Network World reports on the most significant news for infrastructure and operations professionals.
AWS is taking a step towards simplifying multicloud operations with Google; it plans to add Azure to the mix next year.
Most companies expect greater productivity and efficiency from AI, but that hasn't prevented nearly 80% from reverting to human-centric processes after their AI initiatives yielded disappointing results.
Despite the new feature, AWS’s US East region in Northern Virginia will remain at risk, an analyst says.
The three-time No. 1 system maintains its title ahead of Frontier and Aurora in the latest TOP500 list of the world’s most powerful supercomputers.
The company’s restructuring and memory cost warnings signal challenges ahead for enterprise buyers on pricing and service delivery.
Their exits may strain Microsoft’s push to expand AI capacity as energy constraints and infrastructure bottlenecks increasingly dictate the pace of cloud growth.
Flaws in Fluent Bit could let attackers inject fake logs, reroute telemetry, and execute arbitrary code across cloud platforms.
Aria Networks is targeting AI scale-out with a microsecond telemetry approach.
Security leaders call the rollback ‘shockingly incompetent,’ warning it removes critical controls implemented after one of the largest breaches in telecom history.
Gluware's Titan platform coordinates multiple AI agents — from observability platforms, service management tools and security systems — that want to make simultaneous network modifications.
Attackers are moving beyond on-prem systems and now using AWS’s own encryption and key management features to lock organizations out of their cloud data.
There may be better options for enterprises struggling to access enough Nvidia chips to deliver on their AI strategy.
Cisco's Resilient Infrastructure initiative aims to bolster network security by strengthening default device settings and removing insecure capabilities.
Cisco and IBM hope to build a quantum computing internet by the late 2030s.
Counterpoint warns that DDR5 RDIMM costs may surge 100% amid manufacturers’ pivot to AI chips and Nvidia’s memory-intensive AI server platforms, leaving enterprises with limited procurement leverage.
With 132 cores, expanded cache, and 3nm manufacturing, the new chip is designed to consolidate enterprise workloads and deliver up to 40% cost savings.
Broadcom research shows enterprises are accelerating AI adoption, but visibility gaps and network congestion are challenging network operations teams.
The problem is that without Microsoft input, SAP systems will soon not deliver MS-level performance. But if an emergency only lasts a few weeks, analysts said, the plan might be viable.
NaaS startup Meter unveiled new wireless access points, switches, firewalls, and cellular gateways, all of which run identical firmware.
The new Dense Wavelength Division Multiplexing (DWDM) device increases bandwidth by 73% and reduces power consumption by 35%, according to AWS.
A simple database permission change triggered a global failure, highlighting how self-inflicted software errors now cause more disruptions than cyberattacks.
Microsoft’s partnership with Nvidia and Anthropic continues the incestuous couplings that may force changes in how IT makes AI choices.
Dell AI Factory with Nvidia includes PowerEdge servers featuring Nvidia Blackwell and Hopper GPUs.
Kentik launches agentic AI for autonomous network troubleshooting.
IBM expects undisputed quantum advantage next year. Last week, it announced its new Nighthawk quantum computer with 5,000 two-qubit gates and improved error correction.
Analysts say enterprises face a trade-off: reduced operational toil versus increased administrative overhead and governance requirements.
Problems with Cloudflare’s content delivery network followed hot on the heels of an incident with its third-party support portal provider.
VMware’s shift to real-world telemetry data reveals a significant gap between recommended and actual resource usage.
Microsoft says the attack, sourced from more than 500,000 compromised IPs, exposes deep weaknesses in home IoT and raises questions about enterprise DDoS readiness.
Nvidia says it rolled out a combined total of 4,500 exaflops of AI performance for scientific discovery.
Nvidia expands NVQLink quantum interlink as it preps for future quantum processor connectivity
Opening the NVLink ecosystem is a ‘masterstroke’ that leaves Nvidia firmly in control, says analyst.
Germany’s JUPITER Booster marks Nvidia’s first 1-exaflop system on the TOP500 List, earning the fourth spot.
The world’s largest memory chipmaker’s aggressive pricing threatens to push AI server costs up 10-25% as supply constraints remain through 2026.
Microsoft is investing heavily in quantum computing, aiming to bring the world’s most powerful quantum computer online by 2026 to advance security, research, and next-gen technology.
Supercomputer customers have a choice of either company’s processors or a mix of both.
StarlingX cranks open-source cloud network infrastructure to 11.
Unified solution leverages Arista’s AI networking fabric and Palo Alto's next generation firewalls for zero-trust security in the data center.
New AMD CPUs and GPUs are coming in 2026 that will feature improved performance
Extreme envisions offering specialized, easy-to-deploy AI agents for analytics, sustainability and security.
AI, campus, router and security refreshes led to a strong quarter for Cisco.
The year isn't over yet, but we've already seen record-breaking quantum computers, skyrocketing levels of investment, and demonstrations of real-world benefits.
Broadcom unveils a unified networking fabric and integration with Cisco to enhance interoperability and simplify private cloud operations with VMware Cloud Foundation.
Cast AI report says managing cloud GPU resources requires operational and geographic agility.
The tool could help enterprises gain future visibility into planned expansion of services across cloud regions, helping avoid costly replanning and deployment.
Twice the price-performance of comparable modern x86 VMs is Google’s claim for its new N4A instances, now available in preview.
Businesses already writing for TensorFlow, or building from scratch, stand to benefit most, while enterprises with legacy code written for Nvidia’s CUDA may find moving costly.
The latest Cisco AI certifications are targeted at IT pros who are doing the work to deploy, manage and scale AI workloads across enterprise environments.
TransferEngine enables GPU-to-GPU communication across AWS and Nvidia hardware, allowing trillion-parameter models to run on older systems.
Corporations are in a race to analyze data from above the clouds.
