828cloud

Data, Info and News of Life and Economy

Category Archives: Technology

China’s Flying Cars Ready for Liftoff with EV Technology

SHIZUKA TANABE wrote . . . . . . . . .

XPeng, EHang and other Chinese companies are commercializing flying cars this year, tapping the country’s advantages in electric car technologies to claim a large share of the emerging global market.

XPeng AeroHT, a subsidiary of the electric vehicle startup, aims to sell a dual-mode electric vertical take-off and landing (eVTOL) vehicle, which can drive on land like a car and detach a flying module for air travel.

“Normal eVTOL vehicles cannot drive on the ground, but our model is dual use,” said Qiu Mingquan, vice president at XPeng AeroHT.

The Civil Aviation Administration of China has begun its type certification review for the aircraft, a requirement for commercial operation.

XPeng AeroHT will begin accepting pre-orders in China as early as October, with plans to begin mass production as soon as next year. Demand is expected from tourism-related companies and outdoor enthusiasts.

The aircraft will be priced in the 1 million yuan ($138,000) range. Qiu said the company hopes to eventually bring the price down into the hundreds of thousands of yuan.

“If large-scale mass production becomes possible, we can dramatically reduce costs” for materials like carbon fiber, he said.

The company plans to expand overseas. “The Middle East is an important market for us, given the level of regulation, openness to new things and cost,” Qiu said.

The company also aims to commercialize an integrated dual-use eVTOL aircraft that does not require separation, in which the flight propeller is folded and stored on top of the vehicle when driving. A concept model was shown at a Las Vegas technology trade show in January.

EHang obtained type certification for its EH216-S model in October. The two-seater can fly for 25 minutes on a single charge. It went on sale on April 1 for 2.39 million yuan in China and $410,000 elsewhere.

Last month, the startup received authorization to mass produce the aircraft. It hopes to partner with hotels and other businesses to commercialize tourism services.

“We have received inquiries from the Middle East and Southeast Asia,” said Vice President He Tianxing.

China accounts for 50% of the world’s total eVTOL models, according to a China Merchants Securities report last month, far above the U.S.’s 18% and Germany’s 8%.

Behind the rapid progress of China’s flying car sector is the accumulation of EV-related technologies, including batteries.

Batteries for eVTOL vehicles are said to require a high energy density of over 400 watt-hours per kilogram of weight. Chinese battery makers like CATL, the global leader in vehicle batteries, are developing products for eVTOL aircraft, giving Chinese manufacturers an advantage in procurement.

Many technologies can be shared between the automotive and eVTOL businesses. CITIC Securities said that “weight-reducing automobile manufacturing technology is supporting the realization of flying cars.”

Major state-owned automaker Guangzhou Automobile Group is developing the GOVE eVTOL vehicle, in which the aircraft section separates from the vehicle’s chassis, with test flights to begin next year.

Geely subsidiary Aerofugia makes an eVTOL aircraft that can seat six people and fly long distances.

Many Chinese cities contend with serious traffic jams, raising expectations for eVTOL vehicles to offer alternative means of transportation and logistics.

The eVTOL, drone and helicopter businesses in China are collectively known as the low-altitude economy. The Chinese Communist Party said at a conference in December that it would work to foster the low-altitude economy this year, along with the biotech and space industries.

Local governments are also providing support. Shenzhen began measures in December to support companies in the field to boost capital and expand production. Guangzhou has also announced plans to set up an area for field tests for related businesses.

However, hurdles to widespread use remain. There are still few takeoff and landing points, and traffic rules for individual drivers are not yet in place.

“We will be forced to fly relatively infrequently for the next few years,” said an eVTOL company executive.

Generating profit is also not easy. EHang incurred a net loss of 300 million yuan last year. Companies are not yet able to realize cost reductions from mass production.

In November, U.S. short seller Hindenburg Research questioned the validity of EHang’s type certification and preorder figures, accusations that the company denied.

To capture a wide range of demand, makers of eVTOL aircraft are eyeing overseas markets.

The global eVTOL market is expected to reach $23.4 billion in 2030, about 19 times the size last year, according to research companies Global Information and MarketsandMarkets. In addition to China, growth is expected in North America, Europe and the Middle East.

To carry out commercial operations overseas, type certifications are necessary from each country. U.S. startup Joby Aviation has already applied for certification in the U.S., the U.K. and Japan.


Source : Nikkei Asia

Chart: What Are the Biggest Perceived Dangers of AI?

Source : Statista

Meet AI001, Alibaba’s First AI Employee

He Qitong wrote . . . . . . . . .

Complete with an official employee ID (AI001) and a Myers-Briggs Type Indicator personality known as “the Architect,” Chinese tech giant Alibaba Cloud on Tuesday inducted its coding assistant, Tongyi Lingma, as the company’s first AI employee.

The AI-powered assistant, which was also given a lanyard featuring a profile photo, work license, and resume, is accessible to all employees as a plugin. The company announced that it is mainly designed to assist with most programming tasks, such as writing code, optimization, debugging, and testing.

The tech firm projects that Tongyi Lingma will write at least 20% of the company’s code in the future, potentially saving programmers hours of work otherwise spent on test code. An Alibaba member of staff stated: “But programmers are still the core of research and development. The AI will allow them more time to focus on system design and core business development work.”

According to Alibaba, Tongyi Lingma is proficient in more than 200 programming languages, and excels in 16. Official testing showed that it is particularly capable in cross-file perception and adaptability to Alibaba Cloud’s work environment.

Launched in October 2023, the AI is built on the Tongyi large language model by Alibaba Cloud and Tongyi Lab and has already surpassed 2 million downloads.

Infoq, an online programming community, showed that while GitHub Copilot leads in usage among AI programming assistants in China, with a 64.5% rate, Tongyi Lingma stands out as the sole product with a usage rate exceeding 10% in the country, capturing 12.9% of the market.

However, Tongyi Lingma isn’t China’s first AI-powered coding assistant.

Baidu Comate, developed by search giant Baidu from its Ernie large language model and launched internally in April 2023, has already been adopted by over 80% of the firm’s engineers.

Other entries like Racoon from SenseTime, known for its Chinese language capabilities, and Fitten Code by FittenTech, launched in 2023 and 2024, respectively, are both designed to translate natural language into code in a bid to ease the load for programmers while boosting productivity.

Following Alibaba’s announcement, many on social media underscored that such AI tools could help alleviate the heavy workloads on developers. According to a 2023 report by the Chinese Software Developer Network (CSDN), over 60% of Chinese programmers work beyond the standard 40-hour week.

Yet, the rapid adoption of AI assistants has sparked debates about the future of programming jobs. Baidu CEO Robin Li, in a March interview with state broadcaster CCTV, suggested that programmers might even become obsolete as AI capabilities expand. He said: “In the future, there will be no such profession as ‘programmer,’ because as long as you can speak, everyone will have the ability to be a programmer.”

In March 2023, Goldman Sachs reported that generative AI — which can generate text, images, videos, or other data using generative models — could replace 300 million jobs globally, while He Jifeng, director of the Shanghai Academy of Artificial Intelligence Industrial Technology, suggested that up to 50% of professions could be phased out by AI between 2030 and 2060.

Zhang Liaoyuan, the lead for Tongyi Lingma products, however, emphasized in an interview with domestic media that AI’s role will be limited to supporting developers with daily coding tasks and navigating technical challenges.

Reiterating the central role of humans in the development process, he stated: “In the development process, humans are always the main body, and some work cannot be replaced by AI.”


Source : Sixth Tone

Scientists Unveil Breakthrough in Fusion Reactor Technology

Brian Westenhaus wrote . . . . . . . . .

The research, which is featured in a new paper in Nuclear Fusion, includes observations, numerical simulations and analysis from their experiments inside a fusion plasma vessel called the Lithium Tokamak Experiment-Beta (LTX-β).

One team at the U.S. Department of Energy’s Princeton Plasma Physics Laboratory (PPPL) has been asking themselves lately as a metaphor, “How much fuel can we add to the fire while still maintaining control?”

Now, they believe they have the answer for one particular scenario. It’s all a part of the Lab’s work to bring energy from fusion to the power grid.

Building upon recent findings showing the promise of coating the inner surface of the vessel containing a fusion plasma in liquid lithium, the researchers have determined the maximum density of uncharged, or neutral, particles at the edge of a plasma before the edge of the plasma cools off and certain instabilities become unpredictable. Knowing the maximum density for neutral particles at the edge of a fusion plasma is important because it gives the researchers a sense of how and how much to fuel the fusion reaction.

The unique environment of LTX-β

The LTX-β is one of many fusion vessels around the world that holds plasma in a donut shape using magnetic fields. Such vessels are known as tokamaks. What makes this tokamak special is that its inner walls can be coated, almost completely, in lithium. This fundamentally changes the wall behavior, as the lithium holds on to a very high percentage of the hydrogen atoms coming off the plasma. Without the lithium, far more hydrogen would bounce off the walls and back into the plasma. In early 2024, the research team reported that this low recycling environment for hydrogen keeps the very edge of the plasma hot, making the plasma more stable and providing room for a larger volume of plasma.

Richard Majeski, a managing principal research physicist at PPPL and head of LTX-β said, “We are trying to show that a lithium wall can enable a smaller fusion reactor, which will translate into a higher power density.” Ultimately, this research could translate into the cost-effective fusion power source the world needs.

With the new paper the LTX-β team has published additional findings showing the relationship between the fuel for the plasma and its stability. Specifically, the researchers found the maximum density of neutral particles at the edge of a plasma inside LTX-β before the edge starts to cool, potentially leading to stability problems. The researchers believe they can reduce the likelihood of certain instabilities by keeping the density at the edge of the plasma below their newly defined level of 1 x 1019 m-3. This is the first time such a level has been established for LTX-β, and knowing it is a big step in their mission to prove lithium is the ideal choice for an inner-wall coating in a tokamak because it guides them toward the best practices for fueling their plasmas.

In LTX-β, the fusion is fueled in two ways: using puffs of hydrogen gas from the edge and a beam of neutral particles. Researchers are refining how to use both methods in tandem to create an optimal plasma that will sustain fusion for a long time in future fusion reactors while generating enough energy to make it practical for the power grid.

Refining methods for retaining an even temperature across the plasma

Physicists often compare the temperature at its edge to its core temperature to assess how easy it will be to manage. They plot these numbers on a graph and consider the slope of the line. If the temperature at the inner core and outer edge are nearly the same, the line is almost flat, so they call that a flat temperature profile. If the temperature at the outer edge is significantly lower than the temperature at the inner core, scientists call it a peaked temperature profile.

Santanu Banerjee, a staff research physicist at PPPL and lead author on the new paper explained, “The team determined the maximum density of neutral particles beyond the edge of a plasma that still allows for a flat-edge temperature profile. Going beyond that number of neutrals at the edge definitely will drop your edge temperature, and you will end up in a peaked temperature profile.”

“That same neutral density is the threshold for instabilities known as tearing modes. Beyond that density, tearing modes tend to get destabilized, cause threats to the plasma and may stop the fusion reaction if left uncontrolled,” he added.

If the instabilities become too large, the fusion reaction will end. In order to support the power grid, researchers are figuring out the best ways to manage a fusion plasma so that the reaction is stable.

Banerjee and Majeski worked with several other researchers on the paper, including PPPL’s Dennis Boyle, Anurag Maan, Nate Ferraro, George Wilkie, Mario Podesta and Ron Bell.

Meanwhile work on the project continues. PPPL engineer Dylan Corl is optimizing the direction in which the neutral beam, which is used to heat the plasma, is injected into the tokamak. “We’re basically creating a new port for it,” Corl said. He uses a 3D model of the LTX-β, testing different beam trajectories to ensure the beam won’t hit another part of the equipment, such as tools used to measure the plasma. “Finding the best angle has been a challenge, but I believe we’ve got it now,” Corl said.

This is a fascinating account of one part of the effort to contain a plasma at temperatures and pressures needed for a fusion reaction. The idea of the tokamak has, for your humble writer, been a something of a bewilderment. The surface area of a donut vs say a simple sphere is well . . . And compressing the shape seems well . . . The tokamak shape seems to amplify the problems. However, your humble writer is no kind of expert in these matters.

Thus everything developed is interesting to the point of fascination. The best case is the tokamak will be a power producer someday. If now, the effort is going to turn up a whole lot of knowledge and knowhow that might enable other concepts or designs.

Fusion on earth is a long, long, long reach. We’re getting there with each grasp of new understanding and knowhow.


Source : Oil Price

Tech War: How Chinese Scientists Rigged a Low-cost AI Computer Chip to Power a Hypersonic Weapon

From SCMP . . . . . . . . .

A research team in China has created a step-by-step guide that allows anyone with a low-cost artificial intelligence chip to boost the performance of hypersonic weapons.

To accomplish the task, the researchers installed a Nvidia Jetson TX2i GPU computer module – which can be purchased online – into an air-breathing hypersonic aircraft capable of speeds exceeding Mach 7.

Tests have suggested that this particular module can process computational fluid dynamics models with unprecedented efficiency, meaning calculations that previously took seconds to complete could now be done in just 25 milliseconds – four times faster than the blink of an eye.
The module’s response speed made it ideal for “real-time optimisation of the fuel supply system, fault diagnosis, and fault-tolerant control in scramjet engines”, according to a joint project team from Beijing Power Machinery Research Institute and Dalian University of Technology. Their peer-reviewed paper was published in the Chinese academic journal Propulsion Technology.

US ‘may be sending strong message’ to China with hypersonic missile test

Headquartered in the United States, Nvidia is the world’s largest supplier of AI chips. Nvidia began selling the TX2i for industrial applications about six years ago. The module’s peak single-precision performance is 1.26 TFlops, about one-fiftieth of the capability of the company’s most powerful AI chip, the H100.

The H100, however, costs tens of thousands of dollars and is in short supply. The TX2i, on the other hand, can be obtained for a few hundred dollars, is not subject to US export controls, and is widely available online.

When contacted by the South China Morning Post on April 12, Nvidia said it had no comment on the matter.

The project team, led by Professor Sun Ximing, said in their paper that the TX2i module in the scramjet engine control system not only boosted the range and stability of hypersonic vehicles, but also significantly reduced their research and development costs.

This was not the first time that Chinese scientists have used US chips in hypersonic weapon research, according to their paper. Previous studies used Intel CPUs and Nvidia’s high-end graphics cards to simulate complex high-speed flow fields.

“High-performance graphics cards possess excellent computational capabilities but require supporting equipment such as a hosting platform, power supply and radiator. They have disadvantages such as high power consumption, heavy weight and large size, which do not meet the demands of lightweight and small-sized embedded controllers in the aerospace field,” Sun’s team wrote in their paper.

Because of the sequential nature of hypersonic flow field simulations, where one event must occur before another can be calculated, industry experts generally believe that such computational tasks cannot be accomplished using lower-end AI chips adept at performing simple parallel computations.

To solve this problem, Sun’s team introduced a novel CPU plus GPU architecture, which detailed how to tackle sequential parallel computing problems by ensuring the two different chip types worked well together.

The “step-by-step” guide provided in the paper provides detailed formulas and addresses potential engineering challenges, including limiting simulation grid size, memory management, code optimisation, and specific compilation instruction schemes.

For better universality, the engine controller’s interface and communication protocol adhere to international standards.

However, the engine is only one component of the entire weapon platform.

“To apply (the AI chip) to hypersonic vehicles, further work is needed in inlet modelling, shock wave correction, and data modelling,” the team said in the paper.

Some important parameters involved in these tasks usually need to be obtained in extensive wind tunnel testing and actual flights.

Still, the likelihood of the TX2i being used for Chinese hypersonic missiles is low. China’s domestic chip manufacturers can provide chips to the country’s military that perform as well as, or better, than the TX2i, with minimal concerns over supply chain reliability and safety.

It remains unclear why the researchers selected the Nvidia chip for their experiment, and the authors could not be reached for comment. It is possible that the intention was to prove the feasibility of using an inexpensive AI chip for hypersonic weapons, regardless where it is made. However, such weapons can operate with different types of chips.

While few countries are able to design and manufacture such chips, a growing number of nations, including Germany, France, Japan, North Korea and Iran, have launched hypersonic weapon research and development programmes. Even the Houthis, an Iranian-backed rebel group that controls most of Yemen, have claimed they have been testing hypersonic missiles capable of reaching Mach 8.

The proliferation of hypersonic weapon technology has been a major concern for the United States. In 2017, the Rand Corporation suggested that Washington work with Moscow and Beijing to prevent other countries from acquiring such technology.

But while Chinese and Russian military experts admit hypersonic weapons technology poses certain risks, they said it is more likely to accelerate the collapse of the US-centered world order.

They have reasoned that hypersonic missiles could penetrate the defences of America’s massive aircraft carrier fleet, which it has long relied on for global military superiority.

In a war game exercise conducted last year by Chinese scientists, a Ford-class aircraft carrier battle group was destroyed by about 20 hypersonic anti-ship missiles. If more countries possessed hypersonic weapons, the advantage enjoyed by the few countries with significant naval power over the past five centuries may come to an end, according to some experts.

North Korea says it test-fired new solid-fuel hypersonic missile

On April 3, North Korea successfully test-fired a hypersonic gliding missile, and on Monday, Iran claimed to have used hypersonic missiles to attack Israeli military bases in retaliation for Israel’s bombing of the Iranian embassy. Currently, there is no evidence to suggest that China or Russia provided weapons or technical support to either country.

The Beijing Power Machinery Research Institute is affiliated with China Aerospace Science and Industry Corporation, a major supplier of China’s hypersonic weapons. Dalian University of Technology, closely associated with China’s navy, is a major research and development base for advanced manufacturing.


Source : MSN

Samsung to Get Up To $6.4 Billion in US Grants for Chip Plants

Alicia Diaz, Mackenzie Hawkins and Yoolim Lee wrote . . . . . . . . .

The Biden administration plans to award Samsung Electronics Co. as much as $6.4 billion in grants to increase chip production in Texas, as part of US efforts to bolster domestic semiconductor manufacturing.

The South Korean company plans to invest more than $40 billion overall, including in two foundry fabrication sites that will produce 4-nanometer and 2nm logic chips — one generation beyond the current state of the art. The massive project will encompass a research and development site, and an advanced chip packaging facility in Taylor, Texas, that will help produce high bandwidth memory chips, which are critical for artificial intelligence applications.

The award will also be used to expand Samsung’s existing chipmaking facility in Austin, Texas, which will support US aerospace, defense and automotive industries, the Commerce Department said in a statement. The city of Taylor is just outside Austin.

It’s the latest in a series of multibillion-dollar awards by the Biden administration, which is using the 2022 Chips and Science Act to revitalize American chipmaking after decades of production shifting to Asia. Another goal is to counter the technological rise of China, which is building up its own semiconductor industry.

Samsung opted not to tap loans or loan guarantees under the Chips Act, unlike Intel Corp. and Taiwan Semiconductor Manufacturing Co., both of which are also set to receive multibillion dollar grants. Samsung’s project is expected to benefit from an investment tax credit, which US officials said is likely to cover as much as 25% of qualified capital expenditures.

The Samsung award is part of President Joe Biden’s plans “to bring the manufacture of leading-edge semiconductors back to the United States,” Commerce Secretary Gina Raimondo told reporters.

The Taylor project adds to a robust semiconductor ecosystem in Texas, including tens of billions of dollars of additional investment from Texas Instruments Inc. in its home state and Samsung’s existing factory in Austin.

One of the fabs in Taylor is expected to start production in 2026 and the other will begin in 2027, according to a senior administration official. The company originally said the Taylor plant would begin production in the second half of 2024.

The investment is expected to create at least 17,000 construction jobs and more than 4,500 manufacturing jobs, according to Raimondo. It will contribute to local job creation as supply chains move closer to service the new facilities. US officials said $40 million will be dedicated to workforce funding and development.

The Austin expansion will boost output of some of the world’s most advanced semiconductor technologies, powering a range of industries including defense, aerospace and autos, Raimondo said. Samsung is the leading producer of memory chips, which store data on smartphones and computers, and it’s expanding into the foundry business, or producing chips that are designed by customers.

The facility for packaging — the combining and connecting of semiconductors — will use an advanced technique known as 2.5D packaging, also important for artificial intelligence.

The award will further Samsung’s competitive edge with rival chipmakers that have also received US investments. TSMC announced that it will manufacture 2nm technology at an expanded facility in Phoenix. Samsung’s hometown rival, SK Hynix Inc., also plans to build an advanced packaging facility in Indiana.

The Chips Act — which set aside $39 billion in grants plus $75 billion in loans and guarantees — has spurred more than $200 billion in private semiconductor investments. Intel snagged almost $20 billion in grants and loans. TSMC, the main chipmaker for Nvidia Corp. and Apple Inc., got $11.6 billion.

The announcement will set off a months-long due diligence period during which Samsung and the Commerce Department will hammer out final terms. The money will then be disbursed as the project hits key construction and production milestones, with the potential for clawbacks if the firm falls short of its promises.


Source : BNN Bloomberg

CATL Unveils TENER, the World’s First Five-Year Zero Degradation Energy Storage System with 6.25MWh Capacity

On April 9, CATL unveiled TENER, the world’s first mass-producible energy storage system with zero degradation in the first five years of use in Beijing, China. Featuring all-round safety, five-year zero degradation and a robust 6.25 MWh capacity, TENER will accelerate large-scale adoption of new energy storage technologies as well as the high-quality advancement of the sector.

While preventing the degradation of capacity over the first five years of use is a significant advancement in increasing the lifespan of batteries, the zero degradation of power is also important for energy storage power plants aiming to meet the requirements of new electric power systems. Leveraging biomimetic SEI (solid electrolyte interphase) and self-assembled electrolyte technologies, TENER has cleared roadblocks for the movement of lithium ions and achieved zero degradation for both power and capacity, ensuring zero growth of auxiliary power consumption throughout full life cycle, thereby creating “ageless” energy storage system.

Powered by cutting-edge technologies and extreme manufacturing capabilities, CATL has resolved the challenges caused by highly active lithium metals in zero-degradation batteries, which effectively helps prevent thermal runaway caused by oxidation reaction.

Immense Energy in a Compact Space: 20-foot Container with 6.25 MWh Capacity

TENER achieves an impressive 6.25 MWh capacity in the TEU container, representing a 30% increase in energy density per unit area and a 20% reduction in the overall station footprint, thus enhancing energy density and efficiency through innovative design within a limited space.

CATL’s cutting-edge cell technology supports the outstanding performance of the system. TENER is equipped with long service life and zero-degradation cells tailored for energy storage applications, achieving an energy density of 430 Wh/L, an impressive milestone for LFP batteries used in energy storage.

Dedicated quality management system to ensure ultimate safety

To achieve ultimate safety in energy storage, CATL has established a dedicated, end-to-end quality management system that includes technology development, proof testing, operation monitoring, and safety failure analysis. It sets different safety goals as required by different scenarios, and then develop the corresponding safety technology to meet those goals. In order to ensure the effectiveness of such technologies, CATL has built a validation platform to simulate the safety test of energy storage systems in different power grid scenarios. After the project is put into operation, CATL continuously monitors its operation status through AI-powered risk monitoring and intelligent early warning, calculates the failure rate of energy storage products throughout their life cycle, and thus verify the safety design goals and continue optimizing them.

CATL has reduced the failure rate to the PPB level for cells used in TENER, which, when extended to the operation throughout its full lifecycle, can effectively lower operating costs and significantly enhance IRR (internal rate of return).

Energy storage is a pivotal element of the green energy transition. CATL has steadfastly dedicated itself to delivering world-class energy storage solutions for customers around the world. The unveiling of TENER signifies another milestone in CATL’s ongoing commitment to energy transition. Looking ahead, CATL will remain resolute in its pursuit of open innovation and collaborate with global industry partners to forge a path of mutual prosperity as an industry leader in innovation and advanced technology.


Source : Yahoo!

Long Read: A Vigilante Hacker Took Down North Korea’s Internet. Now He’s Taking Off His Mask

ANDY GREENBERG wrote . . . . . . . . .

A LITTLE OVER two years have passed since the online vigilante who would call himself P4x fired the first shot in his own one-man cyberwar. Working alone in his coastal Florida home in late January of 2022, wearing slippers and pajama pants and periodically munching on Takis corn snacks, he spun up a set of custom-built programs on his laptop and a collection of cloud-based servers that intermittently tore offline every publicly visible website in North Korea and would ultimately keep them down for more than a week.

P4x’s real identity, revealed here for the first time, is Alejandro Caceres, a 38-year-old Colombian-American cybersecurity entrepreneur with hacker tattoos on both arms, unruly dark brown hair, a very high tolerance for risk, and a very personal grudge. Like many other US hackers and security researchers, Caceres had been personally targeted by North Korean spies who aimed to steal his intrusion tools. He had detailed that targeting to the FBI but received no real government support. So he decided to take matters into his own hands and to send a message to the regime of Kim Jong Un: Messing with American hackers would have consequences. “It felt like the right thing to do here,” Caceres told WIRED at the time. “If they don’t see we have teeth, it’s just going to keep coming.”

As he sought an outlet to broadcast that message to the Kim regime, Caceres told his story to WIRED while he carried out his attack, providing screen-capture videos and other evidence that he was, in fact, single-handedly disrupting the internet of an entire country in real time. But it was only just before going public that he decided to invent the P4x pseudonym for himself. The handle, pronounced “pax,” was a cheeky allusion to his intention of forcing a kind of peace with North Korea through the threat of his own punitive measures. He hoped that by hiding behind that name, he might evade not just North Korean retaliation but also criminal hacking charges from his own government.

Instead of prosecuting him, however, Caceres was surprised to find, in the wake of his North Korean cyberattacks, the US government was more interested in recruiting him. Caceres would spend much of the next year on a strange journey into the secretive world of America’s state-sponsored hacking agencies. Adopted informally by a Pentagon contractor, he was invited to present his techniques to high-level US defense and intelligence officials. He carried out a long-term hacking project designed to impress his new audience, hitting real foreign targets. And he pitched Department of Defense officials on a mode of US government-sanctioned cyberattacks that, like his solo North Korean takedown, would be far leaner, faster, and arguably more effective than Washington’s slow and risk-averse model of cyberwar.

Caceres’ pitch never got the green light. Now, partly due to his frustration with that experience, he’s finally dropping his pseudonym to send a new message, this one aimed at his fellow Americans: that the US government needs to wield its hacking powers far more aggressively. “Both the NSA and the DOD have a ton of talented hackers, yet when it comes to actually performing disruptive cyber operations, for some reason we as a country are just frozen and scared,” Caceres says. “And that needs to change.”

He points to ransomware actors, mostly based in Russia, who extracted more than a billion dollars of extortion fees from victim companies in 2023 while crippling hospitals and government agencies. North Korea–affiliated hackers, meanwhile, stole another $1 billion in cryptocurrency last year, funneling profits into the coffers of the Kim regime. All of that hacking against the West, he argues, has been carried out with relative impunity. “We sit there while they hack us,” Caceres says.

So Caceres is now arguing that it’s time for the US to try the P4x approach: that a part of the solution to foreign cybersecurity threats is for the American government’s own hackers to show their teeth—and to use them far more often.

Caceres and the Pentagon contractor that partnered with him—whose founder agreed to talk to WIRED on the condition we not name him or his company—spent much of the past two years advocating within the US government for that far-more-brazen approach to state-sponsored cyberattacks. They describe it as a special forces model: single hackers or small teams carrying out nimble, targeted digital disruptions, in contrast to the US’s traditionally slower and more bureaucratic approach to cyberwarfare.

“You can have an impact here, and it can be asymmetrical, and it can occur on a much faster timescale,” summarizes the founder of the hacker startup that worked with Caceres to pitch the Pentagon.

He cites a military principle that each member of a special forces unit should have the effect of 16 conventional soldiers. “With what we and P4x were doing, we wanted to increase that ratio a hundredfold,” he says. “And P4x would teach other operators how to do it.”

P4x Americana

In his public life as a security researcher, Caceres is known as a talented and sometimes brash figure in the hacker community: The second-generation Colombian-American, who used the hacker handle _hyp3ri0n long before he adopted the P4x pseudonym, is the founder of the cybersecurity company Hyperion Gray and a frequent speaker at events like the Defcon hacker conference, where he has shared methods for lone hackers to amplify their reach and effects through cloud services and high-performance computing clusters. He’s also the creator of a somewhat controversial vulnerability scanning tool called PunkSpider, which he announced at Defcon in 2021 that he intended to use to scan every website in the world and publicly reveal all of their hackable vulnerabilities.

From the beginning of his hacker career, Caceres has never been one to shy away from the most aggressive applications of the digital dark arts. His first job out of college, while he pursued a graduate degree in international science and technology policy, was working for a subsidiary of the notorious military contractor formerly known as Blackwater, doing open-source intelligence investigations for corporate security and executive protection—what he describes as a “Google sweatshop.” Within a few years, however, Caceres and his firm Hyperion Gray were getting grants from the Pentagon’s Defense Advanced Research Projects Agency, using his growing prowess in cloud and high-performance computing to scan the dark web as part of Darpa’s Memex program devoted to advancing search technologies for national security applications.

In that dark-web scouring, Caceres says, he regularly came upon child sexual abuse materials forums and even violent extremist content. He admits that he didn’t hesitate to hack some of those sites, pulling data off their backend servers and anonymously handing it to contacts at the Department of Homeland Security. “That probably wasn’t 100 percent legal,” he says, “but I didn’t necessarily give a fuck.”

Even his company’s name, Hyperion Gray, reflected Caceres’ sense of the murky territory in which he traveled: It combined his _hyp3ri0n hacker handle, the name of a Titan from Greek mythology, with a shade in the middle of the spectrum between ideas of whitehat and blackhat hacking.

So when North Korean hackers chose Caceres as one of their targets in 2021, it was inevitable that he wouldn’t let that slight go unanswered. In January of that year, a fellow hacker whom Caceres didn’t know approached him online through a friend and asked if he’d take a look at what appeared to be an interesting software exploitation program, offering it to Caceres to download. A day or so later, Caceres read a blog post from Google’s Threat Analysis Group warning that North Korean hackers were going after US security researchers in what appeared to be an effort to steal their hacking tools and intel. Sure enough, when Caceres checked the file he’d downloaded and run on his own computer, he saw that it included a backdoor—only the fact that he’d quarantined the program on his machine had protected him from being fully compromised.

Caceres, appalled, reported the hacking attempt to the FBI. He says the bureau ultimately did little more than a fact-finding interview with him, with no real follow-up. So after nursing his grudge for a full year, he decided to take matters into his own hands. In late January of 2022, he began running custom hacking scripts designed to target a few key North Korean routers responsible for internet traffic into and out of the country, repeatedly checking whether they were online and, if they were, amplifying his waves of malicious data requests to cause them to crash. He described the project to WIRED at the time as about equivalent in workload to a “small-to-medium” size penetration test of the kind Hyperion Gray would carry out for clients.

North Korean watchers soon began to notice that the Hermit Kingdom’s entire web, from its government portal to its state-run airline booking site, had been knocked offline for days on end by an apparent cyberattack. News outlets published stories that pointed to the country’s recent missile tests, implying that the attack was perhaps the work of another country’s cyber forces—one news site suggested the US or possibly China—to signal to North Korea that it should stop threatening its neighbors. In reality, it was all the work of one aggrieved Florida man in his pajamas.

A Cyber Army of One

The US government may have had nothing to do with North Korea’s man-made internet outage, but it quietly took an interest in it. In the weeks after WIRED published its story about P4x’s solo hacking feat, Caceres began to receive messages from hacker friends with connections to the Pentagon and intelligence agencies—sent not to P4x but to Caceres’ own true-name accounts. Several agencies, he was told, were intrigued by his work and interested in talking to him.

For those in the know, identifying Caceres had been worryingly easy: He had given away hints in posts to his Twitter account about his North Korean targeting prior to his decision to hide it behind the P4x pseudonym. After P4x’s public statements on the cyberattack, one fellow hacker had even posted screenshots of Caceres’ now-deleted tweets, though without spelling out what exactly they revealed.

One friend had discussed Caceres’ work with a high-ranking military official and told Caceres there was someone the official now wanted Caceres to talk to: a longtime military intelligence contractor who had done contract work for the Joint Special Operations Command, which oversees groups like the Army’s Delta Force and the Navy’s Seal Team Six. WIRED agreed to call him Angus, though that isn’t his real name.

A few weeks after his North Korea attack, Caceres met Angus in the offices of Angus’ Pentagon-funded hacker startup. Angus began by warning Caceres that he was potentially in danger of reprisal from the North Korean state and that he should be wary of the possibility of a physical attack that might be made to look like a mugging, or of someone tampering with his prescription medications. “Before that, I was nervous,” Caceres says. “After that, I was shit-scared.” Angus suggested the hacker ought to arm himself. (Caceres, not one for half measures, later bought three guns and multiple bulletproof vests.)

Angus quizzed Caceres about his past hacking activities, his allegiance to other governments—he said he didn’t have any—and his politics. He specifically asked Caceres if he was a Marxist. Caceres confirmed he was not. With that brief vetting out of the way, they went out for drinks and talked late into that night about what a P4x-style US special forces hacker team might look like and what sort of work they might do together to demonstrate that model to the Department of Defense.

Soon after, Angus convened a meeting of military and intelligence staff at his startup’s office, where they listened to a presentation from Caceres. Standing before an audience of officials from Cyber Command, Special Operations Command, the NSA, and the Marines’ Cyberspace Command known as Marforcyber, he detailed his North Korean hacking project as a case study and laid out principles for how it could be replicated: Aim for “easy and impactful.” Minimize the number of “cooks in the kitchen.” Iterate rapidly. He laid out a timeline for operations that suggested assembling teams of two to four hackers, with support from researchers and analysts, and taking just a few days to plan an operation.

“For the US government, any execution on target is typically a six-month process. P4x did it in two weeks,” says Angus. “The whole point was that he can show them how to do it, and if they wanted to they could fund it and see it happen.”

The response to the presentation was positive, if somewhat cynical, according to Angus. “Most of them put their faces in their palms when they realized what he’d done and how he did it, and the only thing that stopped them from doing it was bureaucracy,” he says. Caceres remembers that one audience member responded with a joke: that Caceres forgot the step where he presents a 100-slide PowerPoint deck to someone who doesn’t understand what he’s talking about and then denies him authorization.

Too Hot for the DOD

After that initial presentation, Caceres gave a similar talk at MIT Lincoln Laboratory, a Pentagon-funded R&D lab, and then was invited to speak again near DC, this time to a larger audience of military and intelligence officials. That last speaking engagement was delayed due to paperwork and would ultimately never happen—in part because Caceres was by then busy with a more hands-on demonstration.

In fact, while Angus sought funding for their project, Caceres was targeting another foreign adversary to show the full extent of the cyber mayhem a small team could inflict. Over the next year, he and another hacker who goes by the handle tu3sday undertook a broad intrusion campaign whose details they both declined to describe.

Caceres and tu3sday did much of that hacking in the offices of Angus’ startup, visiting the site dozens of times over the next year. But Caceres says that the work was never officially carried out with the company’s approval or involvement, much less the Pentagon’s. “The message I got was that it was a ‘don’t get caught, we’ll disavow all knowledge’ sort of thing,” Caceres says.

Meanwhile, Angus found that he was running into roadblocks in his attempts to get more official support for their experiment. “The initial response was, ‘We should be doing that,’” he says. But as he “talked it up a very long chain, all the way to very senior people to see if they were amenable to it, most of them were not.”

Angus says he’s still not sure why their project didn’t gain traction. He believes it was only partly a considered decision by the Pentagon not to engage in the more aggressive and freewheeling hacking operations that they’d proposed. He thinks the resistance can be explained at least as much by the Department of Defense’s sclerotic management and the difficulty of convincing it to try anything new that involves risk. “There were forces that were nebulous even to me, and people wanted different things, and there was a lot of risk-aversion,” says Angus. “Bureaucracy was 100 percent a factor. They were trying to dilute liability.”

After nearly a year of Angus’ efforts to find funding for the project and failing, Caceres says he gave up and ended his visits to the startup. “We had something really powerful to share here, and something they really need,” he says of the US military. “And nothing really came of it.”

A Tricky Trilemma

Even after hitting that brick wall, Caceres hasn’t given up on his pitch. He has come forward, now speaking to WIRED under his own name, in an attempt to rekindle the idea of American “special forces” hackers: He still imagines small teams who make the lives of ransomware gangs hell by repeatedly targeting and destroying their servers and personal computers, or breaking into the wallets of North Korea–linked hackers to steal back the hundreds of millions of dollars in cryptocurrency that they routinely take from US victims. He also suggests that US cyber forces could easily, like P4x, take down North Korea’s internet for extended periods in response to every major theft until Kim’s state-sponsored thieves are forced to stop. “Like, every time they steal a hundred million dollars we shut down their internet for a year,” says Caceres. “Let’s figure out what kind of disruption makes them think twice about doing this.”

Caceres even argues that this sort of deterrent response could be used in answer to Russian or North Korean misbehavior in the physical world, such as war crimes or human rights abuses. Most radically, he argues that US cyber targeting shouldn’t be confined to military, government, or even criminal targets: that civilian infrastructure should be fair game, too. He says the effects of those attacks should be thought of not as “cyberwar” but as simply another kind of trade embargo or sanctions. “Exactly as we’re withholding certain goods and businesses from Russia right now,” Caceres says, “we could be withholding the internet.”

Those with a less run-and-gun approach to cyber policy, of course, cite some valid reasons why the Pentagon might hesitate to adopt the P4x model of state-sponsored hacking. If US Cyber Command took up Caceres’ notion of attacking civilian targets, for instance, it could be accused of war crimes, just as some have said about Russia’s cyberattacks against Ukraine, argues Jacquelyn Schneider, a cyber conflict researcher at Stanford’s Hoover Institution. Not only would indiscriminate attacks on civilians be arguably immoral, she points out, they’d potentially invite other countries to do the same, or worse.

“That’s not nice, and it’s not a good norm,” says Schneider. She says that much of the US government’s slow approach to cyberattacks stems from its care to ensure it avoids unintentionally hitting civilians as well as breaking international law or triggering dangerous blowback.

Still, Schneider concedes that Caceres and Angus have a point: The US could be using its cyber forces more, and some of the explanations for why it doesn’t amount to bureaucracy. “There are good reasons, and then there are bad reasons,” says Schneider. “Like, we have complicated organizational politics, we don’t know how to do things differently, we’re bad at using this type of talent, we’ve been doing it this way for 50 years, and it worked well for dropping bombs.”

America’s offensive hacking has, by all appearances, gotten less aggressive and less nimble over the past half decade, Schneider points out. Starting in 2018, for instance, General Paul Nakasone, then the head of Cyber Command, advocated a “defend forward” strategy aimed at taking cyber conflict to the enemy’s network rather than waiting for it to occur on America’s turf. In those years, Cyber Command launched disruptive hacking operations designed to cripple Russia’s disinformation-spouting Internet Research Agency troll farm and take down the infrastructure of the Trickbot ransomware group, which some feared at the time might be used to interfere in the 2020 election. Since then, however, Cyber Command and other US military hackers appear to have gone relatively quiet, often leaving the response to foreign hackers to law enforcement agencies like the FBI, which face far more legal constraints.

Caceres isn’t entirely wrong to criticize that more conservative stance, says Jason Healey, who until February served as a senior cybersecurity strategist at the US Cybersecurity and Infrastructure Security Agency. He responds to Caceres’ cyberhawk arguments by citing the Subversive Trilemma, an idea laid out in a 2021 paper by the researcher Lennart Maschmeyer: Hacking operations have to choose among intensity, speed, and control. Even in earlier, more aggressive years, US Cyber Command has tended to turn up the dial for control, Healey says, prioritizing it over those other variables. But he notes there may in fact be certain targets—such as ransomware gangs or hackers working for Russia’s no-holds-barred GRU military intelligence agency—who might warrant resetting those dials. “For those targets,” says Healey, “you really can release the hounds.”

P4x Is Dead, Viva P4x

As for Caceres himself, he says he’s not opposed to American hacking agencies taking a conservative approach to limiting their damage or protecting civilians—as long as they take action. “There’s being conservative,” he says, “and then there’s doing fuck all.”

On the argument that more aggressive cyberattacks would lead to escalation and counterattacks from foreign hackers, Caceres points to the attacks those foreign hackers are already carrying out. The ransomware group AlphV’s catastrophic attack on Change Healthcare in February, for instance, crippled medical claim platforms for hundreds of providers and hospitals, effects about as disruptive for civilians as any cyberattack can be. “That escalation is already happening,” Caceres says. “We’re not doing anything, and they’re still escalating.”

Caceres says he hasn’t entirely given up on convincing someone in the US government to adopt his more gloves-off approach. Ditching the P4x handle and revealing his real name is, in some sense, his last-ditch attempt to get the US government’s attention and restart the conversation.

But he also says he won’t be waiting for the Pentagon’s approval before he continues that approach on his own. “If I keep going with this alone, or with just a few people that I trust, I can move a lot faster,” he says. “I can fuck shit up for the people who deserve it, and I don’t have to report to anyone.”

The P4x handle may be dead, in other words. But the P4x doctrine of cyberwarfare lives on.


Source : WIRED

TSMC Wins $6.6 billion US Subsidy for Arizona Chip Production

David Shepardson and Stephanie Kelly wrote . . . . . . . . .

The U.S. Commerce Department said on Monday it would award Taiwan Semiconductor Manufacturing Co’s, opens new tab U.S. unit a $6.6 billion subsidy for advanced semiconductor production in Phoenix, Arizona and up to $5 billion in low-cost government loans.

TSMC agreed to expand its planned investment by $25 billion to $65 billion and to add a third Arizona fab by 2030, Commerce said in announcing the preliminary award. The Taiwanese company will produce the world’s most advanced 2 nanometer technology at its second Arizona fab expected to begin production in 2028, the department said.

“These are the chips that underpin all artificial intelligence, and they are the chips that are necessary components for the technologies that we need to underpin our economy, but frankly, a 21st century military and national security apparatus,” Commerce Secretary Gina Raimondo said in a statement.

TSMC, the world’s largest contract chipmaker and a major supplier to Apple, opens new tab and Nvidia, opens new tab, had previously announced plans to invest $40 billion in Arizona. TSMC expects to begin high-volume production in its first U.S. fab there by the first half of 2025, Commerce said.

The $65 billion-plus investment by TSMC is the largest foreign direct investment in a completely new project in U.S. history, the department said.

Congress in 2022 approved the Chips and Science Act to boost domestic semiconductor output with $52.7 billion in research and manufacturing subsidies. Lawmakers also approved $75 billion in government loan authority.

TSMC Arizona has also committed to support the development of advanced packaging capabilities through partners in the U.S. to allow customers to purchase advanced chips that are made entirely on U.S. soil, the department said, adding 70% of TSMC customers were U.S. companies.

TSMC CEO C.C. Wei said the company would help U.S. tech firms “unleash their innovations by increasing capacity for leading-edge technology through TSMC Arizona.”

Commerce expects the projects will create 6,000 direct manufacturing jobs and 20,000 construction jobs. The department said 14 direct TSMC suppliers plan to construct or expand U.S. plants.

At full capacity, TSMC’s three fabs in Arizona will manufacture tens of millions of leading-edge chips in 5G/6G smartphones, autonomous vehicles, and AI data center servers, the department said.

Through its Arizona fabs, TSMC will support key customers like Apple, Nvidia, Advanced Micro Devices, opens new tab and Qualcomm, opens new tab “by addressing their leading-edge capacity demand, mitigating supply chain concerns, and enabling them to compete effectively in the ongoing digital transformation era,” the department added.

TSMC said in a separate statement that its Arizona factories aim to achieve a 90% water recycling rate, adding that the company has started the design phase of building a water reclamation plant with a goal of achieving “near zero liquid discharge”.

Commerce last month announced $8.5 billion in grants and up to $11 billion in loans for Intel, opens new tab to subsidize leading-edge chip production from the same program.


Source : Reuters

TSMC’s First Japan Plant to Hit 60% Local Procurement by 2030

Jane Lanhee Lee wrote . . . . . . . . .

Taiwan Semiconductor Manufacturing Co. told Japan’s Prime Minister Fumio Kishida that the chipmaker expects to reach 60% local procurement for its first chip plant in the country by 2030.

The comment was made by TSMC Chief Executive Officer C.C. Wei during his meeting with Kishida on Saturday when the official visited the company’s plant in Kumamoto. TSMC Spokesman Nina Kao told Bloomberg News that it’s a target for indirect materials used in manufacturing process but not included in the final products, and the goal doesn’t include machinery.

Japanese officials are hoping TSMC’s arrival will help boost local suppliers’ technology and business. Tokyo has already allocated ¥476 billion ($3.1 billion) for TSMC’s first factory which was built though a joint venture between the Taiwanese chipmaker and local companies including Sony Group Corp. Japan’s government has pledged an additional ¥732 billion in subsidies for TSMC to build a second fab.

TSMC said it plans to start shipping logic chips for camera sensors and automobiles from the first facility in Kumamoto by the end of this year.


Source : BNN Bloomberg