Google has released the Gemini 2.5 Pro Preview ahead of its I/O 2025 conference, enhancing capabilities for coding and interactive web applications. This version addresses developer feedback, improving error reduction and function call efficiency, while excelling in web development aesthetics and video understanding benchmarks.
Meta’s Llama 4 is now generally available as a fully managed API on Vertex AI, offering advanced reasoning, coding, and multimodal capabilities. This release simplifies deployment and management, allowing developers to focus on building AI-powered applications with enhanced efficiency and scalability through Vertex AI’s infrastructure.
AWS AI Labs has launched SWE-PolyBench, an open-source multilingual benchmark for evaluating AI coding agents. This framework addresses limitations of existing benchmarks by incorporating real pull requests across multiple programming languages, enabling more comprehensive assessments of coding agents’ capabilities in real-world scenarios.
Huawei plans to begin shipping its advanced AI chip, the 910C, to Chinese customers soon. This development comes as companies face challenges finding local alternatives to Nvidia’s H20 chip, which is now restricted for export. The 910C offers double the computing power and memory capacity of its predecessor, the 910B.
Dr. Rowland Illing highlights how cloud computing and AI are transforming Parkinson’s disease research and care. With advancements in genetic analysis and brain imaging, these technologies promise faster diagnoses and innovative treatments, offering hope to the millions affected by this progressive neurological condition.
In an interview, Mark Johnstone, Google Cloud’s security lead for the Asia-Pacific region, emphasized the company’s commitment to AI-driven cybersecurity. He highlighted the importance of collaboration with governments and startups to enhance security measures, showcasing innovations like the Gemini AI model and the Secure AI framework to protect users effectively.
AI’s growing energy demand is reshaping power infrastructure, with data centers consuming 415 terawatt-hours globally in 2024. This consumption is increasing 12% annually, outpacing global electricity use. Solutions include more efficient models, renewable energy shifts, and edge computing to manage environmental and economic impacts.
The digital transformation driven by AI is reshaping our energy landscape, with data centers consuming 415 TWh of electricity globally in 2024. This demand is growing rapidly, raising sustainability concerns. As AI’s energy appetite expands, innovative strategies and regulatory frameworks are essential for balancing technological advancement with environmental responsibility.
Google Cloud is preparing for the agentic AI era, unveiling its new Ironwood TPU, which boasts significant power efficiency and scalability. The company introduced various AI models and software, including an Agent Development Kit and an AI Agent Marketplace, aimed at enhancing the development and management of AI agents.
Meta Platforms has launched Llama 4, featuring two new models, Scout and Maverick, designed to handle diverse data formats. Llama 4 Scout runs on a single Nvidia H100 GPU, while Maverick competes with top models in coding tasks. Meta aims to integrate these advancements into its AI assistant across platforms.
