Data Centre Infrastructure News & Trends

GNM launches new PoP in Paris

GNM, a backbone internet provider and telecom operator, has launched a new Point of Presence (PoP) at 137 Boulevard Voltaire in Paris, expanding its Western European footprint and giving French operators direct access to its backbone and interconnection services. The company says that strong growth in France over the past year drove the decision to expand into the capital. The Paris site connects regional operators with lower-latency routes to Frankfurt, Amsterdam, London, and Central Europe, integrated with GNM’s 20,000 km backbone and 80+ PoPs in 21 countries. From Paris, GNM says it offers services including IP Transit with Tier-1 upstreams and advanced BGP community control; GNM-IX peering, with more than 650 connected networks; Ethernet / L2VPN with SLAs and end-to-end traffic control; and Remote IX for access to European Internet Exchanges without physical presence. Ahmed Eidarous, International Development Manager at GNM, comments, “Paris is a natural step in our European expansion. “By launching a PoP here, we are bringing French operators closer to our backbone and service portfolio, enabling them to exchange traffic more efficiently and access international hubs with minimal latency.” With the Paris PoP online, GNM says it has both expanded its presence in Western Europe and strengthened connectivity for regional and global customers. For more from GNM, click here.

Uncover the hidden risks in data centre resilience

In July 2024, a lightning arrester failure in Northern Virginia, USA, triggered a massive 1,500MW load transfer across 70 data centres - handling over 70% of global internet traffic. The result? No customer outages, but a cascade of grid instability and unexpected generator behaviour that exposed critical vulnerabilities in power resilience. Powerside’s latest whitepaper - entitled Data Centre Load Transfer Event – Critical Insights from Power Quality Monitoring - delivers a technical case study from this unprecedented event, revealing: • Why identical voltage disturbances led to vastly different data centre responses • How power quality monitoring helped decode complex grid interactions • What this means for future-proofing infrastructure in 'Data Centre Alley' and beyond Whether you are managing mission-critical infrastructure or advising on grid stability, this is essential reading. You can download the full whitepaper by registering below: [ninja_form id=5]

LINX achieves 25 x 400GE port count milestone

The London Internet Exchange (LINX), an internet exchange point (IXP) operator, has successfully deployed their 25th 400GE port across their global network, a major milestone within the wider industry's growth. The first 400GE port was provisioned in 2021, with demand for ultra-high bandwidth connectivity surging since then, driven by the exponential growth in cloud services, video streaming, gaming, and AI workloads. A single 400GE connection delivers four times the capacity of a standard 1000GE port, enabling networks to consolidate traffic, reduce operational complexity, and build in resilience, future-proofing their infrastructure. Investing in the future LINX has invested heavily in Nokia’s next-generation hardware and optical technologies to enable 400GE delivery across its interconnection ecosystems in London and Manchester. This investment aims to ensure members can scale quickly and efficiently while maintaining the resilience and reliability that underpin the global internet. Jennifer Holmes, CEO of LINX, comments, “Reaching 25 active 400GE ports is a testament to the evolving needs of our members and the strength of our technical infrastructure. "We’ve seen a clear shift towards high-capacity services, with our larger delivery networks upgrading first to 400GE. "In the last 12 months UK ISPs are now seeing the demand for the service and are upgrading in London and Manchester, a positive sign of effective network traffic management or regional peering. "This evolution demonstrates how network operators of all sizes are adapting to keep pace with the demands of modern digital services.” As digital transformation accelerates globally, LINX says it continues to extend its reach beyond the UK, also supporting interconnection platforms in the US and Africa. For more from LINX, click here.

Honeywell, LS Electric deal to boost data centre power

Honeywell, a US multinational specialising in building automation, has announced a global partnership with LS Electric, a South Korean manufacturer of electrical equipment and automation systems, to develop and market integrated hardware and software systems for power management and distribution in data centres and commercial buildings. The collaboration aims to simplify the integration of electrical and automation systems, improving operational efficiency, resilience, and energy management for operators of data-intensive and energy-critical facilities. Integrated approach to power and automation The partnership combines LS Electric’s experience in power infrastructure with Honeywell’s expertise in building automation and control systems. Together, the companies plan to offer systems that unify power distribution and building management, ensuring load and capacity are balanced to maintain resilience and uptime. The first joint products will include integrated switchgear and power management systems that help data centre operators control and distribute power more effectively. Future development will focus on new electrical monitoring systems using the Honeywell Forge platform, enhanced with AI and analytics, and LS Electric’s software capabilities. These systems will be designed to regulate energy distribution, improve efficiency, and provide predictive maintenance to reduce downtime and power quality issues. Energy storage and grid resilience Honeywell and LS Electric also plan to develop a grid- and building-aware battery energy storage system (BESS) for commercial and industrial buildings. The modular BESS will integrate LS Electric’s energy storage technology with Honeywell’s dynamic energy control software, allowing users to forecast and optimise energy sourcing and costs based on grid data, weather conditions, and other variables. The companies said the technology will help manage growing energy demand from sectors such as data centres, which currently account for between 1-2% of global electricity consumption. Billal Hammoud, President and CEO of Honeywell Building Automation, comments, “Our collaboration with LS Electric supports our continued focus on delivering smarter, scalable solutions for the world’s most critical industries. "As the global demand for data and energy accelerates, this partnership combines our complementary strengths to deliver intelligent infrastructure that’s both resilient and efficient.” JongWoo Kim, President of LS Electric Power Electric, adds, “Building on our expertise in power infrastructure and energy storage systems, we are expanding globally into the data centre and industrial building markets. "Through our collaboration with Honeywell, we will provide solutions that help large-scale operators achieve both energy efficiency and reliability.” For more from Honeywell, click here.

The shift from standby to strategic energy management

In this article, Laura Maciosek, Director Key Accounts at Cat Electric Power Division, talks about why shifting backup assets into primary power is becoming essential as grid constraints intensify: It’s safe to say the energy landscape is changing, with many prominent and significant changes having taken place in the last 24 months. The data-driven society we live in, from streaming devices and smart appliances to AI processing, continues to move demand for data centres in just one direction: up. As data centres experience this growth, utility power is no longer a given. Today, there’s no guarantee the local electrical grid can meet these increased power needs. In fact, many utilities I’ve talked with say it’ll be three to five years (or longer) before they can bring the required amount of power online. That puts data centre customers in a tricky position. How can they continue to expand and grow if there isn’t enough power and moving sites isn’t an option? The answer includes rethinking power options, and that means considering the transition from using power assets for largely backup purposes to employing them as a primary power source. That’s a big change from the status quo. If you’re in a similar position, you can read our advice on how to navigate the transition on our blog. Whether you’re ready to make the switch from standby to prime power at your data centre today – or simply weighing options for your next development or expansion – we’re here to help. We’ll work with you to find the right combination of assets and asset management software that fulfils your power requirements reliably and cost-effectively. Connect with one of our experts to get the process started. For more from Caterpillar, click here.

CommScope expands SYSTIMAX Constellation platform

CommScope, a US manufacturer of network infrastructure products, has announced the global availability of its evolved SYSTIMAX Constellation platform, an edge-based system for power and data connectivity designed for large-scale enterprise networks. Now compliant with international power and data transmission standards, the Constellation platform combines fault-managed power with hybrid power and data fibre cabling in a simplified star topology. The design enables hundreds or thousands of connected devices to be powered and linked efficiently across buildings or campuses. Simplified architecture for modern infrastructure The platform’s modular, technology-agnostic design supports both AC and DC power applications and multiple IT and operational technology (OT) standards. The company says unlike traditional structured cabling, Constellation removes the need for a telecoms room or wiring closet on every floor, allowing centralised uninterruptible power supply (UPS) management. Constellation Points (CPs) can be installed in ceilings, walls, or equipment racks, reducing space requirements and component use. The system delivers higher power and faster data speeds over longer distances, supporting 10G and above bandwidth and up to 1kW of fault-managed power. According to CommScope, the platform can reduce installation labour by more than 50% compared with conventional cabling, while also lowering material use and the network’s overall carbon footprint through reductions in copper and plastic. Luc Adriaenssens, Vice President of Building and Campus, CommScope Connectivity Solutions, comments, “The Constellation platform enables our customers to support denser urban environments and the growing number of connected devices in their buildings. "By offering a modular, technology-agnostic solution, we’re providing greater flexibility for network design - allowing scalability, sustainability, and cost efficiency.” Technical and sustainability features Part of CommScope’s SYSTIMAX cabling and connectivity range, the Constellation system includes Propel fibre panels, power transmitters, and transition panels connected to CPs via hybrid fibre trunks. End devices are linked through pre-terminated Cat 6A patch cords and cable assemblies. Key performance and sustainability metrics include: • Up to 1,800W of power delivered at distances of 250m• Up to 1,000W at 500m• Up to 700W at 1,000m• Support for 10G and higher data rates• Up to 50 connected devices per CP• 57% reduction in skilled installation labour hours• 59% less copper and 65% less plastic compared with traditional LAN systems• Up to 40% lower embodied network carbon footprint The Constellation platform is now available worldwide. For more from CommScope, click here.

A-Gas completes large-scale DC refrigerant recovery project

A-Gas, a company specialising in Lifecycle Refrigerant Management (LRM), has completed a major refrigerant recovery project for a global technology provider, marking a significant environmental milestone for the data centre sector. More than 73,000 lbs (33,000 kg) of R410A were safely recovered across five buildings containing 222 cooling units scheduled for decommissioning. The work, carried out under challenging summer conditions, prevented the release of greenhouse gases equivalent to 70,226 tonnes of carbon dioxide (CO₂e). The project was managed by A-Gas Rapid Recovery, the company’s on-site refrigerant recovery division, which specialises in high-speed, compliant recovery operations for commercial and industrial facilities. Environmental and operational impact A-Gas said the recovery operation demonstrated its commitment to safe and environmentally responsible refrigerant lifecycle management. The project not only reduced environmental impact, but also delivered financial benefits to the client through the A-Gas buyback programme. Rapid Recovery’s process is designed to complete complex projects quickly, with recovery speeds up to 10 times faster than conventional methods, helping reduce downtime during critical infrastructure transitions. The operation included full Environmental Protection Agency (EPA) documentation, refrigerant analysis, and regulatory compliance checks throughout. A-Gas said its approach combines global expertise with safety-first practices to help technology and data centre clients meet both operational and sustainability goals. For more from A-Gas, click here.

Building AI-ready networks: Smart cabling for the edge era

In this exclusive article for DCNN, Rachid Ait Ben Ali, Product & Solutions Manager, Smart Building & Data Center at Aginode, explores how next-generation fibre and automated management systems are redefining infrastructure for AI and edge computing: Cabling for the future As artificial intelligence and edge computing rapidly reshape data centre architectures, network infrastructure has to evolve to keep up. Critical applications such as autonomous systems, real-time analytics, and GPU-to-GPU communication for AI model training are highly sensitive to latency, signal degradation, and disruptions caused by excessive hops or amplification points. To meet fast-moving, rigorous requirements, cutting-edge cabling solutions - including ultra-dense fibre panels and the latest generation of Automated Infrastructure Management (AIM) systems - are essential. A closer look at the requirements AI-driven infrastructure is unforgiving of downtime and demands built-in redundancy. Workloads are often unpredictable and heavily overloaded, which means they need optical cabling that can absorb traffic spikes without performance degradation. Architectures supporting AI are non-linear and bandwidth-intensive, requiring designs with optimised physical pathways. Fibre networks in these settings must be capable of supporting transmission rates of 400G, 800G, and beyond. At the same time, edge computing introduces a new level of complexity. Its highly distributed nature requires fibre connectivity that is not only robust and high-performing, but also low maintenance and compact enough to function within physical and power-constrained environments. As AI workloads - from training to inference - generate immense volumes of east-west traffic across dense GPU clusters, managing connectivity and thermal performance becomes critical. These clusters produce extreme heat, making thermally optimised cabling and carefully considered airflow vital. Addressing challenges from a technical perspective Everything starts with the right cabling. Pre-terminated fibre links support rapid and reliable deployment at hyperscale and edge sites. High CPR-rated solutions help ensure compliance with stringent fire safety standards. As organisations plan for upgrades to 800G or even 1.6 Tbps networks, deploying a future-ready cabling plant is essential. Smart labelling systems simplify identification, reduce downtime, and help teams operate more efficiently. Ultra-polished UPC and APC connectors minimise reflection and insertion loss - crucial for extremely latency-sensitive AI applications. High-density fibre-optic cabling, such as MPO/MTP for parallel optics, enables scalable bandwidth in compact footprints. OM5 multimode fibre is well-suited for dense AI clusters and edge deployments. Supporting multiple wavelengths over short distances and offering tighter loss budgets, OM5 delivers high bandwidth without dramatically increasing cable volume. Modern cable designs enhance performance through quality shielding, precise construction, shorter channel lengths, and cleaner signal paths that reduce hops and signal loss. The network architecture itself also plays a key role. Ethernet backbones operating at 400G and 800G, combined with direct-connect models like leaf-spine or fully meshed fabrics, reduce latency and support AI’s massive east-west traffic patterns. Minimising patching and interconnection points cuts down on signal attenuation and interference, further improving efficiency. Innovations such as ultra-dense fibre panels and MPO connectors help scale operations without overloading valuable rack space and enable rapid deployment while avoiding costly rewiring. AIM systems are becoming indispensable, offering real-time visibility into port status, topology, and usage, and integrating with Data Centre Infrastructure Management (DCIM) and orchestration tools. By leveraging AI-powered tracking and analytics, AIM systems enable real-time monitoring and automate management tasks. This reduces mean time to repair (MTTR), minimises human error, and supports large-scale, AI-driven operations. In environments where model training may take months, cabling with embedded diagnostics ensures uninterrupted operation and transparency. Different types of fibre can be tracked, diagnosed, and reconfigured with minimal manual intervention. That’s essential for mission-critical systems. Further considerations Smooth integration of high-performance fibre into AI and edge-ready environments demands careful attention to detail. As GPU-dense racks fill with fibre, insertion and connection loss must be minimised, which increases the importance of ultra-low-loss multimode solutions. Thermal and spatial constraints in AI deployments necessitate slim, compact cabling designs that do not compromise on performance. Adhering to industry standards such as ISO/IEC OM5 and MPO ensures compatibility across hardware ecosystems and simplifies deployment at geographically distributed edge sites. Looking ahead, future-proofing and sustainability are equally important. Investing in OM5 fibre today helps avoid the cost and disruption of replacement as bandwidth demands continue to grow. Automated cabling systems reduce operational expenditures but also enable agile provisioning, even in edge locations where power is constrained. Low-loss fibre reduces the need for signal amplification, conserving energy, while structured, durable panels minimise long-term maintenance waste. To accommodate the demanding requirements of AI and edge computing, data centre design must transition from legacy copper and basic point-to-point fibre models to high-density, automated multimode fibre ecosystems. Smart cabling, ultra-dense panels, and intelligent management solutions form the backbone of this transformation. These technologies empower both hyperscale and edge operations with high bandwidth, low latency, and operational agility. Meeting the needs of AI-ready infrastructure doesn’t just require speed; it demands ultra-high bandwidth, minimal latency, granular insights, simplified manageability, and flexible (re)configuration. With the right infrastructure, it’s possible to build networks that are powerful enough for AI at scale and flexible enough enough for edge deployment.

DE-CIX Malaysia extends reach with NTT DATA partnership

Internet exchange (IX) operator DE-CIX Malaysia has announced the launch of its full suite of interconnection services at NTT DATA’s CBJ1 data centre in Cyberjaya, marking a further expansion of Malaysia’s digital infrastructure and the country’s role as a regional connectivity hub. The partnership represents the first local collaboration between DE-CIX Malaysia and NTT Global Data Centres, following successful projects in other international markets. It aims to enhance interconnection options for enterprises, internet service providers (ISPs), cloud providers, and content delivery networks (CDNs) across Southeast Asia. Expanding interconnection capacity in Cyberjaya The collaboration enables customers at NTT DATA’s Cyberjaya facility to access DE-CIX’s peering, cloud exchange, and virtual private network interconnection (VPNI) services. These include premium peering through GlobePEER, direct cloud connectivity via DirectCLOUD, and Blackholing for DDoS mitigation. Weng-Yew Wong, Member of the Board of Management at DE-CIX Malaysia, comments, “This partnership is a strategic move to bring DE-CIX’s interconnection services closer to Malaysia’s enterprises. "By enabling our platform at NTT Global Data Centres’ Cyberjaya site, we are expanding the region’s digital ecosystem and offering greater reach, security, and reliability.” The collaboration will also allow NTT Global Data Centres to attract additional network-focused tenants, including enterprise and hyperscale operators. Yasuo Suzuki, Managing Director, APAC, NTT Global Data Centres, says, “This collaboration reflects our ongoing efforts to bring strategic partners such as DE-CIX into our facilities to enhance connectivity and performance for our clients. "The presence of DE-CIX in our Cyberjaya campus offers improved access to cloud platforms and digital services with minimal latency.” Strengthening Malaysia’s digital ecosystem Cyberjaya has become one of Malaysia’s leading technology hubs, offering high-capacity power, fibre connectivity, and a growing community of network-based enterprises. DE-CIX’s presence at NTT DATA’s CBJ1 data centre provides immediate access to its global interconnection platform via secure ethernet cross-connects and VLAN tagging, supporting low-latency connectivity to international networks. The companies say the collaboration will contribute to the development of resilient, scalable digital infrastructure across Malaysia and strengthen interconnection routes between Asia, Europe, and the United States. For more from DE-CIX, click here.

GNM launches new PoP in Warsaw

GNM (Global Network Management), a backbone internet provider and telecom operator, has launched a new Point of Presence (PoP) at LIM DC in Warsaw, Poland. The new site joins the company’s existing PoPs at Equinix WA1 in Warsaw and 4DC in Katowice, expanding GNM’s footprint in Central Europe. Together, these sites position Poland as a key hub in GNM’s network, directly connected with Germany, Czechia, Ukraine, and other markets in the region. The Warsaw PoP delivers GNM’s full service portfolio, including IP Transit with Tier-1 upstreams, DWDM up to 400G, peering with over 650 ASNs via GNM-IX, ethernet and L2VPN transport, and Remote IX for access to major European exchanges. Krzysztof Krasuski, Head of Region at GNM, comments, “With the launch of our node in LIM DC, Poland is becoming the heart of our Central European network and we are committed to expanding services and connectivity options for all operators in the region.” GNM’s backbone interconnects more than 80 PoPs in 21 countries over 20,000km of optical infrastructure, supporting resilient, low-latency connectivity for carriers, ISPs, and content providers across Europe. For more from GNM, click here.