how, \u2018how do I say this?\u2019 \u2018Open one eye, close one eye.’<\/p>\n – Mr. 5<\/p>\n Mr. 5 stated, \u201cChina has an old saying, \u2018All the hustle and bustle in the world is only for money and interest.\u2019 I also have to add that many of the RTX 5090s are \u2018made in China.\u2019 The video card was born here locally, so a few brands choose to sell domestically in order to digest inventory. Because it takes a few months to ship by ocean freight–we already know it takes 1-2 months to get from China to the USA by boat, which wastes time. But if it\u2019s sold domestically in China, it can be turned-around quickly. Funds return faster. This is also lower pressure and reduced inventory.\u201d<\/p>\n NVIDIA\u2019s Awareness of Smuggling<\/p>\n When we asked Mr. 5 if he thinks NVIDIA knows about what\u2019s going on with all of the banned GPUs being sold in China, Mr. 5 replied:<\/p>\n \u201cOf course they know. For this matter, of course they know, but, how–how do I say this?–\u2018Open one eye, close one eye.\u2019\u201d This translates to \u201cturn a blind eye.\u201d\u00a0<\/p>\n When we asked Mr. 5 if NVIDIA would want to stop it, he simply replied, \u201cNo\u201d since the Chinese market is so big.\u00a0<\/p>\n SHENZHEN: Suppliers, Middlemen, & Fixers of Banned GPUs Our next stop was in Shenzhen Bao\u2019an to meet with a trading company that sits between Hong Kong and Shenzhen warehouses. They\u2019re the trading company\u2019s trading company, and we planned to buy an RTX 5090 from them.\u00a0<\/p>\n Shenzhen was a fishing village just 40 years ago. Now, it\u2019s one of the most technologically advanced cities on earth. Shenzhen has some extreme Cyberpunk vibes with its mix of technology and surveillance.<\/p>\n It has packed tech markets in Huaqiangbei<\/a>, including some of the weirdest computer parts we\u2019ve ever seen. There are entire buildings dedicated only to small phone repair shops with their own kiosks, another building dedicated to a mix of computer hardware, gaming, and miscellaneous components, more still for just integrated circuits, and all of these places have people who know people. That\u2019s their job.\u00a0<\/p>\n But with Shenzhen\u2019s technological rise comes with it an uncomfortable omnipresence of CCTV and facial recognition, which feels more fitting today than ever before now that we know worldwide government facial recognition is a big user of AI GPUs.\u00a0<\/p>\n Buying an \u201cIllegal\u201d GPU We started the day by meeting up with our translator for the next two days, Raymen, who\u2019s helped us on factory tours for years now. Because Uber doesn\u2019t work in China — since it uses the Google Maps API and Google is blocked by the firewall — Raymen also was our man with the locally compatible apps to get us places. This became especially important now that even cash is becoming less acceptable by some cafes and restaurants, as everyone has moved to paying with the WeChat app.<\/p>\n With our ride booked, we set off to see if the wired funds turned into a GPU and tried meeting up with our GPU supplier.<\/p>\n Upon meeting with them, we explained that we were media, and they allowed us to record them talking but didn\u2019t allow us to show their faces or divulge their names or company name.\u00a0<\/p>\n They gave us the RTX 5090 we purchased, which wasn\u2019t even much more expensive than what we would have paid for it in the US.\u00a0<\/p>\n We asked them what the most common GPU is that their customers buy and one of them, whom we\u2019ll call \u201cSilver,\u201d stated the H800, A800, and 5090 D. These are are all banned GPUs (the 5090 D was not originally banned). We then asked them what sells better between the 5090-class cards and the H100 type GPUs, and they replied \u201c5090 D.\u201d\u00a0<\/p>\n We asked them if anyone buys high-end AMD GPUs and they responded, \u201cVery few. AMD\u2019s GPUs are rarely useful. We\u2019ve had second-hand customers, but high-end is very rare.\u201d We then asked to see if people were buying Intel GPUs, and they responded, \u201cIntel is the least!\u201d<\/p>\n ‘\u7741\u4e00\u53ea\u773c, \u95ed\u4e00\u53ea\u773c “https:\/\/gamersnexus.net\/”Turn a Blind Eye’<\/p>\n – “Silver”<\/p>\n We then asked them if they think NVIDIA knows about people buying and selling banned GPUs in China and they echoed what Mr. 5 said and replied, \u201cThey \u2018open one eye, close one eye.\u2019 They know you can buy it.\u201d\u00a0<\/p>\n The most important thing we learned here was where to go next: Just like how the ants move GPUs piece by piece, we\u2019ll have to collect our information piece-by-piece. The company said that some cards have become more difficult to get, but they can still get them.\u00a0<\/p>\n The fact that they texted us a photo of an RTX 6000 PRO Blackwell card — the very same that we just bought for $8,500 in the US<\/a> and that is hard to get at home — shows that they\u2019re resourceful. Their price is $8,600 US, which is actually cheaper when factoring-in US taxes and shipping.\u00a0<\/p>\n So the ban isn\u2019t stopping them, and the 6000 PRO Blackwell GPU is a serious AI card (that we\u2019ve benchmarked<\/a>) with 96GB of memory. It\u2019s banned and in demand. This company isn\u2019t used to single-GPU sales like ours: It mostly transacts with other trading companies and in high volume, interfacing with a Hong Kong distributor to bring the cards into Shenzhen, moving it across one more border, another ant in the chain. That meant they could connect us with their distributor, which would be familiar with where we could find smugglers, but this company itself neither knew many smugglers directly nor knew many end users. That\u2019s OK, because each link in this chain will get us one ant closer to the information we need.<\/p>\n We took note of their information on the distributor and smuggling side to use when we got back home, then they helped point us toward our next stop: Huaqiangbei, which is located in Shenzhen.<\/p>\n We hopped in a car and drove 40 minutes to meet with a GPU trader we found in the city, over near Huaqiangbei.\u00a0<\/p>\n SEG E-Market: The Biggest Tech Market in the World Huaqiangbei is home to the world-famous SEG E-Market, or Saige, where we\u2019ve found some of the strangest computer parts we\u2019ve ever seen<\/a>. Huaqiangbei has the highest concentration of technology and integrated circuits in the world, with a neverending maze of multiple disconnected, multi-story malls specializing in all electronics. If your life depended on getting a complete product made in a single city block, Shenzhen Huaqiangbei is your best bet.<\/p>\n Fortunately, this time, we\u2019ll be with this guy:<\/p>\n This is Vincent. His profile picture on one messaging app is Van Gogh, so he has a sense of humor. He seems to have a natural ability to make people — and cars — get out of his way and he argued with security guards about us being able to film.<\/p>\n Vincent generally seems to have an attitude of getting shit done.<\/p>\n But before we met him, we had some concerns going into this one that it\u2019d be fruitless or that no one would even be there to meet us. Luckily, he met up with us after we arrived.\u00a0<\/p>\n We can only move [GPUs] slowly, but there are a lot of people in China.<\/p>\n – Vincent<\/p>\n We asked Vincent if it was difficult to get the GPUs. He replied, \u201cThe market has a lot of them\u201d and pointed us towards the nearby SEG E-Market. Vincent runs a trading company and buys and sells the GPUs. We asked him how the high-end GPUs get into China, and he said, “They often come through Hong Kong\u201d and added that \u201cTaiwanese people also sell them here.\u201d He confirmed that the GPUs are often brought over one-by-one. When we asked him if this one-by-one movement was enough, Vincent replied, \u201cWe can only move them slowly, but there are a lot of people in China.\u201d<\/p>\n If it\u2019s a common card like a 5090, we have them in China. Chinese factories made those [5090s] and sell them<\/p>\n – VincentConsequences of Smuggling<\/p>\n We then proceeded to ask if there were any consequences from bringing those cards in. He replied, \u201cIf it\u2019s a banned video card, then China doesn\u2019t have any first-party [cards]. You can only bring it in from outside. If it\u2019s a common card like a 5090, we have them in China. Chinese factories made those [5090s] and sell them. That\u2019s the way it is. If you\u2019re talking about high-tech servers, you can only bring them in from outside. China doesn\u2019t have these [high-tech parts].\u201d<\/p>\n When we asked Vincent if China cares, he responded, \u201cIt\u2019s not the Chinese government\u2019s business. It\u2019s a US ban.\u201d<\/p>\n He revealed that \u201ceach person has their own method to get cards in. Normally, I just get it from Hong Kong because it\u2019s close.\u201d<\/p>\n We asked Vincent what his normal order amount was and he replied, \u201cRelatively low. Just 1 or 2 [per customer].\u201d\u00a0 We then asked what was the most in-demand GPU, and he responded, \u201cThe 4090 is relatively popular, but the 5090 is too expensive. No one wants it.\u201d When we pointed out that the 4090 was banned, he was surprised and stated, \u201cReally? 4090? It\u2019s relatively common.\u201d<\/p>\n We found it interesting that some dealers of “banned” GPUs aren’t even aware that they are banned, illustrating just how easy these parts are to get.<\/p>\n Vincent stated, \u201cWe don\u2019t know the reason for the ban, we can only buy and sell what\u2019s on the market.\u201d<\/p>\n We then asked Vincent if we could follow him to SEG E-Market and walked there with him.\u00a0<\/p>\n Vincent led us over to various large wholesale and consumer retail markets. His office is within about a 10-minute walk from SEG E-Market, which is his daily haunt.<\/p>\n There\u2019s one peculiar detail in all of this: Vincent has customers both in and outside of China, but his foreign customer base is relatively high. He mostly deals in RTX 4090s, which seems to be a trend as you\u2019ll see with our smuggler contact at the end.<\/p>\n That means that Vincent is buying 4090s that either never left China — but were supposed to — or were likely illegally re-exported to China, and is then re-re-exporting them to his foreign customers. It\u2019s like an infinite loop.<\/p>\n He used to work as a shopkeeper in the mall itself and would interface with factories or anyone else who wandered in.\u00a0<\/p>\n we went into a shop and asked if they had 5090s. The shopkeeper said that they do<\/p>\n These days, he\u2019s closed shop and is a buyer instead. He mostly takes orders online, doesn\u2019t keep much inventory, and then just walks across the street to buy whatever was ordered that day. It\u2019s basically a concierge GPU-picking service.<\/p>\n While at SEG market, we went into a shop and asked if they had 5090s. The shopkeeper said that they do and they confirmed that they did have them and could sell them for 19,800 RMB, which was roughly $2,757 USD. That\u2019s not bad and is pretty close to US pricing.\u00a0<\/p>\n Speaking to shopkeepers, and like with the other interviews, we learned that people rarely want AMD GPUs and basically no one wants Intel.<\/p>\n Along the way, he introduced us to his cousin in the integrated circuit business.\u00a0<\/p>\n Vincent walked us around and repeatedly expressed confusion at the ban list. He sees RTX 4090s everywhere and RTX 5090s are easy to get<\/p>\n His cousin connected a missing link for us: People like him, selling FETs, inductors, capacitors, resistors, MLCCs, and so on for repair shops, can help to supply the parts needed for upfit or modifications. For GPUs, this can also include sourcing additional VRAM for modifications.<\/p>\n some sellers told us that they\u2019re capable of sourcing devices like the A800 or even A100<\/p>\n Vincent walked us around and repeatedly expressed confusion at the ban list. He sees RTX 4090s everywhere and RTX 5090s are easy to get, just expensive. He was pretty sure that their abundance meant we were mistaken, but we checked and re-calculated, and they are export controlled devices. He told us about how 5090s are easy to get because the local factories supply them, matching Mr. 5\u2019s comments earlier. Vincent pointed out everything from RTX 20-series GPUs, GTX 16 cards, 30-series, A1000 cards, Quadros, BTC mining rigs, and 5090s, and 4090s. Although some sellers told us that they\u2019re capable of sourcing devices like the A800 or even A100, they did not have them on-site. They have them in a more secure spot.\u00a0<\/p>\n That\u2019s because there\u2019s a separate area for the warehousing. Fortunately, we learned where they are from our new friends in SEG. Like every other link in this chain, we kept getting closer to understanding the full story.<\/p>\n On a rainy morning, we headed over to the warehouses containing all of the GPUs to get a better idea for how many there might be. SEG\u2019s guards have never been particularly friendly toward filming to begin with, but ultimately, they\u2019re not police. The worst that happens is we get asked to leave. Besides, there weren\u2019t any signs saying we couldn\u2019t be there.<\/p>\n From the perspective of the United States, the warehouse we visited is filled with highly illegal items. They didn’t have export licenses.<\/p>\n From the perspective of people working here, however, they\u2019re just doing their job. It’s not shady. It’s not some subterranean refuge of firearms. It’s just a poorly lit warehouse like any number of other warehouses around the world, but the perspective of the American government would be different from the people working here, who just want to sell these things and go home.<\/p>\n If GPU sellers don\u2019t have certain GPUs, they purchase them from these warehouses. We saw stacks upon stacks of GPUs for sale. These feed into SEG and get distributed to Alibaba sellers like Vincent and get sent out from there. To be clear, not all of the GPUs we saw in the warehouse are banned, but we did see a couple 5090s.\u00a0<\/p>\n From the perspective of the United States, the warehouse we visited is filled with highly illegal items. They didn’t have export licenses. From the perspective of people working here, however, they\u2019re just doing their job. It’s not shady. It’s not some subterranean refuge of firearms. It’s just a poorly lit warehouse like any number of other warehouses around the world, but the perspective of the American government would be different from the people working here, who just want to sell these things and go home.<\/p>\n The warehouse security did eventually kick us out for filming, although they were oddly polite about it, which we appreciated.<\/p>\n Zhengzhou Next up, we headed to Zhengzhou on a 5 hour and 50 minute high-speed rail ride to visit Brother Zhang\u2019s repair shop, where they build custom 48GB GPUs.\u00a0<\/p>\n Brother Zhang is famous for repairs on Bilibili. He\u2019s got video cards and spare parts all over his shop and he saves everything that he can salvage, then rips the good components from the boards to reuse them for other ones. Brother Zhang is an interesting guy in the chain because he\u2019s not directly part of the GPU black market — he just fixes video cards. He does that for consumers and for companies. Sometimes, that means people want him to modify video cards to, for example, double the VRAM. That\u2019s where it gets interesting for our story.<\/p>\n Speaking to him, we learned that his shop tests about 50 GPUs a day and, while we were there, we saw them testing 4090s and 5090s. Most of the GPUs are from NVIDIA, but there are some from AMD, with relatively few from Intel. Most of the cards that come into his shop come from all throughout China and most of his customers know him because of his videos.<\/p>\n “He converted a 24GB RTX 4090 into a 48GB RTX 4090”<\/p>\n The ultimate reason we\u2019re at Zhang\u2019s shop is to see how video cards are modified. Because if there\u2019s an export ban on GPUs going to China, then it becomes critical for China to be more self-sufficient in keeping those GPUs that they do manage to get to stay in service. Most of the repairs are more typical things you’d expect of soldering. There’s board heaters, soldering irons, surface mount components gathered from suppliers nearby, and a lot of test stations. But one thing that’s unique to China and especially to this shop is the ability to take an existing model video card and completely modify the SKU into something that NVIDIA doesn’t even make. NVIDIA is intentionally restrictive with how much RAM it puts on cards. Part of that is to upsell people to more expensive models. Actually, that’s pretty much all of it. That’s basically why they do it. But at shops like Zhang’s, they double the VRAM on some of the cards that come through, making it into a fully custom SKU. We wanted to see if he had one lying around that he could show us, and he did one better than that: his shop made a modified 48GB RTX 4090 right in front of our eyes.<\/p>\n He converted a 24GB RTX 4090 into a 48GB RTX 4090. The unit that we looked at went through 6 repairs: It originally didn\u2019t work at all, then it had a memory problem, then it had display issues. While the card works now and the GPU itself is functional, because its board has had so many issues, it would be better to harvest the expensive components (like the GPU) and put them on a new board to avoid other potential failures.\u00a0<\/p>\n When we asked Zhang what the success rate was, he told us it was about 99%. Impressive.<\/p>\n Because NVIDIA doesn’t make a 48GB 4090, the repair shop has to source its own PCB and cooling solution. Other shops commonly sell both of these on the open market in China. It needs something that has enough pads for all the memory modules they’re adding to it, more memory, and a pre-populated VRM. They work with a third-party supplier that builds a PCB, uses an SMT line to place all the VRM components, and has the extra wiring, circuitry, and pads to support 48GB of memory.\u00a0<\/p>\n To create a 48GB 4090, the first step is to disassemble the card.<\/p>\n From here, they heat the GPU to 260 degrees Celsius for about 5 minutes, which allows them to pull the GPU off the PCB. Zhang revealed to us that they get about 10-20 orders at a time and, when we asked him if NVIDIA had contacted him to tell him to stop, he responded, \u201cI don\u2019t think they will.\u201d<\/p>\n The next step is to work on the memory. To do this, they place a PCB on a board heater and use a hot air station to heat pinpointed memory modules and remove them without risking damage to other components on the board. From there, they\u2019ll add them to a newer board. This, ultimately, allows them to double the card\u2019s memory capacity.\u00a0<\/p>\n Next, they took out a template for the solder balls and cleaned it with rubbing alcohol. From there, they poured solder balls into the template and positioned it on top of the memory modules. This allowed them to bake the new solder balls onto the memory modules, which they could then attach to the PCB. They then applied flux to wet the solder balls and used a solder wick to pull off the excess solder.<\/p>\n From start to finish, it took [the shop] about 2 hours to make a 48GB 4090<\/p>\n When we asked Zhang what 48GB RTX 4090 cards normally sell for, he told us over 20,000 RMB, which is roughly equivalent to $2,785 USD. That\u2019s not a bad price.\u00a0\u00a0<\/p>\n From there, he put a lid on the template and poured in soldering balls, allowing them to adhere to the memory. They then heated the modules on a hot plate that ran at around 195 degrees Celsius. Because the template only allowed them to do 8 at a time, they had to do the solder ball process twice. They then applied heat directly to the surface where the solder balls were using a hot air station.<\/p>\n The next step involved cleaning the GPU in order to mount it to a new PCB. From there, they added flux to the GPU. They then applied a custom template to the bottom of the GPU and poured solder balls on top of it and used the template to sift the solder balls.\u00a0\u00a0<\/p>\n The next thing the shop did was place the GPU on top of a jig to hold it in place on top of a hot plate to protect it.\u00a0\u00a0<\/p>\n They then placed the memory modules onto the PCB and, again, used the hot air station.<\/p>\n Finally, after about 2 hours, the shop added the GPU to the PCB and placed it under heat to bake the GPU back on.\u00a0\u00a0<\/p>\n From there, the team cooled the GPU off with a fan and added thermal paste. A technician placed a heat sink on top to test the card to see if it worked, leading to an instant success.<\/p>\n From there, they attached the GPU and PCB to the rest of the video card\u2019s cooler and chassis. From start to finish, it took them about 2 hours to make a 48GB 4090.\u00a0<\/p>\n At the beginning of this process, the card was working, but it had been through so many repairs that it was very likely to break again at some point from something they couldn’t really predict. From a waste standpoint, it’s just better to figure out a way to save that card because the PCB is kind of the least valuable part and is also the least precious of the resources. Being able to save the silicon that’s in the memory and in the actual GPU itself provides a lot of value, but it especially offers value in China, where they just don’t have as much supply, even though the supply looked good from what we saw. Purely from the perspective of people in China, not only is this better than what NVIDIA shipped to begin with, and really not that far off in price when the card was at its most scalped, it’s also just a good way to keep the silicon in circulation even as things like MOSFETs, capacitors, or PCBs die.<\/p>\n The silicon tends to be pretty resilient and it doesn’t really die that often on GPUs. What this tells us is that this is maybe an alternative method to getting more GPUs into supply. If they take the broken boards and then put functional silicon back into circulation, it’s a certain level of ingenuity, and there aren’t many places that are set up to handle it. Brother Zhang’s shop is impressive.<\/p>\n Video Card Factory Next, we headed back to Shenzhenbei in a high-speed rail trip that added roughly 1,000 miles to our journey, planning to meet up with Yeston Creative Director of Product Zhou. Yeston is a video card manufacturer.<\/p>\n Zhou and Yeston were not officially contributing to the black market side of our research, but they did show us around their factory and there were a few interesting things we learned that relate to our story. The big takeaway is that factories receive their GPU supply from their partners. That would be AMD, Intel, or NVIDIA, and the memory is often packaged with it. But for the most part, everything else that they use, unless they’re buying reference PCBs, all come from whatever sources they want to find for their supply.\u00a0<\/p>\n for at least the gaming class cards, which the 5090 more or less is one of, there’s enough volume and NVIDIA\u2019s hands off enough that it would be pretty easy to make these disappear<\/p>\n We were also informed that there’s very little oversight in terms of the management of the rejects at these places. They can file for refunds if, for example, a chip is bad. But if they screw up a board in their own process, then it’s going to be on them to deal with the defect.<\/p>\n Factory work for GPUs is highly automated. For at least the gaming class cards, which the 5090 (more or less) is one, there’s enough volume and NVIDIA is hands-off enough that it would be pretty easy to make these disappear. The biggest reason for this is that there are partners for the 5090s, unlike cards that only NVIDIA makes, such as the RTX Pro 6000. Since there are partners, NVIDIA is only tracking when they sell those partners the GPUs. NVIDIA doesn’t necessarily keep tabs on where those get distributed once they’re done being manufactured. For this reason, if a factory wants to just sell the GPU domestically in China and they think they can get away with it without being added to the entity list, which would very likely kill their business with NVIDIA, then they might sell them domestically. That might be how some of these trading companies get a hold of them. It’s an easier way to make them disappear than through channels that are more publicly visible from NVIDIA and the government.\u00a0<\/p>\n Yeston doesn’t make 5090s, but the processes they taught us about illustrates how that could happen.\u00a0<\/p>\n As for the server grade solutions, there are manufacturers for those that are pretty tightly controlled. Basically, the silicon is made in Taiwan, shipped to wherever it’s going to be assembled, and accompanied by the other components sourced in China. As a result, it’s easiest to keep the whole supply chain there, including the assembly.\u00a0<\/p>\n after seeing how a highly skilled group like Brother Zhang’s repair shop can just repurpose all those components that are still good, you could see how there would be value in figuring out a way to make that board disappear, even if it’s supposed to be in the trash<\/p>\n Factories use surface mount technology (SMT) lines with conveyed inflow of PCBs and reels upon reels of components that are placed smallest to largest, with the most valuable typically at the very end. That would often be the GPU and VRAM. Heatsinks are installed often through a manual process at the end. There is often manual assembly even for NVIDIA’s high-end products.<\/p>\n We have seen NVIDIA’s server video cards (and servers) being made in different factories. If the factory finds a defect, however unlikely, it could ‘disappear’ into scrap to be resold to the domestic market later. Dr. Vinci Chow had such a board where he had a broken link, and so there’s a good chance that card stayed in China and never left. If it was reported as a defect to NVIDIA, then it may have effectively just been written off and considered trash. After seeing how a highly skilled group like Brother Zhang’s repair shop can just repurpose all those components that are still good, you could see how there would be value in figuring out a way to make defects — or “defects” — disappear, even if it’s supposed to be in the trash.\u00a0<\/p>\n This comes back to NVIDIA turning a blind eye, because they’d eventually notice it from the serial number popping up somewhere. NVIDIA might choose not to notice it, though. Besides, once someone has a card, especially in China, there’s not a ton NVIDIA can do, anyway.<\/p>\n Taiwan We then headed to Taipei, Taiwan. Most of our stay in Taiwan was for another story; however, we did accidentally unturn one stone with something interesting under it.<\/p>\n A B2B company in Taiwan noted to us that it is commissioned by Chinese companies to import servers for pre-testing and pre-assembly and setup steps. When we asked what GPUs the servers ran, we realized that all of the hardware is export-controlled.<\/p>\n Once the B2B agency completes its testing, it ships the system out to the original buyer. Basically, they act as an intermediary to buy the machine, then mark it up and reship it to the original purchaser.<\/p>\n We also spoke with a company that conducts business in Singapore. The company informed us that they are also aware of similar passthrough so-called \u201ctesting\u201d services.<\/p>\n Smuggler<\/p>\n Our last link in this chain is actually the first: The smugglers themselves, and we\u2019re back in the US for this. We didn\u2019t think we\u2019d be able to find one doing the dirty work, but in the final hours producing this story, one of our viewers provided a lead. This led to about a day of production delay as we were wrapping this project up, but as far as we\u2019re aware, this is the first content piece that actually features someone doing the highest risk dirty work.<\/p>\n Everyone else we spoke to is in China. They\u2019re safe from US retaliation. But large-scale smugglers get arrested, fined millions of dollars, and can spend years in prison if they were serious volume movers.<\/p>\n A viewer contacted us to say that he\u2019d connected with a traveling GPU buyer. The story went that this guy drives around buying specifically RTX 4090s. He doesn\u2019t care about anything else, including 5090s, because those don\u2019t sell as well to China. 4090s are ideal, he told us, because they can be modified into cost-effective 48GB models, like at Brother Zhang\u2019s shop.<\/p>\n The smuggler was extremely open with us in text messages written only in Chinese, but wasn\u2019t open to us flying out to him. He did allow us to share footage of his car, however.<\/p>\n The Plug offers $2,000 flat per RTX 4090, which isn\u2019t a bad price. He told us that he then finds ways to get them into China<\/p>\n The Plug, as we call him, has an ATX test bench, motherboard, simple downdraft cooler, power supply, obviously a CPU, RAM, and an SSD, and gigantic battery in the trunk of his Prius. He also has at least one spare license plate in his trunk in addition to the one on the back of his car. We\u2019re not sure why, but it\u2019s better not to ask.<\/p>\n People like The Plug post on Facebook Marketplace and other common online reseller forums looking to buy GPUs, just like anyone else in your city would. The Plug offers $2,000 flat per RTX 4090, which isn\u2019t a bad price. He told us that he then finds ways to get them into China. He recently was scammed out of $5,500 of payments owed by a Hong Kong buyer that resells in Hong Kong, so lately, he\u2019s been considering hand-carrying the devices in luggage instead.<\/p>\n His margin is slim. He makes just under $300 US per GPU, not counting gas, potential hotels as he travels, and time. Our understanding is that some in his shoes will strip the GPU cooler off and ship just the PCB back to China, increasing their margin. Coolers are available in abundance where they\u2019re made, so it can be cheaper to do this than to pay for the weight and size.\u00a0<\/p>\n He sometimes buys entire computers just to take the 4090s out, as that\u2019s what the bounty is on. He then sells the remaining system back to anyone who\u2019ll buy it. He told us that 5090 prices in China are falling so fast due to oversupply, ironically, that he\u2019d lose money reselling it to China and would do better flipping it to Americans.<\/p>\n The Plug isn\u2019t rich. He seems to be doing OK, but it\u2019s not the type of wealth you might expect for such risk.\u00a0<\/p>\n We can\u2019t overstate how important it was to get this piece into the story, as this allowed us to fully complete the chain. We want to provide a huge thanks to our anonymous viewer for their help.\u00a0<\/p>\n Many of the sources we met in Asia during this trip told us they simply didn\u2019t know how the GPUs actually get in, and now we know. People doing this on a small scale like The Plug are unlikely to be caught, but operations transacting millions of dollars worth of GPUs would have a harder time getting them out undetected.<\/p>\n That brings us to the end of our travels, but not the end of this piece. We still have two more entities to look at, and that\u2019s NVIDIA and the US Government.<\/p>\n Details: Legality<\/p>\n The legality is simple — this is our understanding of it.\u00a0<\/p>\n There is no restriction on purchasers, only on sellers the US has control over. Even Americans buying GPUs in China are not violating any laws, as the purchase of a GPU in China is not restricted.<\/p>\n We\u2019ll start with selling:<\/p>\n The simplest answer to \u201cwho the sale of GPUs is restricted for\u201d is anybody who doesn\u2019t have a specific export license<\/a> and who would be governed by US law, and NVIDIA can\u2019t just bypass it by shipping from a different country<\/a>. It has to do with the sale, not with the shipment location. That includes entities buying from US companies, like European companies, who would be under guidance both from NVIDIA and, if they do business in the US, its government.\u00a0<\/p>\n It\u2019s also illegal for anybody <\/a>in the United States, citizen or not, to sell these GPUs to China, Hong Kong, Macau, or companies in those locations if the seller does not have a re-export license.<\/p>\n It is not illegal for a person of any nationality in America to sell a GPU to any non-restricted entity.<\/p>\n As for buying, as Dr. Vinci Chow stated, \u201cThere\u2019s absolutely nothing illegal for us over here to buy these GPUs.\u201d\u00a0<\/p>\n But it\u2019s also not illegal to sell them <\/a>in China for a Chinese company. The Chinese government doesn\u2019t generally enforce American export laws. Other nations cooperating with the United States might, such as the recent arrests<\/a> in Singapore, but once the GPU is in China, the people in possession of it likely don\u2019t care — buyer or seller or someone who does both, like GPU dealer Vincent.<\/p>\n The only control over Chinese companies that the US has is the Entity List<\/a>, which would hurt their business prospects with American companies — but only if they care about that. Chinese GPU middleman \u601d\u9a30\u5408\u529b\u79d1\u6280\u6709\u9650\u516c\u53f8 (Sitonholy Technology Company) in Tianjin landed<\/a> on the Entity List for transacting banned GPUs and being found out. Sitonholy<\/a> purchases GPUs from anyone who can get them into China, and then they bid on domestic projects<\/a> like data center build-outs.<\/p>\n The Entity List was used<\/a> to restrict, for example, DeepCool<\/a> previously, resulting in their American partners ceasing business with them for fear of frozen assets, audits, or collateral bans. This hurts companies like DeepCool that want to operate<\/a> in the US and even shut down their California office, but for Sitonholy, they mostly want to do business <\/a>with other Chinese companies, and so it\u2019d have limited impact unless they wanted to expand to do business with Americans.<\/p>\n Beyond the entity list, the US really has no control<\/a> over what happens inside Chinese borders. That means the only point at which a GPU could feasibly be intercepted<\/a> and a person<\/a> arrested would be operating in the US or in transit to intermediary countries, such as Singapore, which may have their own export laws. The people buying and selling them within China are not breaking any of their own government\u2019s laws, though.<\/p>\n We think the story of AI GPUs has become a story of corruption between governments, and the wealthiest company on Earth.<\/p>\n NVIDIA is Playing All Sides<\/p>\n And so we come back to NVIDIA. At every turn, it really looks like NVIDIA is playing all sides. If there\u2019s enough money to be made, anybody is NVIDIA\u2019s friend.\u00a0<\/p>\n We think the story of AI GPUs has become a story of corruption between governments and the wealthiest company on Earth.<\/p>\n For example, on April 30th, Amazon-backed AI startup Anthropic called on the US government to increase export control restrictions to China. As part of a blog post<\/a>, Anthropic said the government needs to improve its export enforcement to reduce smuggling. The company cited examples of chips being smuggled with \u201cprosthetic baby bumps\u201d and \u201clive lobsters.\u201d This upset NVIDIA, obviously, because NVIDIA doesn\u2019t like restrictions on making money.<\/p>\n So, NVIDIA shot back<\/a>. Seemingly taking a page out of Trump\u2019s playbook, NVIDIA essentially called this fake news and stated, \u201cAmerican firms should focus on innovation and rise to the challenge, rather than tell tall tales that large, heavy, and sensitive electronics are somehow smuggled in \u2018baby bumps\u2019 or \u2018alongside live lobsters.\u2019\u201d<\/p>\n We unearthed this official Hong Kong customs website<\/a> detailing a white van busted driving on the Zhuhai-Macau bridge, filled with 280 kg of undeclared live lobsters and, yes, 70 smuggled GPUs<\/p>\n Gaslighting by NVIDIA<\/p>\n Anthropic isn\u2019t telling tall tales–it\u2019s right. A 2022 video <\/a>previously showed a security check at Zhuhai port in Guangdong, not far from Shenzhen, wherein a woman with a prosthetic baby bump was shown to have been carrying CPUs and iPhones instead. The report made it to customs.gov.cn<\/a>, stating that she arrived from Macau, a common go-between (similar to Hong Kong). This story got international attention in technical media and we reject the possibility that NVIDIA wasn\u2019t aware of it. We found the official Chinese Government posting about smuggling from Macau, so there’s public record. Imports to most of China are taxed, and so tax evasion coupled with smuggling will increase margin on the electronics rather than sharing it with one of the two governments. If you\u2019re already breaking a US law, it seems some just go for a hat-trick and increase the profits.\u00a0<\/p>\n As for the lobsters, that\u2019s real, too: We unearthed this official Hong Kong customs website<\/a> detailing a white van busted driving on the Zhuhai-Macau bridge, filled with 280 kg of undeclared live lobsters and, yes, 70 smuggled GPUs, complete with photo evidence.\u00a0<\/p>\n NVIDIA\u2019s response to call these ‘tall tales’ then is not only defensive, but serves to gaslight and grossly mislead, we think<\/p>\n The Hong Kong customs itself calls this \u201czousi,\u201d or smuggling, and notes a maximum sentence of 7 years.<\/p>\n NVIDIA\u2019s response to call these \u201ctall tales\u201d then is not only defensive, but serves to gaslight and grossly mislead, we think, and is tantamount to lying for sake of downplaying reality for its own benefit. But then this is a common NVIDIA tactic, including its dishonest approach to reviews<\/a> that we\u2019ve already detailed and its deceptive and we think false advertising of the RTX 5070 as being equivalent to an RTX 4090<\/a>, which is provably and comically false.<\/p>\n Another instance of NVIDIA\u2019s \u201cfake news\u201d defense was following a July 24 publication by the Financial Times<\/a>, reporting that more than $1B worth of NVIDIA\u2019s AI chips had been smuggled to China. In response, NVIDIA, whose blind eye is turning an awful lot lately, downplayed the issue, and stated<\/a>, \u201cTrying to cobble together datacenters from smuggled products is a losing proposition, both technically and economically. Datacenters require service and support, which we provide only to authorized NVIDIA products.\u201d<\/p>\n That sounds like something a company selling support would say. And the first part doesn\u2019t really match the whole \u201cthe more you buy, the more you save<\/a>\u201d assertion. It\u2019s only a winning proposition by all of their prior years of statements. If your only option is a useless insufficient data center or a cobbled-together sufficient data center, then a cobbled-together one is still a winning proposition by comparison. It\u2019s weird for the company to pretend that this isn\u2019t worth doing. It\u2019s worth lots of money.\u00a0<\/p>\n That same day, whitehouse.gov posted an article<\/a> titled, \u201cWide Acclaim for President Trump\u2019s Visionary AI Action Plan.\u201d Near the top of the post, it highlighted a sycophantic quote from Huang, which read, \u201cAmerica\u2019s unique advantage that no country could possibly have is President Trump.\u201d And we could call that sycophantic if it was about any president.\u00a0<\/p>\n On August 5, NVIDIA got another opportunity to talk. The US Department of Justice announced <\/a>it had arrested two people in California for smuggling \u201ctens of millions of dollars\u2019 worth of sensitive microchips used in artificial intelligence (AI) applications\u201d to China. The BBC <\/a>reported that court documents say the shipments included the NVIDIA H100 and RTX 4090. Rather than admit smuggling exists, NVIDIA downplayed the situation and stated<\/a>, \u201cThis case demonstrates that smuggling is a nonstarter.\u201d<\/p>\n Except that anyone who made tens of millions of dollars before getting caught had a pretty good start, and so did their customers. This is a bullshit statement from NVIDIA that seemingly aims to downplay and deflect to reduce lawmaker attention on its monopoly.<\/p>\n NVIDIA also said, \u201cWe primarily sell our products to well-known partners, including OEMs, who help us ensure that all sales comply with U.S. export control rules.\u201d NVIDIA noted that, \u201cany diverted products would have no service, support, or updates.\u201d\u00a0<\/p>\n Again, this is not fully true. Our own sources in this story noted that, although something like an HGX system would be hard to service, a standalone PCIe GPU could be parted-out and covered under a separate warranty, even in China.\u00a0<\/p>\n a government committee called Deepseek a ‘threat to national security’ and said Deepseek had used NVIDIA\u2019s technology<\/p>\n Fears of Deepseek<\/p>\n Meanwhile, the US began scrutinizing NVIDIA\u2019s technology for getting into China, whether or not the company itself was directly involved. In a bipartisan report called \u201cDeepseek Unmasked<\/a>,\u201d a government committee called Deepseek a \u201cthreat to national security\u201d and said Deepseek had used NVIDIA\u2019s technology, \u201cAI model appears to be powered by advanced chips provided by American semiconductor giant NVIDIA and reportedly utilizes tens of thousands of chips that are currently restricted from export to the PRC.\u201d It continued, \u201cNVIDIA designed and manufactured many of these chips to create the most sophisticated possible chip while skirting U.S. export controls. This has allowed these chips to be exported to China as the U.S. government develops stricter restrictions. Since March 2024, it is estimated that NVIDIA has produced over 1 million chips for the Chinese market.\u201d<\/p>\n The government also examined<\/a> NVIDIA\u2019s significant revenue growth in Singapore compared to China based on SEC filings, particularly in years featuring restrictions. The government questioned \u201cwhether PRC customers are arranging for the diversion of sensitive chips that are reportedly sold through Singapore,\u201d since revenue from Singapore had grown from almost nothing since 2021.\u00a0<\/p>\n NVIDIA has defended<\/a> its sales to Singapore by saying that \u201cCustomers use Singapore to centralize invoicing while our products are almost always shipped elsewhere.\u201d According to NVIDIA, shipments destined<\/a> to Singapore were only 2% of the company’s total revenue in 2025. But we also know that Singapore has made numerous<\/a> arrests relating to GPU smuggling, so there appears to be some reason for the concerns, whether or not NVIDIA itself wants to turn a blind eye to it.<\/p>\n Ignorance is Bliss<\/p>\n Despite the allegations, NVIDIA downplayed any smuggling of AI chips.\u00a0<\/p>\n In a video uploaded 2 months ago, Jensen Huang spoke on smuggling, and stated<\/a>, \u201cGovernments understand that diversion is not allowed. And there’s no evidence of any AI chip diversion.\u201d Except that there is evidence of it — not only in this very story, but in readily available reports online for years now.<\/p>\n Huang continued<\/a>, \u201cOur data center GPUs are massive. These are massive systems. The Grace Blackwell system is nearly two tons. And so you’re not going to be putting that in your pocket or your backpack anytime soon. And so these systems are fairly easy to keep track of, but the important thing is that the countries and the companies that we sell to recognize that diversion is not allowed and everybody would like to continue to buy NVIDIA technology. And so, they monitor themselves very carefully and they’re quite careful about that.\u201d<\/p>\n This one is interesting. Huang is right that it\u2019s much harder to smuggle Grace-Blackwell or Hopper HGX-class complete systems. Dr. Vinci Chow\u2019s statements align with this when he said, \u201cIt’s very hard to get like a full HGX system.\u201d But it still happens. At least one of NVIDIA\u2019s GPU and server customers in another country told us that they facilitate intermediary transmission to China and in fact showed us the server racks on-site in their facilities. We weren\u2019t allowed to film them, but we saw them. A separate representative told us that document forgery through third-party countries can also disguise such transshipments.<\/p>\n One middleman told us that an NVIDIA distributor gets parts into China; a downchain factory told us that NVIDIA\u2019s QC rejects sometimes end up repurposed and kept in China, salvaging the GPU and VRAM and scrapping the rest; Dr. Vinci Chow told us that one of his own devices had a defective link on it, contributing to this statement. And when he asked him whether he thinks NVIDIA knows all of this is happening, he replied, \u201cI would be surprised if they don\u2019t, right. I would be surprised. I would be really surprised if they don\u2019t. These are very expensive items. I would imagine you would keep track of everything, right? It\u2019s hard to know what [a person] plans to do with all of these defective parts, but I\u2019ll be very surprised that no one has ever thought of the possibility that, if it\u2019s something so valuable, someone would come up with a use [for] even a defective one.\u201d\u00a0\u00a0\u00a0<\/p>\n NVIDIA\u2019s Hypocrisy<\/p>\n So in one set of statements, NVIDIA said that smuggling doesn\u2019t really happen because the export controls work and keep partners in-line.<\/p>\n But in another statement, Huang called the US export controls a \u201cfailure,\u201d talking out of both sides of his mouth, we think. He spoke<\/a> of competing Chinese GPU brands posing a threat to NVIDIA, \u201cThe local companies are very, very talented and very determined, and the export control gave them the spirit, the energy and the government support to accelerate their development. I think, all in all, the export control was a failure.\u201d<\/p>\n But the stakes, and dollar signs, for NVIDIA had increased. Jensen said NVIDIA\u2019s market share in China had dropped from 95% to 50%, and, in NVIDIA\u2019s May quarterly earnings before the H20 exemption and revenue share, Jensen Huang said<\/a> the company\u2019s data center business in China was done, \u201cHowever, the $50 billion China market is effectively closed to U.S. industry. The H20 export ban ended our Hopper data center business in China.\u201d Or, as Jensen said, the China market is worth one Boeing<\/a>. Boeing is probably not the best example…<\/p>\n Conclusion The October 7, 2022 Biden Administration export controls had a stated goal<\/a> to ‘protect US national security and foreign policy interests’ by implementing new export controls<\/p>\n Let\u2019s look back at the timeline once more between NVIDIA and the US government, highlighting NVIDIA\u2019s relentless appetite for global dominance.<\/p>\n The October 7, 2022 Biden Administration export controls had a stated goal<\/a> to \u201cprotect US national security and foreign policy interests\u201d by implementing new export controls restricting China\u2019s ability to build high-end semiconductors, including for the development of supercomputers. There was an included goal of staving off the potential for China to develop \u201cnuclear weapons and other military technologies<\/a>.\u201d\u00a0<\/p>\n Blocking the H100 and A100 led to NVIDIA creating an export-compliant A800 at about 70% of the speed of an A100 for the Chinese market<\/a>.\u00a0<\/p>\n A year later on October 17, 2023, the US Department of Commerce updated <\/a>its export compliance and restricted NVIDIA\u2019s A800 chip as well along with the newer China-targeted H800.<\/p>\n Weeks later on December 6, 2023, NVIDIA told reporters<\/a> in Singapore that it would be working on another new chip that would comply with the US\u2019 new restrictions.\u00a0<\/p>\n The AI Diffusion Rule On January 13, 2025, which was after Donald Trump won the 2024 presidential election, but a week before he took office, the outgoing Biden administration tightened export controls<\/a> by introducing national chip caps for many countries, except for 18 allies. That\u2019s the AI Diffusion Rule we spoke of earlier and would have gone into effect in May.<\/p>\n NVIDIA, who had remained relatively quiet about the regulations up until this point, criticized the restriction and made an attempt to appeal to the president-elect, and reportedly stated<\/a>, \u201cIt makes no sense for the Biden White House to control everyday datacenter computers and technology that is already in gaming PCs worldwide, disguised as an anti-China move. The extreme \u2018country cap\u2019 policy will affect mainstream computers in countries around the world, doing nothing to promote national security but rather pushing the world to alternative technologies. AI is mainstream computing \u2013 ubiquitous and essential as electricity. This last-minute Biden Administration policy would be a legacy that will be criticized by U.S. industry and the global community.\u201d It seems like NVIDIA tried to set up an appeal to the president-elect, stating, \u201cWe would encourage President Biden to not preempt incoming President Trump by enacting a policy that will only harm the U.S. economy, set America back, and play into the hands of U.S. adversaries.\u201d<\/p>\n We already went over the million dollar dinner and ensuing ban-then-unban of the H20 chip.<\/p>\n NVIDIA Persuades Trump Administration<\/p>\n roughly 1 month after Jensen reportedly spent $1 million to eat dinner at Trump\u2019s Mar-A-Lago estate, the US Department of Commerce confirmed<\/a> that it will not implement the AI Diffusion Rule that NVIDIA campaigned against<\/p>\n Later that month on April 30, Huang said this of Trump, stating<\/a>, \u201cWithout the president’s leadership, his policies, his support, and very importantly, his strong encouragement[\u2026] frankly, manufacturing in the United States wouldn’t have accelerated to this pace.\u201d<\/p>\n On May 7, roughly 1 month after Jensen reportedly spent $1 million to eat dinner at Trump\u2019s Mar-A-Lago estate, the US Department of Commerce confirmed<\/a> that it will not implement the AI Diffusion Rule that NVIDIA campaigned against and that was created under the Biden administration. The rule was supposed to go into effect a week later on May 15. This does not unban GPUs like the H100, 5090, B100, and so on.\u00a0<\/p>\n Following the termination of what was supposed to be a rule to address national security implications, the Department of Commerce, now under President Trump, stated<\/a>, \u201cThe Biden AI rule is overly complex, overly bureaucratic, and would stymie American innovation. We will be replacing it with a much simpler rule that unleashes American innovation and ensures American AI dominance.\u201d\u00a0<\/p>\n NVIDIA predictably celebrated the statement<\/a>. The company, which has been begging to sell to China while also praising Taiwan\u2019s importance, now took an America-first<\/a> posture, collecting countries like Pokemon, writing, \u201cWe welcome the Administration\u2019s leadership and new direction on AI policy. With the AI Diffusion Rule revoked, America will have a once-in-a-generation opportunity to lead the next industrial revolution and create high-paying U.S. jobs, build new U.S.-supplied infrastructure, and alleviate the trade deficit.\u201d Job creation promises coming from this company, in particular, are interesting<\/a>, but they are playing all sides consistently.<\/p>\n Later that month on May 28, Huang spoke with Mad Money host Jim Cramer. He stated<\/a>, \u201cWhen [Trump] rescinded the AI Diffusion Rule, it was a visionary move. It was a bold move, and he recognizes that there\u2019s an AI race and we\u2019re not alone. And he wants America to win.\u201d\u00a0<\/p>\n A Manipulative NVIDIA<\/p>\n June 23, via Reuters<\/a>, an official of the US State Department, which didn\u2019t reply to GamersNexus except the Department did send us 3 out-of-office auto responders, warned of DeepSeek military and intelligence operations and warned of the use of \u201cshell companies\u201d in Southeast Asia to circumvent export restrictions. The report mentioned that DeepSeek had \u201clarge volumes\u201d of high-end H100 chips, which are banned in China.\u00a0<\/p>\n NVIDIA didn\u2019t like that, responded to Reuters, and stated, \u201cWe do not support parties that have violated U.S. export controls or are on the U.S. entity lists,\u201d adding, \u201cWith the current export controls, we are effectively out of the China data center market, which is now served only by competitors such as Huawei.\u201d Turning a blind eye to the situation. NVIDIA added, \u201cOur review indicates that DeepSeek used lawfully acquired H800 products, not H100.\u201d<\/p>\n 3 days later on June 26, The Information<\/a> reported that DeepSeek\u2019s next AI model has been delayed due to a shortage of NVIDIA AI GPUs in China. This directly contradicts Huang\u2019s comments that export controls do not work.<\/p>\n On July 4, Bloomberg reported<\/a> that the Department of Commerce, which also did not reply to GamersNexus\u2019 emails, was preparing new export controls on Malaysia and Thailand to reduce chip smuggling; interestingly, Singapore, which now comprises a significant portion of NVIDIA\u2019s revenue, was not on that list despite being a known smuggling passthrough.<\/p>\n [Jensen Huang] really pulled off something few tech CEOs have managed. He played both Washington and Beijing and he won<\/p>\n – CNBC Business News anchor Deidra Bosa<\/p>\n On July 10, Bloomberg reported<\/a> that Huang and Trump were scheduled to meet again ahead of the CEO\u2019s planned trip to China. Days later on July 14, NVIDIA confirmed that it will resume sales of H20 chips to China<\/a> with Huang stating, \u201cThe U.S. government has assured NVIDIA that licenses will be granted, and NVIDIA hopes to start deliveries soon.\u201d\u00a0<\/p>\n On July 15, CNBC Business News anchor Deidra Bosa gave her synopsis on the situation and said<\/a>, \u201c[Jensen Huang] really pulled off something few tech CEOs have managed. He played both Washington and Beijing and he won.\u201d She added, \u201cJensen has stayed disciplined and diplomatic with a clear message, and that is: NVIDIA\u2019s dominance serves America\u2019s interest.\u201d<\/p>\n We write to express our deep concern over the recent decision to resume exports of NVIDIA\u2019s H20 chips to China. As policymakers and professionals with a background in national security policy, we believe this move represents a strategic misstep that endangers the United States\u2019 economic and military edge in artificial intelligence.<\/p>\n – national security experts to US Secretary of Commerce Howard Lutnick<\/p>\n On July 28, The Financial Times<\/a> reported that the US Commerce Department was not going to make \u201ctough moves\u201d to tighten export controls to China. This is in spite of several congressional members warning the administration not to loosen the US\u2019 export controls<\/a> for AI GPUs. Several national security experts also voiced their concern by sending a letter <\/a>to the US Commerce Department, which read, \u201cWe write to express our deep concern over the recent decision to resume exports of NVIDIA\u2019s H20 chips to China. As policymakers and professionals with a background in national security policy, we believe this move represents a strategic misstep that endangers the United States\u2019 economic and military edge in artificial intelligence.\u201d<\/p>\n NVIDIA might have started off as a much more humble company, but it has become a savvy political player<\/a> in a game that\u2019s seemingly pay-to-win<\/p>\n On August 11, via Bloomberg, Trump said<\/a> he was open to allowing NVIDIA to sell modified versions of the company\u2019s newest Blackwell chips to China.<\/p>\n That brings us to today.\u00a0<\/p>\n There\u2019s no one better equipped to play that game than the most valuable company by market cap<\/a> in the world<\/p>\n Most Valuable Company by Market Cap in the World<\/p>\n NVIDIA might have started off as a much more humble company, but it has become a savvy political player<\/a> in a game that\u2019s seemingly pay-to-win. That seems only fitting for a gaming company to be particularly good at pay-to-win games. There\u2019s no one better equipped to play that game than the most valuable company by market cap<\/a> in the world, now at $4 trillion<\/a>, led by a man whose net worth is estimated at $148.1 billion<\/a>. NVIDIA knows when to bite its tongue and how to effectively appeal to the ego of politicians of all parties and all countries.<\/p>\n We think NVIDIA is playing all sides. We think it is greedy, manipulative, and carefully employs propaganda such as its use of the \u201cfake news\u201d playbook for news which is literally reality. But we don\u2019t think NVIDIA has a particular set of beliefs beyond just making more money. We think NVIDIA will sell anyone out to make a buck.<\/p>\n NVIDIA is in the big leagues now. Inside of one month, reportedly paying $1 million to a sitting President after which followed the unlock of $5.5B of lost H20 revenue, followed next by a 15% split of that unlock going to the US Government, is what raises these new questions of NVIDIA’s integrity in our piece.\u00a0<\/p>\n As for the actual black market side of it and smuggling, it was an exciting story to cover and get to the bottom of. We’ve learned that common methods include factory so-called QC defects, hand-carried items by students, actual smugglers on the ground in the US, and suppliers through third-party countries, among others.![\"\"](\"https:\/\/www.newsbeep.com\/il\/wp-content\/uploads\/2025\/12\/image108.jpg.webp.webp\")
<\/p>\n![\"\"](\"https:\/\/www.newsbeep.com\/il\/wp-content\/uploads\/2025\/12\/ai-gpu-black-market_MAIN_FINAL-MP4.mp4_snapshot_00.04.51.859.jpg.webp.webp\")
<\/p>\n<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
![\"\"](\"https:\/\/www.newsbeep.com\/il\/wp-content\/uploads\/2025\/12\/image133.jpg.webp.webp\")
<\/p>\n<\/p>\n
![\"\"](\"https:\/\/www.newsbeep.com\/il\/wp-content\/uploads\/2025\/12\/toolkit-tear-down.jpg.webp.webp\")
Grab a GN Tear-Down Toolkit<\/a> to support our AD-FREE reviews and IN-DEPTH testing while also getting a high-quality, highly portable 10-piece toolkit<\/a> that was custom designed for use with video cards for repasting and water block installation. Includes a portable roll bag, hook hangers for pegboards, a storage compartment, and instructional GPU disassembly cards.The Smuggling Pipeline<\/p>\n<\/p>\n
<\/p>\n
<\/p>\n
![\"\"](\"https:\/\/www.newsbeep.com\/il\/wp-content\/uploads\/2025\/12\/image138.jpg.webp.webp\")
<\/p>\n<\/p>\n
<\/p>\n
![\"\"](\"https:\/\/www.newsbeep.com\/il\/wp-content\/uploads\/2025\/12\/image141.jpg.webp.webp\")
<\/p>\n![\"\"](\"https:\/\/www.newsbeep.com\/il\/wp-content\/uploads\/2025\/12\/image142.jpg.webp.webp\")
HONG KONG: Dr. Vinci Chow, Chinese University of Hong Kong<\/p>\n<\/p>\n
<\/p>\n
![\"\"](\"https:\/\/www.newsbeep.com\/il\/wp-content\/uploads\/2025\/12\/image145.jpg.webp.webp\")
<\/p>\n<\/p>\n
<\/p>\n
Datacenter<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
![\"\"](\"https:\/\/www.newsbeep.com\/il\/wp-content\/uploads\/2025\/12\/image012.jpg.webp.webp\")
<\/p>\n![\"\"](\"https:\/\/www.newsbeep.com\/il\/wp-content\/uploads\/2025\/12\/image014_0.jpg.webp.webp\")
![\"\"](\"https:\/\/www.newsbeep.com\/il\/wp-content\/uploads\/2025\/12\/image015.jpg.webp.webp\")
<\/p>\n![\"\"](\"https:\/\/www.newsbeep.com\/il\/wp-content\/uploads\/2025\/12\/image016.jpg.webp.webp\")
![\"\"](\"https:\/\/www.newsbeep.com\/il\/wp-content\/uploads\/2025\/12\/image017.jpg.webp.webp\")
<\/p>\n<\/p>\n
![\"\"](\"https:\/\/www.newsbeep.com\/il\/wp-content\/uploads\/2025\/12\/vlcsnap-2025-09-11-13h00m03s311.jpg.webp.webp\")
<\/p>\n<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
![\"\"](\"https:\/\/www.newsbeep.com\/il\/wp-content\/uploads\/2025\/10\/gn-modmat-1000x489_1.jpg.webp\")
<\/a>Grab a GN15 Large Anti-Static Modmat<\/a> to celebrate our 15th Anniversary and for a high-quality PC building work surface. The Modmat features useful PC building diagrams and is anti-static conductive. Purchases directly fund our work! (or consider a direct donation<\/a> or a Patreon contribution<\/a>!)Warehouses<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
![\"\"](\"https:\/\/www.newsbeep.com\/il\/wp-content\/uploads\/2025\/12\/image056.jpg.webp.webp\")
<\/p>\n<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
Tech Exec Sycophancy<\/p>\n
Singapore Suspicions<\/p>\n
<\/p>\n
<\/a>Grab a GN15 Large Anti-Static Modmat<\/a> to celebrate our 15th Anniversary and for a high-quality PC building work surface. The Modmat features useful PC building diagrams and is anti-static conductive. Purchases directly fund our work! (or consider a direct donation<\/a> or a Patreon contribution<\/a>!)<\/p>\n<\/p>\n
<\/p>\n
<\/p>\n
We loved working on this story and meeting all of these unique people. Each person played a key role in helping us find the next person, allowing us to complete the first public, on-record, complete start-to-finish cataloguing of a smuggling pipeline for high-end silicon. We want to thank everyone who made this story possible, including our viewers who funded it<\/a>.<\/p>\n","protected":false},"excerpt":{"rendered":"Acquiring “Illegal” GPUs in China In China, it’s not an illegal market — it’s just a market. We…\n","protected":false},"author":2,"featured_media":184099,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[20],"tags":[345,343,344,85,46,125],"class_list":{"0":"post-184098","1":"post","2":"type-post","3":"status-publish","4":"format-standard","5":"has-post-thumbnail","7":"category-artificial-intelligence","8":"tag-ai","9":"tag-artificial-intelligence","10":"tag-artificialintelligence","11":"tag-il","12":"tag-israel","13":"tag-technology"},"_links":{"self":[{"href":"https:\/\/www.newsbeep.com\/il\/wp-json\/wp\/v2\/posts\/184098","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.newsbeep.com\/il\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.newsbeep.com\/il\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/il\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/il\/wp-json\/wp\/v2\/comments?post=184098"}],"version-history":[{"count":0,"href":"https:\/\/www.newsbeep.com\/il\/wp-json\/wp\/v2\/posts\/184098\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/il\/wp-json\/wp\/v2\/media\/184099"}],"wp:attachment":[{"href":"https:\/\/www.newsbeep.com\/il\/wp-json\/wp\/v2\/media?parent=184098"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.newsbeep.com\/il\/wp-json\/wp\/v2\/categories?post=184098"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.newsbeep.com\/il\/wp-json\/wp\/v2\/tags?post=184098"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}