This topic is being tracked. New articles will appear here as our AI agents discover them.
Cloud computing is currently experiencing rapid advancements in AI agent infrastructure and large language model deployment. Companies are focusing on providing specialized compute for AI agents, enhancing storage scalability for massive user bases, and developing efficient inference engines. Security concerns are also being addressed with the discovery of exposed cloud and AI credentials.
Answers synthesised from 12 recent sources · updated 12d ago
Google Cloud has launched Agent Compute Engine, a service designed specifically for AI agents. It provides each agent with its own hardware-isolated microVM, offering VM-grade isolation with the speed and density of containers.
OpenAI scaled its online storage infrastructure, named Habitat, to support over 1 billion ChatGPT users. This was achieved through reliable multi-tenancy, global caching, and Azure Cosmos DB, resulting in 12x faster read times for ChatGPT.
An independent researcher discovered 26,400 valid cloud and AI service credentials left in public Gitee repositories. These credentials remained functional, posing a potential risk as they could have been used by anyone.
Liquid AI has launched Liquid RIPL, a real-time inference pipeline engine for its Large Foundation Models (LFMs). RIPL is designed to serve LFMs within a live request path, delivering complete generated responses in under 50 milliseconds.
Kazakhstan's AI ecosystem is progressing from infrastructure planning to active deployment across the entire AI stack. This involves an increase in accelerated computing resources, supporting the nation's AI future.
You.com has launched Inference Router, a managed API for deploying open-weight AI models in production. This service is compatible with OpenAI and Anthropic APIs, allowing developers to integrate existing SDK code by updating the base URL and API key.