The $100T Joke: Chinese Tech Company Challenged Apple, Nvidia & Tesla—Collapsed Within A Year

A video showing a supplier allegedly demanding payment from Dream Technology circulated online in early September. In the footage, a man identified as a supplier stands on an office desk and shouts that the company has delayed payments for a year, while employees' family members are hospitalized in the ICU and urgently need money. On September 5, a group of workers gathered outside the entrance of Dream Technology's industrial park in Suzhou, Jiangsu Province. They chanted, The reason the company has attracted attention is that just one year earlier, it had officially announced plans to build the fastest car in the world, and even publicly challenged Apple, NVIDIA, and Tesla. Founded in 2017, Dream Tech initially focused on robot vacuum cleaners and vacuum products. In recent years, however, the company has rapidly expanded into multiple industries, moving from home appliances into EVs, smartphones, large AI models, microchips, smart rings, and is even reportedly exploring areas such as drones, asteroid exploration, and space mining. On August 28, 2025, Dream Tech officially announced its entry into the automotive industry and launched the Starry Sky project. It declared it would be the world's fastest car. Founder Yu Hao wrote in an open letter, At the time, Dream even unveiled a luxury electric supercar positioned as a rival to Bugatti. But this highly publicized car-making expedition took a sharp turn less than a year later. In January, Dream founder Yu Hao posted a video showing the company's new car, the Nebula NEXT-01. Later, the Starry Sky project unveiled a total of three Nebula vehicles, and claimed that its team had grown to nearly 1,000 employees at one point. However, employee Li Da told Chinese media outlet Yicai that the so-called concept cars were actually models that could not operate normally. The company reportedly customized 10 model cars in different colors, with each costing about 3 million yuan. The Starry Sky project's office was located in Building 6 of an industrial park in Wuzhong District, Suzhou. Over the past four months, reporters cross-checked information from multiple sources, including business records, site visits, observations of the office, and statements from several insiders. They found that senior executives from Dream's automotive division had been actively seeking new jobs, while project employees had experienced mass resignations and layoffs. Many desks were empty, personal items had been mostly removed, and the project's original signage had been taken down. Last year, a site for the project in eastern China was publicly announced, covering an area of 414,400 square meters. But during an on-site visit in May this year, reporters found that the grass there had grown taller than a person. Meanwhile, public information showed that Dream's announced overseas factory plans, first-round funding, and manufacturing partnerships had not seen substantial progress. Within the Dream system, there were three business units focused on building sports cars. Besides the Starry Sky project, there was also the Starlight Future and Interstellar Voyage, but all faced major setbacks. During its push into car manufacturing, Dream also incubated two battery companies, Lightyear Engine, focused on consumer applications, and Jinghe Energies, specializing in all-solid-state batteries for vehicles. Currently, Lightyear Engine is also carrying out large-scale layoffs. According to a report by National Business Daily, in response to rumors that its car project had been terminated, people familiar said the company had abandoned plans for mass production and would only retain related technology reserves. The most astonishing moment came on January 12, 2026, when CEO and founder Yu Hao posted a shocking statement on his social media. He claimed that Dream would become the first $100 trillion company ecosystem in human history. Currently, the world's highest-valued company is NVIDIA, with a market cap of $4.5 trillion. Yu's claim is even higher than the nearly $70 trillion, representing the combined market value of the entire U.S. stock market. Yu Hao said, Entrepreneurs represented by Elon Musk and Jensen Huang can raise the market value of a single company to between $8 to $10 trillion. But they are older than me by a generation. As a younger person, his goal was to increase enterprise value by another order of magnitude and create a $100 trillion ecosystem. In response, many users commented, he must have been drunk. Another said, This company has no real technological substance and is full of bubbles. Its collapse was bound to happen overnight. Why did Dreamtech, after making such bold claims, fall from its peak to its current situation? Industry insiders believe the main reasons include the following factors. First, Dreamtech entered the wrong market at the wrong time. After aggressively moving into the new energy sector, it happened to run into the sharp decline of government subsidies. The plan to use subsidies to make quick profits failed, and instead, the company lost significant marketing investments. During the final week of August, Chinese media described several EV startups as they released their first-half financial results. Data released by Li Auto on August 26 showed that it delivered 98,330 vehicles in the second quarter, down 11.5% year-on-year. Revenue reached 25.7 billion yuan, while net losses totaled 1.7 billion yuan. After losing 2.3 billion yuan in the first quarter, the company's total losses in the first half reached about 4 billion yuan. Its cash reserves fell from 94.3 billion yuan at the end of the first quarter to 87.5 billion yuan at the end of the second quarter, a drop of 6.8 million yuan in three months. Data released by Leap Motor on August 24 showed that the company delivered 356,000 vehicles in the first half of the year, up 60% year-over-year, ranking first in sales. However, its net profit was only 210 million yuan. Calculated per vehicle, the profit was less than 600 yuan per car. Leap Motor has partnered with Stellantis Group, using its global distribution network to enter 45 international markets. Sales channels and after-sales services are mainly managed by Stellantis. As the technology provider and vehicle supplier, Leap Motor has to give up part of its profit margin in pricing negotiations. Financial reports showed that exports rose from 0 to 27% of sales, while overall gross margin fell from 14.1% to 11.7%. Industry Insider said that if exports exceed 30% of sales and gross margin drops below 10%, Leap Motor's profitability could enter a danger zone. Data released by XPeng Motors showed that it delivered 166,000 vehicles in the first half of the year, down 15.8% year-over-year, with a net loss of 3.12 billion yuan. Its vehicle sales gross margin was 12.1%, far from enough to cover research and development, as well as distribution costs. Financial reports showed that the company's overall gross margin was supported by its sales and other businesses sector, including EV platforms, software, and electronic and electrical architectures. Data from Xiaomi's automotive business was disclosed along Xiaomi Group's financial report. In the second quarter, Xiaomi delivered 104,000 vehicles, up 28.2% year-over-year, while its automotive and AI businesses recorded an operating loss of 2.6 billion yuan. The company's financial report showed that Xiaomi's car business relies on funding and other forms of support from the parent company. On August 28, William Li, chairman and CEO of NIO, wrote on Weibo that the retail penetration rate of EV passenger vehicles continued to rise, reaching 65% in July. He said China currently has 371 million vehicles on the road with gasoline-powered cars, accounting for 86.8%, and having an average age of 8.2 years. At the same time, the first generation of EVs is approaching the end of their warranty periods. Li's comments were interpreted as meaning that once EVs are out of warranty, owners may have to pay for battery replacements. Replacing a battery can often cost tens of thousands or even more than 100,000 yuan, and the cost could exceed the remaining value of the EV itself. Earlier, Li predicted the eventual outcome of China's auto industry and said the industry's final round would be determined within three to five years. He said, we are about to enter the most brutal stage of the final competition. Over the next three to five years, the last players in the auto industry will basically be decided. Deputy Secretary General of the China Association of Automobile Manufacturers Chen Shihua also said that China's auto industry is currently facing operational challenges, with several profit levels in the industry at a historic low. Industry Insider said the companies mentioned above are among the survivors of intense competition. Yet even these leading companies have suffered heavy losses this year as subsidies declined. For Dreamtec, which entered the market with overly ambitious plans, unless it had unparalleled proprietary technology, trying to rely on marketing or subsidies alone would only lead to failure and humiliation. However, Dreamtec has become another failed entrant into the EV sector. Its core business has come under pressure. On July 29, the U.S. Federal Communications Commission, FCC, officially announced that it would add foreign-made advanced robotic devices and connected inverters to its coverage list. After the announcement, public attention immediately focused on humanoid robots, as they are currently one of Silicon Valley's hottest fields. Boston Dynamics quadruped robots and Figures humanoid robots are attracting widespread attention. However, the industry soon discovered that the product category most affected by the restriction was one that few people expected. Robot vacuum cleaners. Katie Gorshak, FCC Director of Media Relations, confirmed that robot vacuums are indeed covered by the restriction. According to FCC standards, a device is considered an advanced robotic device if it meets all of the following conditions. It is an autonomous robot controlled by software, can sense its surrounding environment, weighs more than 4.4 pounds, including the charging station, moves on the ground, and has wireless connectivity. Robot vacuum cleaners matched every requirement perfectly. They use software to plan routes, rely on LiDAR or cameras to sense their surroundings, the main unit and charging base easily exceed 4.4 pounds, move independently across home floors, and require Wi-Fi connections to be controlled through apps. Not only that, under this definition, robotic lawnmowers, sidewalk delivery robots, and even automated grass-cutting devices could theoretically fall within the scope of this restriction. The FCC has also listed some exempt categories, including self-driving cars and trains, drones and underwater vehicles, which are already covered by separate restrictions, as well as surgical robots, electric wheelchairs, and stationary robotic arms used in industrial manufacturing. In 2025, global shipments of household cleaning robots reached 32.7 million units, up more than 20% year-over-year. The top five brands together accounted for more than 73% of the market, and all of them were Chinese companies. Dreamtech ranked third on the list. This explains why Dreamtech took such a large risk by entering the EV market, only to suffer a major failure. The reason was that the company had already learned that its core business, robot vacuum cleaners, could face a complete ban from the U.S. market. So what is the current situation for Dreamtech's other businesses? In March 2026, Dreamtechnology's ecosystem company, Chip Odyssey, made a high-profile announcement of an autonomous driving chip and large model processor that it claimed was built on a two nanometer process and capable of 2,000 tops of computing power. However, because self-developed chips require heavy investment, long development cycles, and an unproven path to commercialization, the project has now been removed from its independent organizational structure. In the smartphone sector, Dream launched an ultra-expensive mass-produced model in 2026, but later reports from workplace platforms and media outlets revealed that its smartphone business had been completely shut down, with many related research and development staff leaving the company. Currently, chip and smartphone development teams have been merged into a newly established industrial research institute and are no longer being operated as independent business units. Why did Dream's expansion into large AI models and self-developed chips fail? Experts say it is linked to the Chinese authorities' misguided policies. In recent years, under the Chinese Communist Party's push for chip independence, local governments and private companies have been burdened with political missions. They have rushed into the semiconductor industry regardless of cost, creating a false boom resembling the Great Leap Forward. However, this state-driven approach of flooding the market with resources goes against the objective realities and scientific principles of the semiconductor industry. It not only fails to produce advanced chips, but it could also push China toward greater technological isolation. Experts said government subsidies and forced industrial policies have created rapid year-over-year sales growth, but they cannot hide the reality that these companies lack genuine international competitiveness and technological advantages. In the end, they will leave behind nothing but a mess. A Reuters article published on September 7 reported that Suiyuan Technology is advancing plans for a listing in Shanghai, while MoreThreads, MetaX, Integrated Circuits, and Biren Technology have already entered capital markets. Chinese AI chip companies have recently raised large amounts of funding. These four companies are working with firms such as Huawei to build computing platforms that can replace NVIDIA products. The article cited data showing that China's AI chip market is worth about $90 billion, while NVIDIA's share of the Chinese market has fallen from near-monopoly levels to about 55%. Among them, Suiyuan Technology, which has received investment from Tencent, is preparing an initial public offering in Shanghai and plans to raise about $900 million. The company has not yet achieved profitability. In the first quarter of this year, its sales increased more than 14 times year-over-year, with its business focused on AI inference chips and related computing systems. Companies such as MoreThreads are also developing products that can support programs written for NVIDIA's CUDA platform. CUDA includes programming tools, software libraries, and development environments, and has long been a key competitive barrier for NVIDIA in the AI chip market. Chinese companies hope to use code conversion tools to move programs originally developed with CUDA onto domestic platforms, attracting data centers, internet platforms, and AI companies to adopt Chinese-made chips. However, Mr. Jiang, who works in AI development in Jiangsu province, told reporters that his company began using Chinese-made computing chips in recent years. During testing, they found that many technologies still relied on US, Japan, and other countries. Due to insufficient technical support, testing was temporarily halted. He said, during the initial chip testing, we found compatibility problems. After those were solved, the computing performance was not enough. Now another problem with spending huge amounts of money developing chips is that the cost is higher than buying foreign chips. If large-scale products are more expensive than chips from the US and Japan, who will buy them? The software side also doesn't work. This is not something that can be solved simply by throwing money at it. Mr. Qin, a semiconductor industry researcher in Nanjing, said the CCP is trying to force the creation of a computing platform independent from mainstream technology ecosystems, such as NVIDIA's CUDA, and the global open-source community. He said that while it is presented as ecosystem competition, it is essentially creating a digital iron curtain in the technology sector, forcibly separating Chinese researchers and private companies from advanced global technology ecosystems. Qin said, as frontline engineers have pointed out, abandoning globally used development ecosystems means extremely high conversion costs, risks of rebuilding entire systems, and the cost of retraining personnel. To protect its political power and compete against the free world, the CCP is willing to hold the entire industry hostage, forcing domestic companies to pay for this inefficient and outdated backup ecosystem. In January, the U.S. Department of Commerce's Bureau of Industry and Security adjusted export licensing review policies for advanced chips including NVIDIA's H200, AMD's Mi325X, and similar products shipped to China. Chinese buyers are required to establish export compliance procedures, including customer screening, and the products must undergo third-party performance and security testing in the United States. On May 31, the agency issued enforcement guidance reiterating that exports of advanced computing products to companies whose headquarters or ultimate parent companies are located in countries under U.S. arms embargoes require a license even if the receiving company is located in a third country. These countries are called D5 countries and include China, Russia, Iran, and Macau. Mr. Gu, a semiconductor industry insider in Zhejiang province, said that as the United States, the Netherlands, and Japan tighten export controls on advanced lithography machines, key semiconductor equipment, and critical materials, the CCP's push to He said it's like the steelmaking campaign during the 1958 Great Leap Forward when people melted down household items to produce steel. The same scenario is now playing out in the chip industry. Before long, the West will cut off China's access to related technologies, including software and supporting materials. Experts said that if even the four nationally prioritized chip companies are struggling, then Dream, which is not a leading player, will naturally find it harder to secure resources in the field. Being abandoned by the market was therefore expected. Dream went from a company surrounded by hype to one facing debt collection demands. What people are seeing now is simply the final scene. And companies like Dream are not rare in China. Their eventual outcomes may be very similar. A robot company that loses its core business fails to transform successfully and ultimately becoming one of the failed companies swept away by the market.