Mitch Wagner

AI is dramatically increasing the amount of power data centers must deliver within increasingly dense environments. | AI is pushing data center racks toward megawatt-scale power. See why 800 VDC is gaining attention and how rethinking the power path from the utility grid to the GPU could help meet the demands of next-generation AI infrastructure.

The Kubernetes-based container platform telcos built for cloud-native network functions is now facing new demands — AI inferencing, AI-RAN, data-heavy OSS/BSS and edge services that run in a thousand places, from core to edge. Operators today are making architectural choices they’ll live with for a decade or more.Kubernetes orchestrates workloads — but it’s not enough. It doesn’t deliver the persistent storage, high availability, recovery and portability that carrier-grade operations require. |

ServiceNow said network incident volume fell 90% after it moved its network operations to Juniper Mist | ServiceNow said HPE Juniper Mist AI cut its network incident volume 90%. Its next goal is full autonomy by 2028.

HPE said Aruba Central and Mist will remain separate platforms, with Mist for cloud and Aruba Central | HPE said Aruba Central and Mist will never merge. Instead, the platforms will share AI engines and hardware as HPE fights Cisco for campus share.

HPE's self-driving network detects and fixes 70-80% of potential issues on its own, said networking ch | HPE networking chief Rami Rahim said the network resolves most issues autonomously, but security and compliance changes still need humans.

HPE won a $1.2B order from Vultr for AMD Helios racks with six Juniper scale-up switch trays each

Apelogic CEO Boris Renski compares today's crowded AI agent governance market to cloud infrastructure | Everyone agrees AI agents need guardrails. No one agrees whose. Apelogic's Boris Renski sees cloud history repeating itself.

Ericsson laid out a road map for operators to build AI businesses on 5G before 6G arrives

Orange Business shifted sensitive AI work to open-weight models it runs itself, with sovereignty outwe | Sovereignty concerns are motivating Orange Business to move sensitive AI workloads onto open-weight models it runs itself.

Digital Realty is widening hiring to bring in master electricians, HVAC techs and people with no tech | As AI data center construction outpaces the supply of electricians and technicians, operators are getting creative to fill the gap.

Graphiant CEO Ali Shaikh said carriers are moving fast to take advantage of AI, defying their history | Carriers have a reputation for caution, but Graphiant CEO Ali Shaikh said AI has pushed them to move fast over the past year.

Lightyear launched an agentic AI service that automate telecom quoting and installation for enterprise | Lightyear's AI agents automate quoting and installing enterprise network services.

Grokstream's founders built the AIOps vendor on machine learning after finding rules-based expert syst | Grokstream CEO Casey Kindiger says machine learning, not rules, tames CSP operational complexity.

Palo Alto Networks built NOVA, an agentic AI research system, that found 14,090 unreported vulnerabili | After its AI found 14,090 unreported open-source flaws, Palo Alto is shipping firewall software that protects before a patch exists.

Many neoclouds split GPU orders between Nvidia and AMD at rough parity, VAST Data's John Mao said

'Shabby chic' was the design motif at my hotel. Great elevator, though.

AI infrastructure growth is increasingly constrained by skilled labor, according to executives at Schn | AI data center boom exposes labor crunch

The White House accused Moonshot AI of covertly distilling Anthropic's Fable 5 to build Kimi K3, its 2 | Washington accused China's Moonshot of IP theft, Nvidia rallied 77 companies behind open-weight AI and Anthropic denied seeking a ban. Here's a rundown.

Open models win where data control, cost and vendor leverage matter most, according to executives from | Open-weight AI models give enterprises data control, lower costs and leverage over frontier labs — but Chinese models carry regulatory risk.

AMD launches Helios racks, Instinct MI400 series GPUs and 6th Gen EPYC "Venice" CPUs as one integrated | AMD launched Helios racks, MI400 GPUs and 6th Gen EPYC CPUs, with Anthropic, OpenAI and Meta detailing deployments from the keynote stage.