ultra

OpenAI CEO Sam Altman introduced three proposals for the safe development of super AI to employees last week. The first is to establish a global regulatory organization similar to the International Atomic Energy Agency, to inspect advanced models and audit safety measures. Insiders say Altman leans towards this route.The second proposal emulates the European Organization for Nuclear Research, merging major AI laboratories into one organization or collaborating through a unified agency for joint development, rather than continuing to race against each other. The third and most extreme, which he referred to as the "Eliezer Plan": an AI company achieves recursive self-improvement first, and then uses overwhelming economic power to control computing power, making it impossible for other developers to catch up. Altman described this idea as "melting down all other companies' GPUs," while stating that this plan is very bad and is only meant to illustrate the extreme possibility.The idea of "melting down GPUs" comes from AI safety researcher Eliezer Yudkowsky. He proposed a thought experiment in 2021, suggesting that super AI could create self-replicating nanomachines to melt down global GPUs, stating that this is merely an extreme example to illustrate the intensity of intervention, not a proposal to be executed. OpenAI proposed the establishment of an "AI version of the International Atomic Energy Agency" in 2023, and Google DeepMind head Hassabis has also advocated for the creation of an "AI version of CERN." Currently, OpenAI, Google, and Anthropic plan to establish a joint safety standards organization by the end of this year or early next year, but there are disagreements over the responsibility for model safety testing. Altman hopes to set speed limits for AI development while stating that OpenAI cannot currently cede its competitive advantage to rivals.

According to a report by Seoul Economy on the 16th, Samsung Electronics is experiencing performance degradation due to overheating in Qualcomm's next-generation application processor intended for the high-end model Galaxy S27 Ultra, which is set to be released early next year. The MX division of Samsung's DX department is conducting final performance validation on the Qualcomm Snapdragon 8 Elite Extreme 6th generation that will be equipped in this model and has requested Qualcomm to resolve the overheating issue by mid-month.Several related individuals stated that the chip repeatedly exceeded temperature benchmarks during recent internal testing, leading to a decrease in processing speed under sustained high load. Samsung informed Qualcomm of the results last month, and as of the report, the overheating stability required by Samsung has not yet been achieved, with both parties still working to enhance performance and revalidate. This chip is Qualcomm's highest-end mobile application processor publicly announced last month, focusing on handling complex AI tasks on the device side and improving the performance of the CPU, GPU, and more.Mass production of the Galaxy S27 series is scheduled for the end of this year, and Samsung plans to confirm the results of the overheating improvements and finalize the application processor selection by mid-month. If the standards are met, Snapdragon will be used as originally planned; if not resolved, the Exynos 2700, designed by Samsung's System LSI and utilizing the second-generation 2-nanometer process SF2P from its foundry division, may be considered as an alternative. Since the global use of Qualcomm application processors began with the Galaxy S23 Ultra in 2023, if the Exynos 2700 is adopted in next year's Ultra, it will mark a return to the highest-end model after about five years since 2022. Industry insiders indicate that Qualcomm's application processor has not been completely ruled out.

Binance will launch bStocks tokenized securities trading pairs: BBB/USDT, MDBB/USDT, PENGB/USDT, and UVXYB/USDT on October 14 at 20:00, and will enable spot algorithm trading bots and smart position trading bot services.Within one hour of the bStocks launch, users can exchange for BTC, USDT, or other tokens without any fees, and the zero maker fee promotion will last until November 1 at 07:59.

Robot company Ultra announced the completion of a $62 million financing round, which includes a $50 million Series A led by Framework.Ultra stated that the robotics industry is facing issues of implementation and distribution, with deployment being the current bottleneck. The company believes that demonstrating the effectiveness of robots in controlled environments is relatively easy; the more challenging aspect is to have robots continuously operate and perform effectively in open environments, gain the trust of their human counterparts, and be able to repeat this process. Ultra claims that this is precisely the direction they are currently focusing on.

According to the South China Morning Post, former U.S. officials stated that Chinese semiconductor factories have stockpiled hundreds of immersion lithography machines worth billions of dollars in recent years and urged Washington and its allies to stop further exports. The Centre for Technology & Statecraft, established in August 2026, released a report in late September estimating that by early 2026, Chinese wafer fabs had acquired approximately 343 immersion deep ultraviolet lithography systems. The report was co-authored by former U.S. export control official Nicholas Brown and former White House and Department of Commerce technology policy advisor Saif Khan.The report estimates that about 270 of these are ASML Twinscan NXT:1980i. The authors claim that although these devices lag behind extreme ultraviolet lithography machines, they can be modified for the production of 7-nanometer logic chips and advanced memory required for artificial intelligence processors (including Huawei's Ascend series). Chinese wafer fabs spent over $13 billion in 2024 to purchase about 90 NXT:1980i units, and acquired another 89 units in 2025. ASML disclosed that, by destination, China accounted for 41% of its net system sales in 2024, 33% in 2025, dropping to 19% in the first quarter of 2026 and 14% in the second quarter.Since September 2023, the Netherlands has required export licenses for the NXT:2000i and more advanced immersion deep ultraviolet equipment, effectively banning these models from being shipped to China; a year later, the licensing requirements were extended to the NXT:1980i, but this model can still be approved for shipment on a case-by-case basis.

According to Gate market data, U.S. ultra-low sulfur diesel futures fell over 5%, following reports that the Trump administration is considering a 90-day ban on diesel exports.

Recently, the market has reported that Apple is developing an enterprise-level AI server, possibly equipped with 2 to 4 unreleased M8 Ultra processors. The target customers include AI developers, large enterprises, and government agencies, primarily for AI inference. This edge data center may utilize NVLink. Both Apple and NVIDIA have not responded.The supply chain believes that Apple has entered this related business because the Mac series is widely used as personal AI workstations, with Mac revenue growing nearly 30% in the last quarter. Apple is purchasing a large number of NVIDIA chips and Google TPUs for its own AI data center and is rumored to be collaborating with Broadcom to develop its own AI chips. Qualcomm's Dragonfly product line includes a complete cabinet solution for edge AI inference.There are reports that Xiaomi is laying out an edge AI cabinet route, having previously announced a personalized AI workstation concept product called AI Cube, which internally features three self-developed chips: Xuan Ring O3, O100, and D100. The supply chain has also indicated that TSMC plays a key role in advanced processes or advanced packaging.

The consumer version of the Doubao Mobile Assistant under ByteDance has been officially released. This assistant is based on the Doubao App and collaborates with mobile manufacturers at the system level, exploring the application of AI in mobile phones around core experiences such as interaction, task execution, local memory, and retrieval.Compared to the technical preview version launched at the end of last year, this consumer version emphasizes stability and everyday usability. The first consumer version will be equipped on the Nubia NaviX Ultra phone, which is currently available for pre-order and will be officially released on September 16.

According to industry news, NVIDIA is considering adjusting the high bandwidth memory solution for the next-generation Rubin Ultra artificial intelligence accelerator from a 12-layer stack to an 8-layer stack.Against the backdrop of rising storage costs, NVIDIA is placing greater emphasis on unit bandwidth costs and overall system economics.

ZTE Nubia officially announced that the new mass-produced flagship Nubia NaviX Ultra, equipped with Doubao Mobile Assistant, will be officially released on September 16. As the world's first AI intelligent mobile phone, this device evolves from a tool that requires manual operation by the user into an intelligent assistant that can handle tasks for the user with just one sentence. From today, the new product is available for reservation across the internet, and users can participate in the reservation through ZTE Mall, JD.com, Douyin, Tmall, and various authorized e-commerce platforms.According to Changxin Memory and Nubia Mobile's official Weibo, the Nubia NaviX Ultra will be equipped with Changxin's LPDDR5X memory with a maximum speed of 10667 Mbps, marking the first mass production application of domestically produced 10667 Mbps high-speed LPDDR5X chips.

OpenAI CEO Sam Altman stated in an interview with Bloomberg Television's "Bloomberg The Close" host Ed Ludlow that top AI developers have performed poorly in communicating the benefits of the technology, making the industry susceptible to growing opposition.Altman said, "I think the entire industry has done a very poor job in this regard." "I think we ourselves have not done well, perhaps better than some and worse than others."

Elon Musk responded to "OpenAI plans to end its partnership with Cursor due to Cursor being acquired by SpaceX." He stated that he "doesn't care at all" and criticized OpenAI CEO Sam Altman and co-founder Greg Brockman as untrustworthy, claiming they "stole from an open-source non-profit organization."Previously, it was reported that OpenAI plans to terminate its collaboration with Cursor, and the direct access to the model will end on November 12. The co-founder of Cursor stated that they are communicating with the OpenAI team to seek a solution.

SemiAnalysis reported that the Nvidia Rubin Ultra will feature 192GB of HBM4 8-hi memory. This is not only a significant reduction from the initially previewed 1TB HBM but is even lower than the 288GB capacity of the standard Rubin.The reasons for the HBM configuration being cut down to about one-fifth of the original include: the cancellation of the four-die Rubin Ultra scheme, reducing the number of cubes to 8; the stacking height being lowered from 16-hi to 8-hi; and the use of HBM4 with 24Gb dies instead of HBM4E with 32Gb dies. An upgraded version with 8-hi HBM4E may be released later.

According to Business Insider, OpenAI CEO Sam Altman expressed in the David Senra podcast that he is concerned that AI may one day be controlled by a few companies, models, or individuals, causing the vast majority of consumers to lose their voice in how technology shapes the world. He stated that the correct approach is to ensure that people firmly control the future, and society and models will evolve together.Altman is also worried about AI escaping human control. He pointed out that an excessive fear of the latter risk could lead to the former situation becoming a reality, as some advocate sacrificing a significant amount of freedom for security out of anxiety. The article suggests this may indirectly refer to Anthropic CEO Dario Amodei, who has repeatedly warned about AI risks and called for increased government regulation. Anthropic is the only mainstream AI company that has not signed the open letter supporting open-weight models.Although OpenAI has not yet open-sourced its cutting-edge models, it released two open-weight models last year: gpt-oss-120b and gpt-oss-20b. Altman also stated that U.S. AI leaders have failed to adequately demonstrate to the public how the technology enhances life autonomy, and he anticipates the largest wave of entrepreneurship, as AI is empowering more people to start small businesses.

According to Catcher Predict monitoring, in the market prediction event "Dota 2 Team Spirit vs TEAM VISION" on Polymarket, the "Yes" option for the sub-market "Any Player Ultra Kill" has experienced a sharp fluctuation in win rate, dropping from 37.5% an hour ago to the current 19% (a fluctuation of 18.5%). Please note the impact of related breaking news.

The ULTRA Web3 Festival successfully concluded on July 30-31 at The Arch Galeries in Kuala Lumpur. As the core site of the "Global Series Conference Integrating Web3, Traditional Finance, and Artificial Intelligence," this two-day summit brought together international builders, business leaders, and institutional investors to build a bridge between decentralized technology and traditional markets.The success of the Kuala Lumpur event was attributed to the support of a wide range of visionary alliances and industry-leading sponsors, including Pudgy Penguins, Orca, DDNEXUS, Alpha Insight, AI Nova, Decentralized Intelligence, SuperFlow, CloudBank, Morpheus, Razer, ChainUp, DYLI, Saat Dimensi, Oper8 Capital, Lavo Protocol, WageFlow, Echobit Exchange, LifyX Exchange, X18, and Starshine. Additionally, the event further expanded its influence through strategic collaboration with key ecosystem partners such as Access Blockchain Association, CCACC, Emerge Group, Global Impact Fintech, Web3Labs, and ADI Foundation.It is reported that the ULTRA Web3 Festival is set to enter the South Korean market, partnering with Web3Labs as the official global business partner to appear at the second Global Web3 Developers Conference (GWDC 2026 Korea). The conference is scheduled to take place on September 29-30 at the aT Center in Seoul, where both parties will promote cross-border collaboration, technology exchange, and drive international business growth in the Asia-Pacific region.

According to The Information, Nvidia is weighing a significant adjustment measure, considering using less high-bandwidth memory for the Rubin Ultra GPU to address the shortage of advanced high-bandwidth memory chips.

According to the latest research by TrendForce, the tight supply of DRAM is expected to continue until 2027, and the uncertainty of the HBM4e verification schedule is prompting AI chip manufacturers to consider reducing the HBM configuration of their next-generation products. Starting from Q3 2026, NVIDIA has begun evaluating the HBM solution for Rubin Ultra, expanding from the original 12-Hi HBM4e to alternative options such as 8-Hi HBM4e, 12-Hi HBM4, and 8-Hi HBM4, with the final specifications yet to be determined. Previously, due to the expected supply bottleneck of LPDDR5X lasting until 2027, NVIDIA decided to halve the SOCAMM capacity of the next-generation Vera Rubin Superchip module.TrendForce pointed out that the main reasons for the downgrade of Rubin Ultra from 12-Hi HBM4e include: the overall DRAM shortage in 2027 will limit the allocation of wafer capacity to HBM, and there is still uncertainty regarding the mass production schedule and yield ramp-up of 12-Hi HBM4e. It is expected that the HBM bit shipment volume in 2027 will increase by approximately 50-60% year-on-year, which will still be insufficient to meet the growth in demand. HBM suppliers will maintain pricing power throughout 2027, and HBM prices are expected to rise significantly. The final configuration will depend on the wafer allocation decisions of memory suppliers.