Google unveils TxGemma, a specialized branch of its Gemma AI family, designed to accelerate drug discovery. These open-source models understand both text and molecular structures, enabling researchers to predict therapeutic properties, identify drug targets, and repurpose existing therapies. This initiative aims to revolutionize the lengthy and expensive process of drug development.
AMD CEO Lisa Su's visit to China highlights the company's aggressive strategy to dominate the emerging AI PC market. Through strategic partnerships, collaborations with leading Chinese AI model developers, and a focus on both hardware and software innovation, AMD is positioning itself as a key player in the future of AI-powered computing.
Meta's Llama AI models reach 1 billion downloads, celebrated for their open-source nature and widespread adoption. Despite this milestone, Meta's stock price fell, raising questions about investor sentiment and the balance between open-source contribution and monetization in the competitive AI landscape. The future of Llama hinges on this balance and strategic revenue generation.
France's Mistral AI is collaborating with Singapore's Ministry of Defence, Defence Science and Technology Agency (DSTA), and DSO National Laboratories (DSO) to develop and deploy generative AI (genAI) solutions. This partnership focuses on refining large language models (LLMs) and creating a specialized mixture-of-experts (MoE) model for enhanced defense capabilities and secure operations.
Nvidia's CEO, Jensen Huang, refuted concerns that DeepSeek's R1 AI model would reduce demand for high-end chips. He stated the model's computational needs are 'much higher.' Nvidia remains confident in its market position, citing strong demand for its Blackwell platform and continued investment in AI infrastructure by major customers, despite custom chip development.
NVIDIA announced the Blackwell Ultra GB300 and Vera Rubin superchips at GTC 2025. Blackwell Ultra offers 1.5x FP4 inference performance over H100, and significantly faster AI reasoning. Vera Rubin, arriving in late 2026, features a custom CPU and GPU, achieving 50 petaFLOPS of FP4 inference performance per chip, pushing AI capabilities further.
From a modest academic gathering to the epicenter of the AI world, Nvidia's GTC reflects the company's meteoric rise. Jensen Huang's vision and the relentless pursuit of AI innovation have transformed GTC into a must-attend event, showcasing cutting-edge technologies, fostering collaboration, and shaping the future of artificial intelligence across diverse industries.
Explore OpenAI's Sora with five prompts designed to ignite your AI filmmaking. From intense samurai duels to tranquil coffee rituals and dynamic cityscapes, discover how to craft compelling video content using text, and learn techniques like 'Remix' and storyboarding to enhance your creations and unleash Sora's full cinematic potential for your projects.
STORY, a blockchain-based IP protocol, integrates Anthropic's AI to revolutionize intellectual property management. This partnership aims to streamline registration, licensing, and revenue distribution, empowering creators and fostering innovation. The fusion of blockchain and AI promises a more equitable and efficient future for the creative economy, with Anthropic's valuation soaring to $90 trillion.
Tencent Cloud is investing over $650 million to build data centers in Saudi Arabia and Indonesia, marking a major expansion into the Middle East and Southeast Asia. This strategic move aims to meet the growing demand for cloud services, empower local businesses, and solidify Tencent's global presence in the competitive cloud computing market.
