AI Infrastructure

NVIDIA released a workflow for converting FP8-quantized CLIP model checkpoints into TensorRT engines, enabling faster inference and higher GPU throughput for production deployment. The process involves exporting the quan…

Anthropic's latest frontier model, Claude Fable 5, is now generally available on Google Cloud. The model is designed for complex, multi-step reasoning tasks including advanced software development and deep multimodal doc…

European AI companies like Mistral AI are pursuing sovereign AI strategies to maintain control over models, data, and compute power, but remain dependent on U.S.-made chips as no European equivalent exists, according to …

Goose, an open source AI agent runtime hosted at the Agentic AI Foundation, released versions 1.36 and 1.37 in quick succession, adding improved Agent Client Protocol (ACP) support, nine new model provider integrations, …

IBM, Supermicro, and Nvidia engineers published an IBM Redbook detailing a reference architecture that uses IBM Storage Scale Erasure Coding Edition on Supermicro servers as an external shared storage tier for Nvidia Dyn…

A March 2026 paper by Vera V. Vishnyakova introduces five criteria for evaluating AI agent context: relevance, sufficiency, isolation, economy, and provenance. The framework treats context as the agent's operating system…

Efinix has launched the Titanium Edge family of FPGAs for edge AI applications, based on soft or hard RISC-V cores with AI instruction extensions. The company's XLR logic cell dynamically reconfigures between logic and r…

Google Cloud announced Gemini for Government, an integrated AI stack designed to help public sector organizations move from isolated AI pilots to full-scale production deployments. The platform includes security features…

Apple announced its new Apple Intelligence system at WWDC, powered by next-generation Apple Foundation Models built in collaboration with Google's Gemini models. The system uses on-device processing and Private Cloud Com…

NVIDIA CEO Jensen Huang told Stanford students that his company achieved a 1,000,000x improvement in computing performance over the past decade through co-design — simultaneously optimizing hardware, software, and networ…

XAI raised billions to build AI models but now rents its Memphis data center to competitors Anthropic and Google, generating $26 billion annually. Apple admitted it cannot run its own AI assistant and will pay Google $1 …

Lindy, an AI agent startup, switched its underlying model from Anthropic to DeepSeek and reports saving millions of dollars in inference costs. The company made the change to address the rising expense of running AI mode…

ClickHouse announced the public beta of ClickHouse Agents, a fully managed agentic analytics service powered by Claude, at Open House 2026 in San Francisco. The no-code service allows users to build and deploy AI agents …

Claw Patrol, a new open-source security firewall for AI agents, intercepts agent traffic at the network level and enforces rules written in HCL before requests reach production systems. The tool can block destructive SQL…

Google Cloud announced the general availability of activity insights within Storage Insights datasets, providing administrators with operational visibility into how storage objects are accessed, moved, and modified. The …

Google Cloud's GKE Inference Gateway delivers up to 92.8% shorter wait times and 62.6% lower inter-token latency compared to the next leading managed Kubernetes service, according to an independent benchmark report. The …

Broadcom reported $22.19 billion in revenue for its fiscal second quarter, a 47.9 percent increase year over year, driven by surging demand for its custom AI chip design and networking components. The company is on track…

OpenAI filed confidentially for an IPO, joining SpaceX and Anthropic in the public-market pipeline, as the AI industry faces growing public skepticism. The filing raises questions about revenue recognition, gross margins…

Chris Lattner, a lead engineer on Apple's original OpenCL implementation, explains why OpenCL and other C++-based GPU programming models failed to become dominant in AI, citing slow committee-driven development and the c…

Zep launched the Batch API for loading large datasets into agent memory, supporting batches up to 50,000 items with a progress dashboard and no impact on real-time ingestion. The API replaces earlier batch methods and is…