Recently, cloud computing company Lambda announced a large-scale AI infrastructure partnership with Microsoft. According to the latest press release, Lambda will deploy tens of thousands of NVIDIA GPUs as part of the collaboration, although the specific transaction amount was not disclosed. These GPUs include the recently launched NVIDIA GB300 NVL72 systems, which began shipping in the past few months. Stephen Balaban, CEO of Lambda, stated in the press release: "It's exciting to see the Microsoft and Lambda teams working together to deploy these massive AI supercomputers. Our partnership with Microsoft has spanned over eight years, and this collaboration is a significant milestone in our relationship." Microsoft first launched the NVIDIA GB300 NVL72 cluster in October, demonstrating its continued investment in AI computing power. Founded in 2012, Lambda has emerged as a standout in the current AI boom, raising $1.7 billion in venture capital—reflecting strong market demand for AI infrastructure and computing capabilities. The announcement comes just hours after Microsoft revealed a $9.7 billion AI cloud computing capacity deal with Australian data center company IREN. Meanwhile, OpenAI also announced a $38 billion cloud computing agreement with Amazon, under which it will use Amazon's cloud services over the next seven years. Additionally, OpenAI signed a landmark $300 billion cloud computing deal with Oracle in September. Amazon's AWS division reported in last week's Q3 earnings that it is on track for its best operating income in three years, with $33 billion in sales so far this year. Andy Jassy, CEO of Amazon, stated: "AWS is growing at a pace we haven't seen since…

In the journey of web browser innovation, Arc was a brilliant but overly radical experiment. Now, its "spirit" is being reborn in a more powerful successor. Named Dia, this new AI browser draws inspiration from the "failures" of its predecessor Arc. Josh Miller, founder of The Browser Company, confirmed on Sunday that the new browser will bring "Arc's greatest works" (such as the iconic sidebar mode) to Dia, integrating them deeply with AI-native memory and agent capabilities. This strategic move is significant. Following The Browser Company's acquisition by Atlassian for a staggering $610 million, Dia's roadmap has become extremely clear. It suggests that Dia may already hold a first-mover advantage in the fierce AI browser race, as it is not starting from scratch but building on the company's valuable experience and user insights gained from developing Arc over more than a year. Looking back at Arc's launch in mid-2023, it was hailed as a complete reinvention around modern internet usage. It offered separate workspaces for professional and personal browsing (Spaces), supported pinned tabs, a command bar similar to Apple's Spotlight search, and a powerful sidebar integrating search, tab lists, bookmarks, and audio controls. However, Arc may have gone too far in its innovation. Miller later admitted that Arc ultimately proved "too complex" for most people. In a blog post earlier this year, he reflected that Arc was simply too different, requiring users to learn too many new things for too little reward. More importantly, Arc lacked cohesion in its core features and values. Its experimental nature was part of its…

Coca-Cola has once again used AI generation technology to reinvent its iconic "Holidays Are Coming" holiday ad, but the move has seemingly sparked negative feedback from consumers, who describe the visual effects of the latest Christmas ad as "unsettling." New Ad "Even Creepier": Animal Characters and Rough Animation Last year, Coca-Cola faced criticism for its AI-generated holiday ad featuring "sliding wheels and eerie faces." This year, Coca-Cola attempted to avoid the challenge of creating realistic human characters by releasing a new ad titled "Holidays Are Coming" starring primarily animal characters. However, the visual style of the new ad has been described as "chaotic," alternating between realistic and exaggerated aesthetics. The movements of the polar bears, pandas, and sloths are "highly unnatural," resembling rough 2D image animations rather than refined 3D CG models. Compared to advanced tools like OpenAI's Sora 2 or Google's Veo 3, Coca-Cola's video appears "very outdated." The only "improvement" is that the wheels of the iconic Coca-Cola truck finally spin this year, rather than sliding statically over the snow as in previous years. 70,000 Clips, 100-Person Team: AI Projects Still Require Human Input According to The Wall Street Journal, Coca-Cola collaborated with two AI studios, Silverside and Secret Level, on its latest holiday campaign. Notably, Coca-Cola declined to disclose the specific cost of the campaign but stated that approximately 100 people were involved in the project—an effort similar in scale to the company's previous non-AI production projects. The team included five "AI specialists" from Silverside, who were responsible for prompting and refining over 70,000 AI-generated video clips.…

Recently, Apple insider Mark Gurman revealed that iOS 27 is under active development and will make its debut at next year's Worldwide Developers Conference in June. This new system will include not only iOS 27 but also macOS 27, watchOS 27, and other operating systems, with the most anticipated update being significant enhancements to Apple Intelligence. Notably, while the newly released iOS 26 introduced over 20 new Apple Intelligence features, this update is not considered a major breakthrough in the AI field. Apple is laying the groundwork for the upcoming iOS 27, aiming for more in-depth innovations in AI technology. With the iOS 26.4 update, Apple will launch a personalized Siri, marking a comprehensive overhaul of the voice assistant. The new Siri will feature three core capabilities: screen awareness, personal context, and app control, allowing users to obtain more accurate information through the intelligent assistant. For example, users can ask Siri about their mother's flight information or lunch reservations, and Siri will provide intelligent responses based on relevant data in email and messaging apps.

Google's Gemini AI has recently received a major update, with its built-in interactive workspace Canvas now supporting direct generation of professional presentations (PPT). This feature marks a significant leap for AI from text assistance to visual content creation. Users only need to input simple prompts or upload documents to automatically generate well-structured, visually rich slides, greatly simplifying the traditional PPT production process. According to the official announcement, this feature is powered by the Gemini 2.5 Pro model and is currently rolling out to Gemini Advanced subscribers (Google One AI Premium plan). Free users are expected to gain access in the coming weeks. Since its launch in March this year, Canvas has expanded from document and code editing to multimodal creation. The addition of the PPT generation feature further strengthens its positioning as an "all-in-one AI creative studio." Simple Operation: Dual Input Modes Unlock Efficiency Canvas's PPT generation feature is designed to be highly user-friendly, supporting two main input methods to ensure a seamless experience from idea to final product. First, generation via text prompts: Users open Canvas (located in the toolbar) within the Gemini app and input prompts such as "Create a sales presentation based on Google's Q3 earnings report" or "Design a product launch PPT." The AI will automatically build a framework, including a title page, key data visualizations, and a summary page. The generation process takes only a few seconds, with slides automatically matching theme colors, layout designs, and embedding relevant images (with clear source attribution). Second, the document upload mode prioritizes practicality: It supports uploading files such…

Hyundai Motor and NVIDIA recently announced plans to deepen their collaboration by jointly establishing an artificial intelligence (AI) factory based on NVIDIA's latest Blackwell architecture. First revealed at this year's Consumer Electronics Show (CES), the partnership will focus on co-developing multiple projects, including autonomous vehicles, smart factories, and robotics technology. Supported by the South Korean government, the collaboration was detailed at the 2025 APEC Korea Summit. Hyundai Motor and NVIDIA stated that their partnership will enter a new phase, focusing on co-innovating core physical AI technologies. The planned AI factory will be equipped with 50,000 NVIDIA Blackwell GPUs to enable integrated AI training, validation, and deployment. The factory will integrate in-vehicle AI technology, autonomous driving, manufacturing plant automation, and robotics, while leveraging several key NVIDIA platforms. The NVIDIA DGX platform will support large-scale AI model training and software development, while the Omniverse and Cosmos platforms will power digital twin technology for optimizing manufacturing and autonomous driving testing. Meanwhile, DRIVE AGX Thor will serve as the real-time AI brain for vehicles and robots. In addition to building the AI factory, Hyundai Motor and NVIDIA have signed a memorandum of understanding with South Korea's Ministry of Culture, Sports and Tourism and Ministry of Science and ICT to collaborate with stakeholders in advancing the national AI ecosystem. The investment, approximately $3 billion, will be used to establish Hyundai Motor's Physical AI Application Center, NVIDIA's AI Technology Center, and multiple physical AI data centers. Bae Kyung-hoon, Vice Minister of South Korea's Ministry of Science and ICT, welcomed the collaboration, stating that the development of…

Microsoft recently announced that its Azure ND GB300v6 virtual machine has achieved an industry-leading new record of 1.1 million tokens per second in inference speed on Meta's Llama 2 70B model. Satya Nadella, CEO of Microsoft, stated on social media: "This achievement is the result of our long-standing partnership with NVIDIA and our expertise in running AI at production scale." The Azure ND GB300 virtual machine is powered by NVIDIA's Blackwell Ultra GPUs, specifically the NVIDIA GB300 NVL72 system. It features a single-machine architecture with 72 NVIDIA Blackwell Ultra GPUs and 36 NVIDIA Grace CPUs. Optimized specifically for inference workloads, this virtual machine delivers a 50% increase in GPU memory and a 16% boost in Thermal Design Power (TDP). To validate the performance improvement, Microsoft ran the Llama 2 70B model (FP4 precision) on 18 ND GB300v6 virtual machines within a single NVIDIA GB300 NVL72 domain, using NVIDIA TensorRT-LLM as the inference engine. Microsoft stated: "An Azure ND GB300v6 cluster with one NVL72 rack achieved a total inference speed of 1.1 million tokens per second." This new record surpasses Microsoft's previous achievement of 865,000 tokens per second on the NVIDIA GB200 NVL72 rack. Based on the system configuration, each GPU delivers approximately 15,200 tokens per second. Microsoft has also provided detailed simulation processes, along with all log files and results. The performance record has been verified by Signal65, an independent performance validation and benchmarking company. Russ Fellows, Vice President of Labs at Signal65, noted in a blog post: "This milestone not only breaks the million-tokens-per-second barrier but does so on…

Hello Everyone, Today, as remote operation and maintenance, development collaboration, and multi-host scenarios become increasingly complex, traditional SSH tools can no longer meet engineers’ demands for efficiency, security, and intelligence. The launch of Aeroshell is precisely to break through this limitation. As a new-generation intelligent terminal management software, it is more than just a tool for connecting to servers—it is a comprehensive work platform integrating terminal access, file management, status monitoring, and AI enhancement. 👉 Official Website: https://termdev.com/ Pain Points of Traditional Tools Inefficient multi-host switching: Separate session maintenance is required for each login. Tedious file transfer: Lack of resumable transfer and verification, resulting in poor stability. Lack of intuitive monitoring: System status relies on additional scripts or tools. Insufficient intelligence: Complex issues still require researching materials and trial-and-error. Core Feature Highlights Unified Multi-Host Management Support for grouping and tag management Automatic session recovery, eliminating repeated input Batch host import Intelligent Terminal & Multi-Screen Splitting 4-screen split support Host map for intuitive display of latency and status Support for command broadcasting and SFTP directory following High-risk command interception Enhanced File Management & Transfer Drag-and-drop transfer Resumable transfer and transfer history MD5/SHA256 verification support Support for hiding directories and files Real-Time Status Monitoring Visualization of disk, memory, latency, and other metrics No need to build additional monitoring systems Multiple Intelligent Modes Enable "ai-on" in the terminal to enter AI mode for intelligent prompts and natural language-to-command conversion Directly chat with AI to generate and explain command parameters Task orchestration: Generate dedicated DSL language via natural language for batch execution of complex tasks…

It automatically adjusts the screen color temperature and filters blue light — more powerful and easier to use than the system's built-in night mode.

An extremely useful PDF processing tool with 27 features in total. It’s completely free, ad-free, with no hidden tricks, and supports offline use. https://tools.pdf24.org/en/