Flex (NASDAQ: FLEX) said this week it has reached a definitive agreement to buy EPC Power for $4.4 billion, folding in differentiated power-conversion and grid-forming technology built for the next generation of 800V data center power architectures. Flex expects the deal to close in the fourth quarter of calendar 2026,… Read More »Flex Acquires EPC Power: A $4.4B Bet on 800V AI Data Center Power
This morning, Arm positioned the new Mali G2-Ultra NX GPU as its first AI-native Mali graphics core, integrating dedicated neural accelerators directly into the shader pipeline rather than bolting an NPU alongside it. It arrives with a third-generation ray tracing unit and a new execution engine Arm calls the largest… Read More »Arm’s Mali G2-Ultra NX: Neural Accelerators Take Mobile Graphics AI-Native
Nokia today announced Mobile Core Early Access, a live, hosted environment that lets telecommunications providers and enterprises evaluate the software behind AI-era mobile core networks before committing to deployment. The program launched in June with a select group of 30 companies; with today’s announcement, registration is now open to the… Read More »Nokia Opens Its Mobile Core Early Access Program to the Whole Industry
by Chris DePuy, 650 Group / August 30, 2026 On August 11, 2026, xAI introduced Grok Bot, an early-beta product it frames as a roster of always-on AI agents that each run on a cloud computer of their own, sign into the customer’s existing tools, and carry multi-step jobs through… Read More »Grok Bot: Always-On Agent Compute and the Infrastructure It Demands
IntroductionGPU performance from 2025 through the end of the decade is growing exponentially, but the more fundamental driver of infrastructure change is what is happening to the models themselves. The shift from dense transformer architectures to Mixture of Experts designs, now dominant across frontier model development, is transforming not just… Read More »Mixture of Experts and Disaggregated Architectures Are Rewriting the Rules of Network Infrastructure
In recent weeks, Cato Networks made clear it is deepening its position in AI security while building on its SASE foundation. The company launched Agentic Threat Prevention, disclosed two critical vulnerabilities in the Cursor IDE through its Cato AI Labs, and reported that annualized recurring revenue has now passed $415… Read More »Cato Networks Pivots to Agentic AI Security as ARR Passes $415M
Celona today announced Orion, a new enterprise platform that unifies private 5G, Wi-Fi, public cellular, and satellite into a single wireless fabric designed for AI-native operations. Celona is positioning Orion as its Agentic Converged Wireless Platform, creating a new category that brings together previously separate wireless domains under a unified… Read More »Celona Launches Orion, an Agentic Converged Wireless Platform
Networking now determines how effectively memory reaches GPUs Introduction Memory has become a critical constraint in AI infrastructure. As models scale, solutions involve a mixture of experts, and context windows expand, the way memory is used must change. If we don’t change, GPUs stall waiting for data rather than computing.… Read More »Marvell Expands AI Memory and Storage Infrastructure with Bravera SC6
At Black Hat USA 2026, Cribl unwrapped a set of AI-era security capabilities for its existing platform. The headline is the AI Observability app, which puts token counts, spending, model uptake, and risk into a single view across teams and applications and flags sensitive data as it enters prompts and… Read More »Cribl Announces AI Observability at Black Hat
IntroductionAI clusters deliver value only when GPUs stay productive. Too many installations still run below 30-50% utilization because networks cannot keep up with job dynamics, tenant sprawl, and constant workload movement. While that number improves with time, those early idle times can quickly move to millions of dollars of waste… Read More »Full Stack Security and AI Job-ID-aware Segmentation: Cisco Raises the Bar for AI Fabric Efficiency
