AWS unveiled its new Graviton5 Arm-based chips at re:Invent this week, reports TNS Senior Editor for AI, Frederic Lardinois. The new chips feature 192 cores, delivering up to 25% higher performance than the previous generation, demonstrating the company’s sustained commitment to custom silicon. Read Frederic’s full story to find out what AWS announced.

In other technology challenges we’re exploring today:

• AI agents are becoming internal “operating systems” that orchestrate data and workflows, which fundamentally changing the way enterprises function, writes Angie Jones, who is the vice president of Engineering, AI Tools & Enablement at Block. The company engineers built an AI agent framework called “goose,” which uses the Model Context Protocol (MCP) to synthesize data across 60% of the workforce. In today’s contributed piece, Angie explains how goose does this.
• While AI is revolutionized application development by enabling tools like Cursor to write code rapidly, DevOps infrastructure operations remain manual. That’s creating a growing gap in the software delivery pipeline, according to Zafar Abbas, the CTO of DuploCloud. In this sponsored post, Zafar looks at the challenges and opportunities of using AI to automate infrastructure.

For Kubernetes platform engineers or DevSecOps leads, the experience is all too familiar: You open your security dashboard and are greeted by a list of 10,000 deployments, all flagged with critical vulnerabilities, configuration issues and suspicious activities. The sheer volume of alerts creates a paradox: When everything is a priority, nothing is. Traditional risk scoring solutions evaluate the...

Recent AWS, Azure, and Cloudflare outages proved the ugly truth: backup protects data, not uptime. Teams need to know what truly recovers, how long it actually takes, and where hidden failures will make it painfully clear that the RTO was never real.

• Learn how the AI Cloud Services Portal speeds up building and customizing services
• Watch a live demo of launching a GPU service
• Use ready-made components to accelerate time to market