With Trainium4, AWS Will Crank Up Everything But The Clocks

The AI model makers of the world have been waiting for more than a year to get their hands on the Trainium3 XPUs, which have been designed explicitly for both training and inference and which present a credible alternative to Nvidia’s “Blackwell” B200 and B300 GPUs as well as Google’s “Trillium” TPU v6e and “Ironwood” TPU v7p accelerators. …

With Trainium4, AWS Will Crank Up Everything But The Clocks was written by Timothy Prickett Morgan at The Next Platform.

TCG064: Governing AI Agents for Real-World Infrastructure (Sponsored)

In this sponsored episode recorded live at AutoCon 4 in Austin, we sit down with Peter Sprygada, Chief Architect at Itential, to discuss Itential’s on-stage announcement of FlowAI. Peter shares his journey from network engineering skeptic to AI advocate, explaining how Itential securely connects AI agents to infrastructure with enterprise-grade governance and traceability. We dive... Read more »

Cloudflare WAF proactively protects against React vulnerability


Cloudflare has deployed a new protection to address a vulnerability in React Server Components (RSC). All Cloudflare customers are automatically protected, including those on free and paid plans, as long as their React application traffic is proxied through the Cloudflare Web Application Firewall (WAF).

Cloudflare Workers are inherently immune to this exploit. React-based applications and frameworks deployed on Workers are not affected by this vulnerability.

We strongly recommend that customers immediately update their systems to the most recent version of React, despite our WAF being designed to detect and prevent this exploit.

What you need to know

Cloudflare has been alerted by its security partners to a Remote Code Execution (RCE) vulnerability impacting Next.js, React Router, and other React frameworks (security advisory CVE-2025-55182, rated CVSS 10.0). Specifically, React version 19.0, 19.1, and 19.2, and Next.js from version 15 through 16 were found to insecurely deserialize malicious requests, leading to RCE.

In response, Cloudflare has deployed new rules across its network, with the default action set to Block. These new protections are included in both the Cloudflare Free Managed Ruleset (available to all Free customers) and the standard Cloudflare Managed Ruleset (available to all paying Continue reading

Cloudflare’s 2025 Q3 DDoS threat report — including Aisuru, the apex of botnets

Welcome to the 23rd edition of Cloudflare’s Quarterly DDoS Threat Report. This report offers a comprehensive analysis of the evolving threat landscape of Distributed Denial of Service (DDoS) attacks based on data from the Cloudflare network. In this edition, we focus on the third quarter of 2025.

The third quarter of 2025 was overshadowed by the Aisuru botnet with a massive army of an estimated 1–4 million infected hosts globally. Aisuru unleashed hyper-volumetric DDoS attacks routinely exceeding 1 terabit per second (Tbps) and 1 billion packets per second (Bpps). The number of these attacks surged 54% quarter-over-quarter (QoQ), averaging 14 hyper-volumetric attacks daily. The scale was unprecedented, with attacks peaking at 29.7 Tbps and 14.1 Bpps.

Key insights

Other than Aisuru, additional key insights in this report include:

  1. DDoS attack traffic against AI companies surged by as much as 347% MoM in September 2025, as public concern and regulatory review of AI increases. 

  2. Escalating EU-China trade tensions over rare earth minerals and EV tariffs coincide with a significant increase in DDoS attacks against the Mining, Minerals & Metals industry as well as the Automotive industry in 2025 Q3.

  3. Overall, in the third quarter of 2025, Cloudflare’s autonomous Continue reading

Worth Watching: AI/ML Data Center Design

What could be better than watching 0x02 Jeffs discuss networking? How about having Petr Lapukhov of the RFC 7938 fame as a guest discussing AI/ML Data Center Design?

Note: Petr disappeared into the information black hole called Facebook over a decade ago, so I wondered how they allowed him to chat on a podcast for hours. It turns out he moved to NVIDIA, which might influence the podcast content a bit, but I’m pretty sure Petr is still Petr ;)

AI Meets Kubernetes Security: Tigera CEO Reveals What Comes Next for Platform Teams

Kubernetes adoption is growing rapidly, but so are complexity and security risks.

Tigera CEO, Ratan Tipirneni, on Calico AI and the Push for Simpler, Unified Kubernetes Security

Platform teams are tasked with keeping clusters secure and observable while navigating a skills gap. At KubeCon + CloudNativeCon North America, The New Stack spoke with Ratan Tipirneni, President and CEO of Tigera, about the future of Kubernetes security, AI-driven operations, and emerging trends in enterprise networking. The highlights from that discussion are summarized below.

Portions of this article are adapted from a recorded interview between The New Stack’s Heather Joslin and Tigera CEO Ratan Tipirneni. You can watch the full conversation on The New Stack’s YouTube channel. Watch the full interview here

How Can Teams Better Manage the Kubernetes Blast Radius and Skills Gap?

Tipirneni emphasizes the importance of controlling risk in Kubernetes clusters. “You want to be able to microsegment your workloads so that if you do come under an attack, you can actually limit the blast radius,” he says.

Egress traffic is another area of concern. According to Tipirneni, identifying what leaves the cluster is critical for security and compliance. Platform engineers are often navigating complex configurations without decades of Continue reading

AWS, Google Build a Multicloud Bridge

Addressing a long-standing perceived roadblock in enabling systems to span multiple cloud services, Amazon Web Services and Google Cloud have jointly developed a standard for customers to easily bridge their cloud deployments with Layer 3 connectivity. The idea, according to both companies, is to make it easy for their customers with cloud operations in both clouds to network them together in a private network, reducing the burden of maintaining multicloud connectivity, and perhaps even dispelling fears of cloud lock-in. Such easy connectivity may even spur customers to create more multicloud applications, theConnection Coordinator API specification is built on OpenAPI 3.0 customized for easily provisioning dedicated bandwidth between two cloud providers. The two cloud giants want other cloud providers to use the API as well. AWS implemented the spec in Google Cloud’s Cross-Cloud Interconnect. Both companies pledge to “engage in continuous monitoring to proactively detect and resolve issues,” according to the AWS website. The private lines between Google and AWS will be built on

HS119: Securing 2026: How AI, Quantum, and the AI-Powered Browser are Driving Enterprise Defense (Sponsored)

Anand Oswal, Executive Vice President at Palo Alto Networks, joins Johna Johnson and John Burke for a wide-ranging exploration of two emerging focal points of enterprise risk: cryptographically relevant quantum computing, and browser-mediated agentic AI. The looming arrival of quantum computers that can break legacy encryption has already created the threat of “harvest now, decrypt... Read more »

NB554: AWS, Google Link Public Clouds; Trading Data Center Has Zero Chill

Take a Network Break! We start with listener follow-up on Fortinet’s vulnerability numbering, and sound a red alert about an authentication bypass vulnerability in ASUS’s AiCloud service. AWS and Google announce a joint cross-cloud interconnect offering (other cloud providers are invited to play), Microsoft and Ciena pitch a new design to boost optical network resiliency,... Read more »

Multi-Pod EVPN Troubleshooting (Part 3)

Last week, we fixed the mismatched route targets in our sample multi-pod EVPN fabric. With that fixed, every PE device should see every other PE device as a remote VTEP for ingress replication purposes. We got that to work on Site-A (AS 65001), but not on Site-B (AS 65002); let’s see what else is broken.

Note: This is the fifth blog post in the Multi-Pod EVPN series. If you stumbled upon it, start with the design overview and troubleshooting overview posts. More importantly, familiarize yourself with the topology we’ll be using; it’s described in the Multi-Pod EVPN Troubleshooting: Fixing Next Hops.

Ready? Let’s go. Here’s our network topology:

Read more …

Computer Network Design Complexity and Tradeoffs Training

I’m teaching a “one off” special event class over on O’Reilly’s platform (via Pearson) this coming Friday, the 5th of December. From the Description:

Join networking engineer and infrastructure expert Russ White for this exclusive, one-time event exploring the critical role of tradeoffs in network design. We’ll begin by unpacking how complexity shapes the decisions architects and designers must make, and how tradeoffs are often an unavoidable part of navigating that complexity. Through real-world examples, you’ll learn how different network design choices impact overall system complexity, and how to approach these decisions with greater clarity and confidence. We’ll wrap up with an in-depth discussion of unintended consequences—how they arise, how to anticipate them, and how they relate to designing in complex, adaptive environments.

As always, if you register for the course you can watch later.

Register here.

The Road To HPC And AI Profits Is Paved With Good Intentions

With a profitable PC business that has 25 percent of global shipments (thanks in large part to its acquisition of IBM’s PC business two decades ago) plus a respectable smartphone business (by virtue of its Motorola acquisition), the client device business at Lenovo is finally back to where it was during the peak of the coronavirus pandemic and is consistently delivering what are decent profits for this cut-throat part of the IT sector. …

The Road To HPC And AI Profits Is Paved With Good Intentions was written by Timothy Prickett Morgan at The Next Platform.

Tech Bytes: Bringing AI Reasoning to Infrastructure with Itential FlowAI (Sponsored)

Itential has announced FlowAI, a new offering that brings agentic AI to Itential’s network automation platform. On today’s Tech Bytes podcast Ethan Banks talks with Peter Sprygada, Chief Architect at Itential, about how FlowAI works, its components, and how Itential uses the Model Context Protocol (MCP). They also dig into how FlowAI supports AI-driven orchestration... Read more »

netlab as the Universal Configuration Translator

Dan Partelly, a heavy netlab user (and an active contributor), sent me this interesting perspective on how one might want to use netlab without ever building a lab with it. All I added was a bit of AI-assisted editing; my comments are on a grey background.

In all podcasts and interviews I listened to, netlab was referred to as a “lab management solution”. But this is misleading. It’s also a translator, due to its ability to abstract devices, and can easily generate perfectly usable configs for devices or technologies you have never worked on.

Read more …

Why Replicate is joining Cloudflare

We're happy to announce that as of today Replicate is officially part of Cloudflare.

When we started Replicate in 2019, OpenAI had just open sourced GPT-2, and few people outside of the machine learning community paid much attention to AI. But for those of us in the field, it felt like something big was about to happen. Remarkable models were being created in academic labs, but you needed a metaphorical lab coat to be able to run them.

We made it our mission to get research models out of the lab into the hands of developers. We wanted programmers to creatively bend and twist these models into products that the researchers would never have thought of.

We approached this as a tooling problem. Just like tools like Heroku made it possible to run websites without managing web servers, we wanted to build tools for running models without having to understand backpropagation or deal with CUDA errors.

The first tool we built was Cog: a standard packaging format for machine learning models. Then we built Replicate as the platform to run Cog models as API endpoints in the cloud. We abstracted away both the low-level machine learning, and the complicated Continue reading

1 2 3 4 5 6 … 3,832