"The number of discrete GPU developers from the U.S. and Western Europe shrank to three companies in 2025," notes Tom's Hardware, "from around 10 in 2000." (Nvidia, AMD, and Intel...) No company in the recent years — at least outside of China — was bold enough to engage into competition against these three contenders, so the very emergence of Bolt Graphics seems like a breakthrough. However, the major focuses of Bolt's Zeus are high-quality rendering for movie and scientific industries as well as high-performance supercomputer simulations. If Zeus delivers on its promises, it could establish itself as a serious alternative for scientific computing, path tracing, and offline rendering. But without strong software support, it risks struggling against dominant market leaders.
This week the Sunnyvale, California-based startup introduced its Zeus GPU platform designed for gaming, rendering, and supercomputer simulations, according to the article. "The company says that its Zeus GPU not only supports features like upgradeable memory and built-in Ethernet interfaces, but it can also beat Nvidia's GeForce RTX 5090 by around 10 times in path tracing workloads, according to slide published by technology news site ServeTheHome." There is one catch: Zeus can only beat the RTX 5090 GPU in path tracing and FP64 compute workloads. It's not clear how well it will handle traditional rendering techniques, as that was less of a focus. In speaking with Bolt Graphics, the card does support rasterization, but there was less emphasis on that aspect of the GPU, and it may struggle to compete with the best graphics cards when it comes to gaming. And when it comes to data center options like Nvidia's Blackwell B200, it's an entirely different matter.

Unlike GPUs from AMD, Intel, and Nvidia that rely on proprietary instruction set architectures, Bolt's Zeus relies on the open-source RISC-V ISA, according to the published slides. The Zeus core relies on an open-source out-of-order general-purpose RVA23 scalar core mated with FP64 ALUs and the RVV 1.0 (RISC-V Vector Extension Version 1.0) that can handle 8-bit, 16-bit, 32-bit, and 64-bit data types as well as Bolt's additional proprietary extensions designed for acceleration of scientific workloads... Like many processors these days, Zeus relies on a multi-chiplet design... Unlike high-end GPUs that prioritize bandwidth, Bolt is evidently focusing on greater memory size to handle larger datasets for rendering and simulations. Also, built-in 400GbE and 800GbE ports to enable faster data transfer across networked GPUs indicates the data center focus of Zeus.

High-quality rendering, real-time path tracing, and compute are key focus areas for Zeus. As a result, even the entry-level Zeus 1c26-32 offers significantly higher FP64 compute performance than Nvidia's GeForce RTX 5090 — up to 5 TFLOPS vs. 1.6 TFLOPS — and considerably higher path tracing performance: 77 Gigarays vs. 32 Gigarays. Zeus also features a larger on-chip cache than Nvidia's flagship — up to 128MB vs. 96MB — and lower power consumption of 120W vs. 575W, making it more efficient for simulations, path tracing, and offline rendering. However, the RTX 5090 dominates in AI workloads with its 105 FP16 TFLOPS and 1,637 INT8 TFLOPS compared to the 10 FP16 TFLOPS and 614 INT8 TFLOPS offered by a single-chiplet Zeus...
The article emphasizes that Zeus "is only running in simulation right now... Bolt Graphics says that the first developer kits will be available in late 2025, with full production set for late 2026."

Thanks to long-time Slashdot reader arvn for sharing the news.

China has introduced regulations requiring service providers to label AI-generated content, joining similar efforts by the European Union and United States to combat disinformation. The Cyberspace Administration of China and three other agencies announced Friday that AI-generated material must be labeled explicitly or via metadata, with implementation beginning September 1.

"The Labeling Law will help users identify disinformation and hold service suppliers responsible for labeling their content," the CAC said. App store operators must verify whether applications provide AI-generated content and review their labeling mechanisms. Platforms can still offer unlabeled AI content if they comply with relevant regulations and respond to user demand.

OpenAI has warned the U.S. government that restricting AI models from learning from copyrighted material would threaten America's technological leadership against China, according to a proposal submitted [PDF] to the Office of Science and Technology Policy for the AI Action Plan.

In its March 13 document, OpenAI argues its AI training aligns with fair use doctrine, saying its models don't replicate works but extract "patterns, linguistic structures, and contextual insights" without harming commercial value of original content. "If the PRC's developers have unfettered access to data and American companies are left without fair use access, the race for AI is effectively over. America loses, as does the success of democratic AI," OpenAI stated.

The Microsoft-backed startup criticized European and UK approaches that allow copyright holders to opt out of AI training, claiming these restrictions hinder innovation, particularly for smaller companies with limited resources. The proposal comes as China-based DeepSeek recently released an AI model with capabilities comparable to American systems despite development at a fraction of the cost.

An anonymous reader shares a report: Germany's foreign intelligence service in 2020 put at 80%-90% the likelihood that the coronavirus behind the COVID-19 pandemic was accidentally released from China's Wuhan Institute of Virology, two German newspapers reported on Wednesday.

According to a joint report by publications Die Zeit and Sueddeutscher Zeitung, Germany's spying agency BND had indications that the institute had conducted gain-of-function experiments, whereby viruses are modified to become more transmissible to humans for research purposes.

It also had indications that numerous violations of safety regulations had occurred at the lab, the papers said. The spy agency assessment's was based on an unspecified intelligence operation code-named "Saaremaa" as well as on publicly-available data. It had been commissioned by the office of Germany's chancellor at the time, Angela Merkel, but never published, the report said.

An anonymous reader quotes a report from TechCrunch: Anthropic's CEO Dario Amodei is worried that spies, likely from China, are getting their hands on costly "algorithmic secrets" from the U.S.'s top AI companies -- and he wants the U.S. government to step in. Speaking at a Council on Foreign Relations event on Monday, Amodei said that China is known for its "large-scale industrial espionage" and that AI companies like Anthropic are almost certainly being targeted. "Many of these algorithmic secrets, there are $100 million secrets that are a few lines of code," he said. "And, you know, I'm sure that there are folks trying to steal them, and they may be succeeding."

More help from the U.S. government to defend against this risk is "very important," Amodei added, without specifying exactly what kind of help would be required. Anthropic declined to comment to TechCrunch on the remarks specifically but referred to Anthropic's recommendations to the White House's Office of Science and Technology Policy (OSTP) earlier this month. In the submission, Anthropic argues that the federal government should partner with AI industry leaders to beef up security at frontier AI labs, including by working with U.S. intelligence agencies and their allies.

Several Asian airlines have tightened restrictions on portable battery chargers amid growing concerns about fire risks, following a January blaze that destroyed an Air Busan aircraft in South Korea. South Korean airlines now require passengers to keep portable chargers within arm's reach rather than in overhead bins, a rule implemented March 1 to ease public anxiety, according to the Transportation Ministry. Taiwan's EVA Air and China Airlines have banned using or charging power banks on flights but still allow them in overhead compartments.

Thai Airways announced a similar ban last Friday, citing "incidents of in-flight fires on international airlines." Battery-related incidents on U.S. airlines have increased from 32 in 2016 to 84 last year, with portable chargers identified as the most common culprit, according to Federal Aviation Administration data. The International Civil Aviation Organization has banned lithium-ion batteries from cargo holds since 2016, though no industry standard exists for regulating power banks.

Morgan Stanley has reduced its iPhone shipment forecasts after Apple confirmed the delay of a more advanced Siri personal assistant, dampening prospects for accelerating phone upgrades. The investment bank now predicts 230 million iPhone shipments in 2025 (flat year-over-year) and 243 million in 2026 (up 6%), down from previous estimates.

An upgraded Siri was the most sought-after Apple Intelligence feature among prospective buyers, according to the bank's survey data. "Access to Advanced AI Features" appeared as a top-five driver of smartphone upgrades for the first time, with about 50% of iPhone owners who didn't upgrade to iPhone 16 citing the delayed Apple Intelligence rollout as affecting their decision. The firm also incorporated headwinds from China tariffs in its assessment, noting Apple is unlikely to fully offset these costs without broader exemptions.

The Ballista botnet is actively exploiting a high-severity remote code execution flaw (CVE-2023-1389) in TP-Link Archer AX-21 routers, infecting over 6,000 devices primarily in Brazil, Poland, the UK, Bulgaria, and Turkey. Tom's Hardware reports: According to a new report from the Cato CTRL team, the Ballista botnet exploits a remote code execution vulnerability that directly impacts the TP-Link Archer AX-21 router. The botnet can lead to command injection which then makes remote code execution (RCE) possible so that the malware can spread itself across the internet automatically. This high severity security flaw (tracked as CVE-2023-1389) has also been used to spread other malware families as far back as April 2023 when it was used in the Mirai botnet malware attacks. The flaw also linked to the Condi and AndroxGh0st malware attacks.

Ballista's most recent exploitation attempt was February 17, 2025 and Cato CTRL first detected it on January 10, 2025. Of the thousands of infected devices, the majority of them are concentrated in Brazil, Poland, the United Kingdom, Bulgaria and Turkey; with the botnet targeting manufacturing, medical/healthcare, services and technology organizations in the United States, Australia, China and Mexico.

An anonymous reader quotes a report from the BBC: A former senior Facebook executive has told the BBC how the social media giant worked "hand in glove" with the Chinese government on potential ways of allowing Beijing to censor and control content in China. Sarah Wynn-Williams -- a former global public policy director -- says in return for gaining access to the Chinese market of hundreds of millions of users, Facebook's founder, Mark Zuckerberg, considered agreeing to hiding posts that were going viral, until they could be checked by the Chinese authorities.

Ms Williams -- who makes the claims in a new book -- has also filed a whistleblower complaint with the US markets regulator, the Securities and Exchange Commission (SEC), alleging Meta misled investors. The BBC has reviewed the complaint. Facebook's parent company Meta, says Ms Wynn-Williams had her employment terminated in 2017 "for poor performance." It is "no secret we were once interested" in operating services in China, it adds. "We ultimately opted not to go through with the ideas we'd explored." Meta referred us to Mark Zuckerberg's comments from 2019, when he said: "We could never come to agreement on what it would take for us to operate there, and they [China] never let us in."

Facebook also used algorithms to spot when young teenagers were feeling vulnerable as part of research aimed at advertisers, Ms Wynn-Williams alleges. A former New Zealand diplomat, she joined Facebook in 2011, and says she watched the company grow from "a front row seat." Now she wants to show some of the "decision-making and moral compromises" that she says went on when she was there. It is a critical moment, she adds, as "many of the people I worked with... are going to be central" to the introduction of AI. In her memoir, Careless People, Ms Wynn-Williams paints a picture of what she alleges working on Facebook's senior team was like.

August, 2023: Thousands of Crypto Scammers are Enslaved by Human-Trafficking Gangsters, Says Bloomberg Reporter. ("They'd lure young people from across Southeast Asia...with the promise of well-paying jobs in customer service or online gambling.")

February, 2025: A coordinated response begins by Thai, Chinese and Myanmar authorities, which includes cutting power, internet, and fuel supplies to the scam centers.

Today: The Associated Press reports that thousands of the people liberated from locked compounds in Myanmar now "have found themselves trapped once again, this time in overcrowded facilities with no medical care, limited food and no idea when they'll be sent home." Thousands of sick, exhausted and terrified young men and women, from countries all over the world squat in rows, packed shoulder to shoulder, surgical masks covering their mouths and eyes. Their nightmare was supposed to be over... The armed groups who are holding the survivors, as well as Thai officials across the border, say they are awaiting action from the detainees' home governments. It's one of the largest potential rescues of forced laborers in modern history, but advocates say the first major effort to crack down on the cyber scam industry has turned into a growing humanitarian crisis...

An unconfirmed list provided by authorities in Myanmar says they're holding citizens from 29 countries including Philippines, Kenya and the Czech Republic. Authorities in Thailand say they cannot allow foreigners to cross the border from Myanmar unless they can be sent home immediately, leaving many to wait for help from embassies that has been long in coming. China sent a chartered flight Thursday to the tiny Mae Sot airport to pick up a group of its citizens, but few other governments have matched that. There are roughly 130 Ethiopians waiting in a Thai military base, stuck for want of a $600 plane ticket. Dozens of Indonesians were bused out one morning last week, pushing suitcases and carrying plastic bags with their meager possessions as they headed to Bangkok for a flight home... The recent abrupt halt to U.S. foreign aid funding has made it even harder to get help to released scam center workers...

It's not clear how much of an effect these releases will have on the criminal groups that run the scam centers. February marked the third time the Thais have cut internet or electricity to towns across the river. Each time, the compounds have managed to work around the cuts. Large compounds have access to diesel-powered generators, as well as access to internet provider Starlink, experts working with law enforcement say.
The article also points out that "The people released are just a small fraction of what could be 300,000 people working in similar scam operations across the region, according to an estimate from the United States Institute of Peace. Human rights groups and analysts add that the networks that run these illegal scams will continue to operate unless much broader action is taken against them..."

"The United Nations Office on Drugs and Crimes estimates that between $18 billion and $37 billion was lost in Asia alone in 2023, with minimal government action against the criminal industry's spread."

The Washington Post reports: Meta was willing to go to extreme lengths to censor content and shut down political dissent in a failed attempt to win the approval of the Chinese Communist Party and bring Facebook to millions of internet users in China, according to a new whistleblower complaint from a former global policy director at the company.

The complaint by Sarah Wynn-Williams, who worked on a team handling China policy, alleges that the social media giant so desperately wanted to enter the lucrative China market that it was willing to allow the ruling party to oversee all social media content appearing in the country and quash dissenting opinions. Meta, then called Facebook, developed a censorship system for China in 2015 and planned to install a "chief editor" who would decide what content to remove and could shut down the entire site during times of "social unrest," according to a copy of the 78-page complaint exclusively seen by The Washington Post.

Meta chief executive Mark Zuckerberg also agreed to crack down on the account of a high-profile Chinese dissident living in the United States following pressure from a high-ranking Chinese official the company hoped would help them enter China, according to the complaint, which was filed in April to the Securities and Exchange Commission [SEC]. When asked about its efforts to enter China, Meta executives repeatedly "stonewalled and provided nonresponsive or misleading information" to investors and American regulators, according to the complaint.

Wynn-Williams bolstered her SEC complaint with internal Meta documents about the company's plans, which were reviewed by The Post. Wynn-Williams, who was fired from her job in 2017, is also scheduled to release a memoir this week documenting her time at the company, titled "Careless People: A Cautionary Tale of Power, Greed, and Lost Idealism." According to a memo in the complaint, Meta leaders faced aggressive pressure by Chinese government officials to host Chinese users' data to local data centers, which Wynn-Williams alleges would have made it easier for the Chinese Communist Party to covertly obtain the personal information of its citizens.
Wynn-Williams told the Washington Post that "for many years Meta has been working hand in glove with the Chinese Communist Party, briefing them on the latest technological developments and lying about it."

Reached for a comment, Meta spokesman Andy Stone told the Washington Post it was "no secret" they'd been interested in operating in China. "This was widely reported beginning a decade ago. We ultimately opted not to go through with the ideas we'd explored, which Mark Zuckerberg announced in 2019." Although the Post shares new details about what a Facebook privacy policy staffer offer China in negotations in 2014. ("In exchange for the ability to establish operations in China, FB will agree to grant the Chinese government access to Chinese users' data — including Hongkongese users' data.")

The Post also describes one iteration of a proposed agreement in 2015. "To aid the effort, Meta built a censorship system specially designed for China to review, including the ability to automatically detect restricted terms and popular content on Facebook, according to the complaint...

"In 2017, Meta covertly launched a handful of social apps under the name of a China-based company created by one of its employees, according to the complaint."

"The ubiquitous ESP32 microchip made by Chinese manufacturer Espressif and used by over 1 billion units as of 2023 contains an undocumented 'backdoor' that could be leveraged for attacks," writes BleepingComputer.

"The undocumented commands allow spoofing of trusted devices, unauthorized data access, pivoting to other devices on the network, and potentially establishing long-term persistence." This was discovered by Spanish researchers Miguel Tarascó Acuña and Antonio Vázquez Blanco of Tarlogic Security, who presented their findings yesterday at RootedCON in Madrid. "Tarlogic Security has detected a backdoor in the ESP32, a microcontroller that enables WiFi and Bluetooth connection and is present in millions of mass-market IoT devices," reads a Tarlogic announcement shared with BleepingComputer. "Exploitation of this backdoor would allow hostile actors to conduct impersonation attacks and permanently infect sensitive devices such as mobile phones, computers, smart locks or medical equipment by bypassing code audit controls...."

Tarlogic developed a new C-based USB Bluetooth driver that is hardware-independent and cross-platform, allowing direct access to the hardware without relying on OS-specific APIs. Armed with this new tool, which enables raw access to Bluetooth traffic, Targolic discovered hidden vendor-specific commands (Opcode 0x3F) in the ESP32 Bluetooth firmware that allow low-level control over Bluetooth functions. In total, they found 29 undocumented commands, collectively characterized as a "backdoor," that could be used for memory manipulation (read/write RAM and Flash), MAC address spoofing (device impersonation), and LMP/LLCP packet injection.

Espressif has not publicly documented these commands, so either they weren't meant to be accessible, or they were left in by mistake.
Thanks to Slashdot reader ZipNada for sharing the news.

The White House is weighing measures to restrict Chinese artificial-intelligence upstart DeepSeek, including banning its chatbot from government devices because of national-security concerns, WSJ reported Friday, citing people familiar with the matter. From the report: U.S. officials are worried about DeepSeek's handling of user data, which the Chinese company says it stores in servers located in China, the people said. Officials also believe DeepSeek hasn't sufficiently explained how it uses the data it collects and who has access to the data, they said.

The Trump administration is likely to adopt a rule that would bar people from downloading DeepSeek's chatbot app onto U.S. government devices, the people said. Officials are also considering two other possible moves: banning the DeepSeek app from U.S. app stores and putting limits on how U.S.-based cloud service providers could offer DeepSeek's AI models to their customers, people close to the matter said. They cautioned that discussions about these two moves were still at an early stage.

A U.S. congressional committee has urged Americans to remove Chinese-made wireless routers from their homes, including those made by TP-Link, calling them a security threat that opened the door for China to hack U.S. critical infrastructure. From a report: The House of Representatives Select Committee on China has pushed the Commerce Department to investigate China's TP-Link Technology Co, which according to research firm IDC is the top seller of WiFi routers internationally by unit volume. U.S. authorities are considering a ban on the sale of the company's routers, according to media reports.

Rob Joyce, former director of cybersecurity at the National Security Agency, told Wednesday's committee hearing that TP-Link devices exposed individuals to cyber intrusion that hackers could use to gain leverage to attack critical infrastructure. "We need to all take action and replace those devices so they don't become the tools that are used in the attacks on the U.S.," Joyce said, adding that he understood the Commerce Department was considering a ban.

AmiMoJo writes: China plans to issue guidance to encourage the use of open-source RISC-V chips nationwide for the first time, Reuters reports, citing two sources briefed on the matter, as Beijing accelerates efforts to curb the country's dependence on Western-owned technology.

The policy guidance on boosting the use of RISC-V chips could be released as soon as this month, although the final date could change, the sources said. It is being drafted jointly by eight government bodies, including the Cyberspace Administration of China, China's Ministry of Industry and Information Technology, the Ministry of Science and Technology, and the China National Intellectual Property Administration, they added.

In a policy paper, former Google CEO Eric Schmidt, Scale AI CEO Alexandr Wang, and Center for AI Safety Director Dan Hendrycks said that the U.S. should not pursue a Manhattan Project-style push to develop AI systems with "superhuman" intelligence, also known as AGI. From a report: The paper, titled "Superintelligence Strategy," asserts that an aggressive bid by the U.S. to exclusively control superintelligent AI systems could prompt fierce retaliation from China, potentially in the form of a cyberattack, which could destabilize international relations.

"[A] Manhattan Project [for AGI] assumes that rivals will acquiesce to an enduring imbalance or omnicide rather than move to prevent it," the co-authors write. "What begins as a push for a superweapon and global control risks prompting hostile countermeasures and escalating tensions, thereby undermining the very stability the strategy purports to secure."

Half of the world's climate-heating carbon emissions come from the fossil fuels produced by just 36 companies, analysis has revealed. From a report: The researchers said the 2023 data strengthened the case for holding fossil fuel companies to account for their contribution to global heating. Previous versions of the annual report have been used in legal cases against companies and investors.

The report found that the 36 major fossil fuel companies, including Saudi Aramco, Coal India, ExxonMobil, Shell and numerous Chinese companies, produced coal, oil and gas responsible for more than 20bn tonnes of CO2 emissions in 2023. If Saudi Aramco was a country, it would be the fourth biggest polluter in the world after China, the US and India, while ExxonMobil is responsible for about the same emissions as Germany, the world's ninth biggest polluter, according to the data.

Global emissions must fall by 45% by 2030 if the world is to have a good chance of limiting temperature rise to 1.5C, the internationally agreed target. However, emissions are still rising, supercharging the extreme weather that is taking lives and livelihoods across the planet. The International Energy Agency has said new fossil fuel projects started after 2021 are incompatible with reaching net zero emissions by 2050. Most of the 169 companies in the Carbon Majors database increased their emissions in 2023, which was the hottest year on record at the time.

Microsoft warned that an advanced Chinese hacking group is waging a campaign of supply-chain attacks. From a report: The company's threat intelligence division said in a blog post Wednesday that the group, known as Silk Typhoon, was targeting remote management tools and cloud applications in order to spy on a range of companies and organizations in the US and abroad.

Microsoft said it observed in late 2024 that hackers were targeting cloud storage services, from which they would steal keys that could be used to access customer data. The group breached state and local government organizations and companies in the technology sector, seeking information on US government policy and documents related to law enforcement investigations. Silk Typhoon was behind a December hack that targeted the US Treasury Department, compromising more than 400 computers, Bloomberg News previously reported.

China plans to commercialize nuclear fusion for emissions-free power generation by 2050, with its first operational project expected around 2050 after a demonstration phase starting in 2045. Bloomberg reports: China National Nuclear Corp. (CNNC) last year formed an industry alliance and set up a new national fusion company, the China Fusion Corp. It has attracted about 1.75 billion yuan ($240 million) in investment from CNNC and Zhejiang Zheneng Electric Power Co. for cutting-edge tokamak devices, which use magnetic fields to confine and control superheated plasma to produce power without emissions or significant radioactive waste. CNNC also plans to scale up production of its homegrown designs for regular nuclear fission reactors and small modular reactors over the next five years, the company's Vice General Manager Xin Feng said at the briefing.

China is set to leapfrog the US and France as the owner of the world's biggest reactor fleet by 2030. About 10 new reactors have been approved every year since power shortages emerged in 2022 and the country is expected to keep up that pace through 2030 to meet climate goals, CNNC said on Friday.

China's highest court has called for a crack down on the activities of paper mills, businesses that churn out fraudulent or poor-quality manuscripts and sell authorships. Nature: Some researchers are cautiously optimistic that the court's guidance will help curb the use of these services, while others think the impact will be minimal. "This is the first time the supreme court has issued guidance on paper mills and on scientific fraud," says Wang Fei, who studies research-integrity policy at Dalian University of Technology in China.

Paper mills sell suspect research and authorships to researchers who want journal articles to burnish their CVs. They are a significant contributor to overall research misconduct, particularly in China. Last month, the Supreme People's Court published a set of guiding opinions on technology innovation. Among the list of 25 articles, one called for lower courts to crack down on 'paper industry chains,' and for research fraud to be severely punished. Further reading:
Research Reveals Data on Which Institutions Are Retraction Hotspots;
Paper Mills Have Flooded Science With 400,000 Fake Studies, Experts Warn.