News stories tagged with #KI
OpenAI Shuts Down Sora: End of AI Video Platform After Just Six Months
OpenAI has announced the discontinuation of its AI-powered video platform Sora, launched in December 2024 and released as a standalone app in September 2025. Despite an initial surge of over one million downloads within five days, Sora is being shut down due to high operational costs, declining user engagement, regulatory challenges, and a strategic shift toward more profitable ventures. Disney, which recently invested $1 billion in Sora, will retain its investment but will abandon joint projects. The app and web version will cease operations on April 26, 2026, with the API ending on September 24, 2026.
Nvidia CEO Jensen Huang made several significant remarks in a podcast, proposing $250,000 worth of AI tokens as a new productivity benchmark for developers, naming the 1993 game Doom the most influential PC game of all time, and confirming he declined a 2013 leadership offer from TSMC. He also emphasized that traditional succession planning is less important than continuous knowledge transfer and widely distributed responsibility within the company.
Elon Musk’s Terafab: $20 Billion Chip Factory for Tesla, SpaceX, and xAI – Ambition and Skepticism
Elon Musk has unveiled the ambitious Terafab project, a $20 billion semiconductor mega-factory in Austin, Texas, designed to produce chips for Tesla, SpaceX, and xAI. The vertically integrated facility aims to consolidate the entire value chain—from chip design to manufacturing—under one roof and reduce reliance on external supply chains. However, the planned 2-nanometer technology is seen as extremely challenging both technically and financially. Experts express skepticism, noting Musk’s lack of proven success in scaling complex production processes.
UGREEN iDX6011 Pro Review: High-Performance NAS with Local AI for Home and Professional Users
UGREEN unveils the iDX6011 Pro, a high-end NAS that also functions as a compact workstation and private AI server. Equipped with an Intel Core Ultra 7 255H, 64 GB RAM, and 96 TOPS of local AI performance, the device supports six hard drives and two SSDs, along with 10-Gbit Ethernet and Thunderbolt 4. While AI features showed limitations in testing, the system is priced at €2,599 and scheduled for release in May, targeting premium users.
Microsoft Responds to Criticism: Windows 11 to Receive Major Improvements and AI Adjustments
Microsoft is responding to widespread criticism of Windows 11 with a major quality offensive by the end of 2026, including abandoning mandatory Microsoft account sign-ins, reducing AI integration, and introducing a movable taskbar. The company is also revising its Copilot strategy and halting automatic app installations due to user backlash. Meanwhile, Microsoft is considering legal action against OpenAI and Amazon over a cloud partnership that may violate an exclusivity agreement.
NVIDIA DLSS 5: Revolution or AI Slop? New Technology Sparks Controversial Reactions
NVIDIA unveiled DLSS 5 at GTC 2026, a new technology described as 3D-guided neural rendering that enables fully AI-driven image generation. The system uses 2D color data and motion vectors to create photorealistic lighting and materials in real time, set to launch in autumn 2026 on the new GeForce RTX 5000 graphics cards. While praised for its technical innovation, DLSS 5 has sparked intense criticism for altering developers' artistic vision and creating a uniform, often unexpected visual style.
NVIDIA Unveils Vera-CPU Rack at GTC 2026: New Benchmark for CPU-Only AI Infrastructure
At GTC 2026, NVIDIA unveiled the Vera-CPU Rack, a system designed for CPU-only inference featuring 256 Vera CPUs based on a custom Arm v9.2-A core called Olympus. Each CPU includes 88 cores with SMT, FP8 support, and up to 1.5 TB of RAM. The rack provides 400 TB of LPDDR memory and a total memory bandwidth of 300 TB/s, with individual chips achieving 1.2 TB/s. Optimized for agentic workloads, reinforcement learning, and AI training, the solution is being co-developed with partners like HPE, with the Cray GXC240 supporting up to 640 Vera CPUs per rack.
As global memory prices continue to rise due to supply chain disruptions and high demand from the AI sector, Australian retailers are refusing to replace defective Corsair DDR5 RAM modules, citing financial benefits from price increases. DDR5 memory prices have surged over 400% and are not expected to normalize until 2028, according to analysts. In response, Taiwanese manufacturer V-Color has launched 1+1 DDR5 kits featuring a real memory module and an optical RGB filler module, offering gamers a cost-effective and visually appealing way to upgrade their systems.
SK hynix Unveils 1c LPDDR6 Memory with 16 Gb and 10.7 Gbps – Advancing Mobile AI Technology
SK hynix has unveiled its new 1c LPDDR6 memory with 16 Gb capacity and a data rate of over 10.7 Gbps, based on an advanced manufacturing process that delivers more than 20% improved energy efficiency and is designed for mobile devices with integrated AI acceleration. The chips are set to enter mass production in the first half of 2026. Meanwhile, researchers at the University of Sydney have developed an ultra-compact photonic chip that processes data at the speed of light with high precision, achieving up to 99% classification accuracy in biomedical image tests.
OpenAI has unveiled the new AI model GPT-5.4, which integrates native computer control, programming, and complex logical reasoning into a single system. The model outperforms competitors such as Anthropic Opus 4.6 in desktop control and achieves high scores in benchmarks like OSWorld-Verified and GDPval. With a context window of up to one million tokens and dynamic tool search, it reduces token usage and lowers costs. Meanwhile, Anthropic is challenging the U.S. Pentagon’s supply chain risk designation, while Nintendo has sued the U.S. government over tariffs on Switch consoles.
Intel bets on 18A and 14A: AI-driven CPU demand boosts production and profitability
Intel anticipates strong CPU demand, particularly in data centers driven by the AI boom, and plans to increase production of its new Intel 14A and 18A processors. The company aims to make its Foundry division profitable by 2027 through improved yields, advanced packaging, and growing external customer base. Demand for Panther Lake CPUs already exceeds supply, highlighting the strategic importance of these new technologies.
Windows 12: Microsoft's AI Revolution with CorePC Architecture and Subscriptions
Microsoft is preparing for the release of Windows 12, codenamed 'Hudson Valley Next', which marks a radical shift toward AI technologies and subscription models. The new operating system is built on a modular CorePC architecture and requires at least 40 TOPS on a dedicated NPU, potentially forcing users to upgrade their hardware. Designed as a foundational platform for the future PC market, Windows 12 integrates AI agents and hybrid cloud processing to set new industry standards.
Amazon and OpenAI have announced a strategic partnership worth over $150 billion, involving investments, the use of Trainium chips, and custom AI model development. OpenAI has also signed an agreement with the U.S. Pentagon to deploy its AI models in classified networks, while implementing safeguards against mass surveillance. Meanwhile, Anthropic is advancing AI capabilities with its free Memory feature for Claude and the discovery of 22 security vulnerabilities in Firefox using its Claude Opus 4.6 model. In parallel, OpenAI has postponed the launch of ChatGPT's adult mode, and performance issues with the GeForce RTX 4090 in Resident Evil Requiem have been resolved.
The Trump administration has banned the U.S. military from using AI models from Anthropic, classifying the company as a supply chain risk. OpenAI capitalized on the gap by securing a Pentagon contract for handling sensitive data, sparking criticism over mass surveillance and autonomous weapons. Meanwhile, Caitlin Kalinowski, head of OpenAI’s robotics division, resigned after criticizing the company’s military ties. In a separate incident, a Mexican startup incurred over $82,000 in cloud costs within two days due to a stolen Google Cloud API key, highlighting critical security vulnerabilities.
AMD Ryzen AI 400: New Desktop APUs with Zen 5 and RDNA 3.5 for AM5 Socket Arrive in 2026
AMD has unveiled the Ryzen AI 400 APUs for desktop PCs, featuring Zen 5 architecture and RDNA 3.5 graphics, which will be available in socketed form for the AM5 socket for the first time. The Pro series includes twelve models with TDPs from 35 to 65 watts, based on the Gorgon Point design and equipped with a 50 TOPS NPU. Initially, the processors will only be available in OEM prebuilt systems, not for the DIY market. Additionally, MSI has introduced a new AIO liquid cooling prototype optimized for AMD's AM5 platform.
AI company Anthropic is facing intense pressure after the Pentagon classified it as a supply chain risk and demanded the removal of security filters for military use. The company has filed a lawsuit against the U.S. government, arguing the designation is legally invalid and threatens its business. Meanwhile, competitors like OpenAI have secured military contracts, while studies reveal that most firms have not seen measurable productivity gains from AI despite heavy investments. The conflict highlights growing tensions between AI safety, national security, and economic implications.
Samsung has unveiled the Galaxy S26 series, including the S26, S26+, and S26 Ultra, featuring the Exynos 2600 chip, enhanced camera systems, larger batteries, and a new color palette. The S26 Ultra stands out with a 200 MP main camera sensor, S Pen, and the world’s first built-in Privacy Display using selectively dimmable pixels. The devices offer a more intuitive Galaxy AI experience, improved thermal management, 16 GB RAM, and Gorilla Armor 2 glass. Available starting March 10, 2026, prices range from €999 to €1,499, with limited-time pre-order promotions such as free storage upgrades.
AMD and Meta Announce Billions-Deep AI Deal with 6 GW GPU Infrastructure and Shareholding
AMD and Meta have entered a long-term strategic partnership to build a powerful and energy-efficient AI infrastructure using up to 6 gigawatts of AMD Instinct MI450 GPUs, custom EPYC processors based on the Zen-6 architecture (codenamed Venice and Verano), and Helios racks. The deal, valued at up to 60 billion USD, includes a performance-based warrant for up to 160 million AMD shares and an option to acquire up to 10 percent of AMD's shares. This collaboration strengthens AMD's position in the AI data center market and reflects the shared vision of both companies to build a trillion-dollar AI infrastructure, with deliveries expected to begin in the second half of 2026.
Firefox 148 Launches with Central AI Controls and Kill Switch, Giving Users Greater AI Oversight
Mozilla has released Firefox 148, introducing centralized AI controls and a global kill switch to disable all AI features. Users can now manage individual AI tools such as translations, image alt text, and chatbot providers. The update also brings new features including split view, HDR video streaming on Windows, and a redesigned customizable sidebar, along with security improvements and enhanced profile management.
Meta Strikes Multi-Billion Dollar Deal with Nvidia for AI Chips and Infrastructure
Meta has signed a multi-year, up to $50 billion deal with Nvidia to procure millions of Blackwell and Rubin GPUs, along with additional components such as CPUs, switches, and Vera processors for its AI data centers. The partnership also includes the development of complete rack systems and strengthens Meta's position in the AI race, displacing AMD as the primary supplier. This comprehensive agreement highlights Nvidia's strategic dominance in AI infrastructure.