China AI & robotics developments, rewritten weekly-context-free for US readers. Published daily at 07:00 Beijing time.
Tue, Sep 1, 2026
Dell reports $7.57 billion net profit for first half of fiscal 2027Dell Technologies reported a net profit of $7.57 billion for the first half of its fiscal 2027, marking a 256% year-over-year increase driven by strong server and AI infrastructure demand.
The Round Rock-based hardware giant posted total revenue of $90.81 billion for the six-month period ending July 31, 2026. This performance reflects sustained momentum in its Infrastructure Solutions Group, which continues to benefit from enterprise adoption of artificial intelligence workloads.
Analysts note that Dell’s profitability surge underscores the ongoing capital expenditure cycle among major cloud providers and enterprises investing in high-performance computing clusters. The company’s second-quarter results, covering May through July, further solidified this trend.
Apple's new CEO Ternus promises impressive iPhone 18 launchJohn Ternus, Apple's newly appointed chief executive, issued his first internal memo signaling a strong upcoming launch for the iPhone 18 Pro and Max, marking the start of his leadership era.
On Sept. 1, John Ternus succeeded Tim Cook as Apple's CEO, while Cook transitioned to the role of Executive Chairman. In his inaugural address to staff, Ternus emphasized that the forthcoming iPhone 18 Pro and Max releases would be highly impactful.
This leadership change represents a significant milestone for the Cupertino-based tech giant. The focus on the next-generation flagship devices underscores the company's priority in maintaining hardware innovation momentum under its new operational structure.
Apple CEO John Ternus targets $58M pay in 2027SEC filings reveal Apple's incoming CEO John Ternus has a target compensation of roughly $58 million for fiscal 2027, with 75% tied to equity and performance metrics.
The documents submitted on Sept. 1 outline a base salary of $3 million, with the majority of earnings linked to long-term incentives. This structure aligns executive rewards with shareholder value growth over several years.
Ternus succeeds Tim Cook, whose current package remains undisclosed in this specific filing. The shift highlights Apple's continued emphasis on performance-based remuneration for its top leadership.
SoftBank-backed SB Energy files for IPO to fund AI data centersSB Energy, an AI infrastructure firm backed by SoftBank, Nvidia and OpenAI, filed for a public offering, signaling growing investor appetite for the power assets underpinning the generative AI boom.
The company provides critical energy infrastructure for large-scale computing clusters. Its backers include Japanese conglomerate SoftBank, chipmaker Nvidia and ChatGPT developer OpenAI, according to CNBC. The filing comes as tech giants race to secure stable power supplies for their expanding AI training operations.
Anthropic launches Claude Fable 5.1, Mythos 5.1 with lower cache costsAnthropic released two new AI models on Sept. 1, claiming superior performance for coding and knowledge work while cutting cache read fees by 75%.
The San Francisco-based firm introduced Claude Fable 5.1 and Claude Mythos 5.1, positioning them as its most advanced tools for developers and professionals. Both models share the same underlying architecture but differ in specific capabilities.
A key commercial update is the 75% reduction in cache read pricing, a move likely aimed at attracting cost-sensitive enterprise clients. This follows Anthropic’s ongoing strategy to compete directly with OpenAI and Google in the high-performance LLM market.
UBS: China unlikely to match ASML top tool in decadeA UBS report indicates China will not replicate ASML's extreme ultraviolet lithography machines within ten years, underscoring the persistent efficacy of Western export controls on advanced semiconductor manufacturing.
The analysis highlights that despite state-backed efforts by firms like SMEE, the complexity of EUV systems remains a significant barrier. This reinforces the strategic advantage held by Dutch firm ASML, the sole producer of these critical tools.
For US investors, this suggests continued reliance on existing supply chain bottlenecks as a defensive moat against Chinese chip advancement. The timeline implies no near-term disruption to current market dominance.
John Deere tests AI chatbot to boost farmer profitsAgricultural machinery giant John Deere is piloting an AI assistant named "JD" designed to help farmers increase profitability by providing data-driven insights on best practices and historical trends.
The company said the tool aims to translate complex agricultural data into actionable advice for users. This initiative marks a significant step in integrating generative AI into traditional farming operations, moving beyond hardware sales toward software-as-a-service models.
For US investors, this signals how legacy industrial firms are leveraging AI to deepen customer retention. The pilot reflects a broader industry trend of embedding intelligence directly into operational workflows rather than offering standalone consumer chatbots.
Avatr delivers 8,082 units in August; launches 07L and OTA updatesAvatr Technology reported delivering 8,082 vehicles in August while introducing the new Avatr 07L model. The Beijing-based electric vehicle maker also rolled out over-the-air software upgrades across its entire lineup.
Avatr, a joint venture between Changan Automobile, Huawei, and CATL, announced the August delivery figures on Sept. 1. The company simultaneously launched the Avatr 07L, expanding its premium EV portfolio.
The automaker confirmed that all current models received OTA updates to enhance functionality. Separately, the Avatr 06T chassis earned two awards from the China Automotive Technology and Research Center, including "Best Handling," signaling continued focus on driving dynamics.
Baidu CFO says AI profits to match search soonBaidu Chief Financial Officer Haijian He stated that the company's substantial investments in artificial intelligence are expected to generate profit and cash flow comparable to its core search business in the near term.
According to a Bloomberg report on Sept. 1, He emphasized that this financial milestone would validate Baidu’s strategic pivot from an internet portal to an AI-centric technology firm. The executive noted that current AI-related revenue streams already span multiple sectors, including cloud computing and consumer applications.
This projection comes as Baidu faces pressure to demonstrate tangible returns on its heavy capital expenditure in generative AI models and infrastructure. By aligning AI profitability with its legacy search engine economics, the Beijing-based tech giant aims to reassure investors of its long-term viability in the competitive global AI landscape.
Mon, Aug 31, 2026
OpenAI buys tens of thousands of Apple Macs for AI infrastructureOpenAI has reportedly purchased tens of thousands of Apple Mac computers, signaling a shift in how the company sources hardware for its AI development and operations.
According to reports from 24/7 Wall St., OpenAI is acquiring Macs by the tens of thousands. This move positions Apple not just as a consumer device maker, but as a key supplier for large-scale AI infrastructure.
The purchase suggests that Apple’s silicon, particularly the M-series chips, may offer efficiency advantages for specific AI workloads or local inference tasks. For investors, this adds a new dimension to Apple’s growth story beyond traditional smartphone sales.
Pentagon deploys ChatGPT and Grok to 3 million personnelThe US Department of Defense has rolled out secure versions of OpenAI's ChatGPT and xAI's Grok, granting access to approximately 3 million civilian and military staff. This move signals a significant shift in how the Pentagon integrates generative AI into daily operations.
According to TechCrunch, the deployment provides authorized users with tailored interfaces for both models, aiming to enhance productivity while maintaining strict security protocols. The initiative follows broader federal efforts to adopt large language models for administrative and analytical tasks.
This integration places major commercial AI tools directly within the defense infrastructure, raising questions about data handling and model alignment. It represents one of the largest enterprise-scale adoptions of generative AI by any government entity to date.
Apple Accuses OpenAI of Destroying Evidence in Trade Secrets SuitApple has filed a court motion accusing OpenAI of destroying evidence related to its trade secrets lawsuit, escalating tensions between the two tech giants over alleged poaching and intellectual property theft.
The Cupertino-based company alleges that San Francisco-based OpenAI failed to preserve critical documents. This claim arises from Apple’s ongoing suit against former employees who joined the AI developer. The dispute centers on whether proprietary information was improperly transferred during hiring.
Legal experts note such accusations can lead to severe sanctions if proven. Both companies declined immediate comment on the specific filing details. The case remains under active litigation in US federal courts.
Zhipu AI targets full-scale deployment on day one with next-gen modelChinese AI developer Zhipu AI announced a 'large foundation' strategy for its upcoming model, aiming for immediate enterprise readiness upon release. This approach signals a shift from experimental scaling to production-grade reliability in China's competitive LLM market.
During its semi-annual earnings briefing on Aug. 31, Zhipu stated the new architecture focuses on maximizing base model capabilities before mid-training and post-training phases. The company did not disclose parameter counts but emphasized that the goal is to support full-scale business calls immediately after launch.
Zhipu, known for its GLM series of models, faces intensifying competition from domestic rivals like Moonshot AI and Baidu. By prioritizing day-one operational stability, Zhipu aims to differentiate itself in an enterprise sector increasingly wary of models that require extensive tuning before commercial use.
WeChat Pay AI card adds DeepSeek, OpenClaw integrationTencent’s WeChat Pay launched an AI-exclusive virtual card that now supports DeepSeek Harness and OpenClaw, enabling users to complete purchases directly within chat interfaces.
The update follows earlier integrations with WorkBuddy and QClaw. According to IT Home, the feature allows authorized users to trigger a full cycle from conversational recommendation to checkout without leaving the AI interface.
This move positions Tencent to embed payment infrastructure deeper into third-party large language models. It reflects a broader strategy in China to streamline agentic commerce by linking financial rails directly to emerging AI assistants.
Momenta reports near-breakeven in first post-IPO half-yearAutonomous driving firm Momenta reported a 75.9% revenue jump and narrowed adjusted losses to under $2 million for the first half of 2026, signaling its transition from heavy R&D spending to financial viability.
The Beijing-based company, known for its data-driven autonomous driving solutions, posted 1.6 billion yuan (roughly $224 million) in revenue for the period ending June 30, 2026. Gross margin held at 73.2%, according to the filing.
Excluding non-operational items like share-based compensation, adjusted net loss shrank to 14.1 million yuan (approx. $2 million). This marks a critical milestone for Momenta, which had previously prioritized technological development over profitability. Investors are now scrutinizing whether its commercial model can sustain growth without further capital injections.
Nvidia invests $3.5 billion in MediaTek to expand AI chip reachNvidia has committed a $3.5 billion investment in Taiwanese semiconductor firm MediaTek, aiming to integrate its proprietary technology into MediaTek's chips as Big Tech accelerates custom silicon development.
The deal positions Nvidia to embed its architecture within MediaTek’s product line, potentially broadening the ecosystem for AI inference and edge computing. This move comes as major tech firms increasingly seek alternatives to pure Nvidia hardware for their data centers.
MediaTek, known primarily for mobile processors, will utilize Nvidia’s technology to enhance its chip design capabilities. The partnership signals a strategic shift for Nvidia, leveraging capital to secure deeper ties with key supply chain partners amidst rising competition from AMD and custom ASICs.
Minimax H3 model powers AI-generated livestream on consumer GPUsChinese AI firm Minimax demonstrated its H3 model generating continuous video content from live comments using two Nvidia RTX 5090 cards, signaling progress in real-time generative media.
The project, dubbed "SlopTV," streams AI-generated visuals driven by viewer input. It highlights the capability of Minimax's latest multimodal architecture to operate within accessible hardware constraints. Minimax is a Shanghai-based startup known for its conversational and video generation models.
Running on dual RTX 5090s suggests a push toward democratizing high-fidelity generative video. This contrasts with previous enterprise-only requirements for similar real-time rendering tasks. The demonstration underscores the rapid iteration pace of Chinese AI labs in the multimodal space.
China's AI fund, CP Robot invest in Kuaishou's Kling unitKuaishou announced that China’s National Artificial Intelligence Fund and Charoen Pokphand Robot Limited will inject 1.4 billion yuan into Beijing Kling, signaling state-backed support for the company’s generative AI ambitions.
According to a Hong Kong exchange filing on Aug. 31, both entities signed agreements to become additional investors in Beijing Kling. The National AI Fund is a sovereign-backed vehicle established to accelerate domestic large model development. Charoen Pokphand Robot Limited is part of the Thai conglomerate’s robotics arm.
The investment totals 1.4 billion yuan (roughly $198 million) in cash capital. This move provides critical liquidity to Kuaishou’s AI division as it competes with rivals like ByteDance in the video generation sector.
Sun, Aug 30, 2026
BYD pushes OTA update adding Huawei HiCar to Han EVBYD has released a software update for its 2025 Han models, integrating Huawei's HiCar mirroring and enhancing security features. This move signals continued deepening of hardware-software partnerships in China's competitive electric vehicle market.
On Aug. 28, BYD rolled out version 1.5.1 over-the-air updates for the 2025 Han EV and DM-i trims equipped with LiDAR. The patch introduces seven new functions, including Huawei HiCar mirror mode for seamless phone-to-car integration and automatic uploading of high-risk footage from the vehicle’s Sentry Mode.
The update also includes six performance optimizations. For US readers, this highlights how Chinese automakers are rapidly iterating on software-defined vehicle capabilities. Partnering with tech giants like Huawei allows BYD to offer connectivity features that compete directly with global standards, reinforcing the trend of cross-industry collaboration in China’s auto sector.
Apple boosts Guizhou data center investment to billionsApple executives announced a multi-billion dollar expansion of its Guizhou data center, which is now fully powered by renewable energy. This move underscores the tech giant's commitment to sustainable infrastructure in China amidst ongoing regulatory scrutiny.
Isabel Ge Mahe, Apple’s Vice President and Managing Director for Greater China, disclosed the additional investment at the 2026 China International Big Data Industry Expo. The original facility opened in 2018. While specific figures were not detailed beyond "several billion dollars," the upgrade highlights Apple's long-term operational footprint in the region.
Ge also noted that Apple has operated in China for over 30 years. She claimed the local App Store ecosystem doubled developer revenue and sales over the past six years. These statements come as Apple navigates complex US-China tech tensions and domestic data compliance requirements.
Physix Frontier