Data Center News

Google has introduced Cloud Run instances, a new product now in preview. The offering gives developers a single dedicated Cloud Run runtime for workloads that need to stay active for extended periods, rather than scale up and down in response to web traffic. Cloud Run instances are aimed at long-lived, stateful applications such as AI agents that often run continuously for one user at a time.

NVIDIA has expanded NVLink Fusion with a new high-bandwidth memory technology called NVHBM, extending its semi-custom AI infrastructure offering into memory design. NVHBM moves the memory controller from the XPU die into the HBM base die, changing the long-established arrangement in which the controller sits on the processor itself. NVIDIA says that shift can free more silicon area for compute while improving memory throughput and reducing power use.

Amazon Web Services and NVIDIA will deploy 2 million additional graphics processing units for artificial intelligence workloads and expand their collaboration into central processing units, networking and robotics. The move further scales infrastructure tied to surging demand for AI computing, as cloud providers and chipmakers race to add processing capacity for model training and deployment.

Rafay has achieved NVIDIA-Certified Hypervisors status for its Virtual Machines-as-a-Service offering, covering use with NVIDIA HGX systems and NVL72 rack-scale systems. The designation applies to Rafay's VM service within its broader AI infrastructure management platform and is aimed at organisations using virtualisation to run AI workloads on shared GPU systems.

NETSCOUT has added new features to its Arbor Edge Defence product for distributed denial-of-service protection, targeting attacks that pass through or around content delivery networks. The update is designed to help enterprises identify malicious traffic sources hidden behind shared CDN infrastructure and block those attacks without cutting off legitimate users on the same network path.

MSI has begun shipping the XpertStation WS300, a deskside AI system based on NVIDIA DGX Station architecture. It is available through ASI, D&H and Newegg. The system uses NVIDIA's GB300 Grace Blackwell Ultra Desktop Superchip and is aimed at local AI development and deployment inside businesses. MSI says it combines up to 748GB of coherent memory with dual 400GbE networking through NVIDIA ConnectX-8 SuperNIC, allowing two WS300 units to be clustered for larger AI workloads.

Why is the internet fast? And with changing use cases for the internet, will it stay fast? Those questions go straight to both the start and the future of a company providing what it calls a Global Edge Cloud Platform - Fastly. Humans, said Fastly co-founder Simon Wistow, are astonishingly sensitive to latency. "It's a funny thing.

Virtana and Carahsoft have partnered to supply Virtana's Agentic Observability Platform to government agencies. The agreement covers federal, state and local bodies, as well as education, defence and intelligence organisations. As Virtana's public sector distributor, Carahsoft will provide access to the software through its reseller network and established government purchasing frameworks, including SEWP V, ITES-SW2, TIPS, OMNIA Partners, E&I Cooperative Services Contract and The Quilt contracts.

Omnissa will host Omnissa ONE 2026 in Orlando, with a focus on autonomous digital work and trust in AI. Its annual gathering for the end-user computing market will bring together IT professionals, customers and partners as organisations assess how to manage AI in the workplace. Omnissa linked the agenda to concerns about unsanctioned AI adoption on company devices and the broader question of how businesses can oversee software agents that act with greater independence.

The global data centre physical infrastructure market is projected to reach USD $128.27 billion by 2034, up from an estimated USD $36.20 billion in 2025, according to Polaris Market Research. That forecast implies a compound annual growth rate of 15.09% from 2026 to 2034, driven by rising spending on the equipment and systems that support data centres.