Network World reports on the most significant news for infrastructure and operations professionals.
ECL FlexGrid platform is targeting AI deployments in locations with limited electricity supply.
Two years after its acquisition by Broadcom, VMware is shedding customers but maximizing value from those who remain, says a CloudBolt survey.
The Koi deal comes one week after Palo Alto closed its acquisition of CyberArk to bolster its AI security portfolio.
The new certification is aimed at bolstering cybersecurity skills and addressing the spike in AI-driven security threats.
The conglomerate's sovereign AI play comes as power constraints and regulatory pressure stall data center growth in the West.
Arista exceeded its AI networking and campus expansion goals in 2025 and delivered revenue of $9 billion, which is up 28.6% compared to 2024.
A new report from Gartner argues that by 2028, misconfigured AI will shut down the national critical infrastructure in a G20 country. Consultants counter that it will likely happen much sooner.
The federal agency has begun issuing CIDs to Microsoft competitors in the business software and cloud computing markets.
Pairing Nvidia's Blackwell GPU with open-source models from Baseten, DeepInfra, Fireworks AI, and Together AI cut costs, vendor says
Edge cloud provider posts strongest quarterly growth in three years as AI-driven traffic patterns create new revenue streams and validate edge economics.
Roles that require using judgment, taking risk ownership, and architectural thinking yield the highest bonuses, according to data from Foote Partners.
But even the CEO of AWS is skeptical of plans to put data centers in space.
Dell’Oro Group estimates that while the top four hyperscalers will be spending big, so too will be the neo cloud service providers.
AI networking demand rose, as did deployments of Cisco's Silicon One architecture.
Intel’s INT31 and Google’s Cloud Security teams identified five vulnerabilities and 35 less critical weaknesses
It depends upon what the meaning of the word ‘ready’ is.
Researchers say the victims are systems with weak SSH password authentication.
Cisco’s expansion of AgenticOps impacts a number of its core data enter, networking and security platforms.
Programs designed to enable load flexibility and demand response in data centers are becoming more critical as constraints on the electric grid intensify.
AI will represent 30% of global WAN traffic by 2034, while IoT applications that span facilities will drive traffic growth, Nokia predicts. Some experts aren’t quite convinced.
Agentic AI capabilities in the NetBrain 12.3 release can autonomously diagnose network issues, suggest fixes, and support IT teams in the remediation process.
FlashSystem.ai enables autonomous AI agents to manage, monitor, diagnose, and remediate enterprise storage issues.
Targeting massive AI clusters, Cisco unveiled the 102.4 Tbps G300 chip along with G300-powered switches designed for hyperscalers, neoclouds, sovereign clouds, service providers, and enterprises.
Updates to the VersaONE Universal SASE Platform include AI-enhanced data protection, AI-guided troubleshooting, and expanded OS support.
As infrastructure automation demand surges, NetBox Labs is delivering an AI agent that enables natural language interaction with infrastructure data and the ability to automate complex operational procedures.
Major tech companies could be asked to absorb power, water, and grid costs tied to rapid AI data center expansion.
New ZAM memory design will compete with HBM used in AI data centers
Active since at least 2019, the China-linked framework operates at network gateways to inspect and manipulate in-transit traffic, allowing attackers to redirect updates, disrupt security tooling, and deliver backdoors.
Capital expenditure and not new features could now become the clearest signal of cloud resilience, analysts say.
There are too many problems for Space X’s vision to get off the ground.
Ruckus is expanding its ICX portfolio with switches designed to streamline management of audio and video content over Ethernet networks.
Two carry CVSS scores of 8.8, demanding that admins upgrade to the latest versions of the controller.
Nvidia CEO Jensen Huang dismisses talk that he is having second thoughts about his massive investment in the ChatGPT maker.
Tech leaders from Nvidia, OpenAI, Microsoft, and Intel examine the many ways AI will change the industry.
A misconfiguration in Microsoft-managed storage accounts triggered cascading failures across virtual machine operations, managed identities, and developer workflows.
Analysts say Intel’s success will hinge less on hardware and more on overcoming entrenched software lock-in and buyer inertia.
At Cisco's second annual AI Summit, Jeetu Patel said 70% of all AI products being developed at Cisco are using code that's generated by AI.
Nvidia plans include use its technology to build AI factories developed and operated by CoreWeave
As attacks on the operating system grow more serious, the company is teasing a plan to bring “verifiable integrity” to Linux.
The new Forward AI capability builds on the vendor's digital twin and is designed to allow network teams to ask complex questions, understand network behavior, validate outcomes and safely automate workflows.
Oracle is considering workforce cuts and selling Cerner to alleviate financial pressure, warns investment bank TD Cowen.
ThousandEyes released its list of the 11 most notable Internet outages and application issues of 2025, along with recommendations to help network teams improve resilience.
Cisco's Mesh Policy Engine supports its own firewalls as well as third-party firewalls from vendors including Palo Alto Networks, Fortinet, and Juniper.
Additions to Netscout's nGeniusONE observability platform extend deep packet inspection capabilities to Wi-Fi 7 and allow organizations to more accurately inventory their deployed SSL/TLS certificates.
The limited approvals can ease immediate pressure on local hyperscalers but leave longer-term constraints in place, with implications for global enterprise AI infrastructure.
US nuclear regulators have secretly relaxed rules to hasten the construction of new generating capacity needed to power AI, NPR reports.
Researchers at Pillar Security say threat actors are accessing unprotected LLMs and MCP endpoints for profit. Here’s how CSOs can lower the risk.
Mplify has updated its certification program to address the performance requirements of AI across complex network environments, rebranding MEF 3.0 and adding a new Carrier Ethernet for AI certification.
AWS staff learned of the cuts early when an internal meeting invitation was posted in error.
AI infrastructure limits are shifting from compute to networking, as fiber capacity becomes critical to data center scale.
