Data Centre Infrastructure News & Trends

Huber+Suhner expands Microsoft Azure fibre collaboration

Huber+Suhner, a Swiss fibre optic cable manufacturer, has strengthened its collaboration with Microsoft Azure to support the wider deployment of hollow core fibre (HCF) connectivity across the Azure network. The company plans further investment in production capabilities to increase manufacturing volumes as Microsoft expands the use of HCF across additional Azure regions. The collaboration is focused on supporting cloud and AI infrastructure requirements. Huber+Suhner has worked with Microsoft’s Azure fibre team in Romsey, UK, since 2017, following the acquisition of Lumenisity, a University of Southampton spin-out. Together, the organisations have developed HCF cable and connector technologies which are already deployed within the Azure network. Higher-capacity variants are also in development to support future infrastructure growth. The two companies have jointly developed and qualified a range of outside plant (OSP) and inside plant (ISP) cable designs for field deployment. Work is also ongoing to develop higher-density HCF cable designs for future network requirements. At Huber+Suhner’s manufacturing facility in Herisau, Switzerland, dedicated processes have been introduced to integrate HCF into multi-fibre loose-tube cables, with scope to increase capacity as demand grows. Connector development supporting HCF deployment Alongside cable development, Huber+Suhner has developed a mode-converting HCF connector designed for hyperscale and metro optical environments. These connectors are manufactured at the company’s Cube Optics facility in Mainz, Germany, with further investment planned to expand production capacity. With both HCF cable and connector designs qualified, Huber+Suhner says it is extending its portfolio to support end-to-end fibre connectivity across cloud infrastructure. Jürgen Walter, COO Communication Segment at Huber+Suhner, comments, “Huber+Suhner is proud to support Microsoft as HCF connectivity solutions move to deployment at scale. "Building on our foundations of innovation and quality, we can expect further advances in our HCF connectivity portfolio as the pace of adoption accelerates. Together, we look forward to shaping the future of cloud connectivity and unlocking the full potential of HCF.” Colin Wallace, GM Cloud Network Engineering at Microsoft Azure, adds, “We value our long-standing collaboration with Huber+Suhner, which has helped us transition HCF technology from advanced research into operational deployment in the Microsoft Azure network. "These HCF cable and connector technologies are already deployed and carrying live traffic over Azure HCF links today, and this integrated capability will help us rapidly co-design and scale connectivity solutions for the future of cloud and AI network infrastructure.” The relevance of HCF HCF technology enables data to be transmitted through air rather than glass, allowing for significantly lower latency in optical networks. Microsoft’s Double-Nested Anti-Resonant Nodeless Fibre design also supports lower signal loss and higher launch powers compared to standard single-mode fibre, reducing the need for optical amplification in some metro networks. The use of HCF in data centre environments is expected to support greater flexibility in site location, as well as improved efficiency in distributed AI workloads by reducing latency between compute clusters. However, wider deployment presents technical challenges, including the need for robust cable designs and compatible termination methods. Huber+Suhner says its HCF connectors are designed to interface with standard single-mode fibre systems while protecting the hollow core structure and maintaining performance in operational environments. For more from Huber+Suhner, click here.

STL launches Neuralis US data centre platform

STL, an optical and digital systems company, has launched its Neuralis data centre connectivity portfolio in the United States, targeting infrastructure designed for artificial intelligence and high-density computing environments. The announcement was made by STL Optical Connectivity NA, the company’s US subsidiary, at Data Center World 2026 in Washington, D.C. Neuralis is designed to support evolving data centre requirements, particularly the shift towards AI workloads, hyperscale computing, and edge deployments. These trends are increasing demand for high-speed, high-density connectivity within and between facilities. The portfolio focuses on managing the transition from traditional north–south traffic flows to more intensive east–west traffic, driven by GPU-based architectures and AI training processes. Designed for high-density AI infrastructure The Neuralis portfolio is structured around two main areas: The first focuses on maximising data centre space through the use of high-density, pre-terminated fibre cabling. This approach moves connection work into manufacturing environments, reducing on-site installation time and complexity. The second area addresses data centre interconnect (DCI), supporting large-scale data transfer between sites. This includes fibre infrastructure designed for high-capacity environments, with cables capable of supporting large fibre counts for AI deployments. STL has developed the portfolio through collaboration with customers, with a focus on addressing space, density, and deployment challenges in modern data centres. The company’s manufacturing process covers the full fibre lifecycle, including preform production, fibre drawing, cabling, and connector integration. Production for the US market is supported by STL’s facility in Lugoff, South Carolina. Ankit Agarwal, Managing Director of STL, notes, "AI demands a level of precision and density that traditional cabling simply cannot meet. "With STL Neuralis, we are providing the high-speed, low-latency foundation that allows GPU clusters to perform at their peak, moving complexity out of the field and into a controlled, high-precision factory environment." The launch reflects increasing demand for infrastructure capable of supporting AI-driven workloads, as operators continue to scale data centre capacity across North America. For more from STL, click here.

DE-CIX, Ooredoo link Doha IX to Marseille

Internet exchange (IX) operator DE-CIX and Qatari telecommunications company Ooredoo have connected Doha IX to DE-CIX Marseille, expanding international interconnection for networks in Qatar. The link connects Qatar’s first commercial internet exchange with a wider European ecosystem, enabling direct access to networks in Marseille and remote connectivity to those linked via DE-CIX Frankfurt. Doha IX is operated by Ooredoo under the DE-CIX-as-a-Service model and is hosted in one of the company’s data centres. The interconnection is intended to improve access to cloud platforms and digital services not currently available locally. The connection allows networks in Qatar to exchange data directly with almost 120 networks in Marseille, as well as access a broader pool of networks connected through Frankfurt, one of Europe’s largest internet exchanges. This supports lower-latency connectivity and provides additional resilience for cloud and content delivery. It also enables access to major cloud providers through dedicated and private connections, alongside tools designed to support hybrid and multi-cloud environments. Expanding low-latency access to global networks Since its launch in October, Doha IX has developed as a carrier-neutral interconnection hub, supporting local and international data exchange. The platform also offers services including cloud connectivity, IP transit, hosting, and colocation. Ivo Ivanov, CEO of DE-CIX, says, “The direct interconnection between the IXs in Doha and Marseille brings the world closer together. “By providing even better performance and user experience for internet-based content and applications, our collaboration with Ooredoo opens up new opportunities for Qatar’s digital economy. "Enhanced connectivity will further strengthen the digital ecosystem in the GCC, supporting economic growth and innovation while paving the way for the amazing digital decades ahead of us.” Hassan Ismail Al Emadi, Chief Business Officer at Ooredoo Qatar, adds, “The direct interconnection between Doha IX and DE-CIX Marseille represents a strategic expansion of Qatar’s global digital reach. “By linking our national interconnection platform with one of Europe’s leading internet exchange ecosystems, we are enabling differentiated digital performance through lower latency, enhanced resilience, and secure, seamless access to global cloud and content networks. "This collaboration reinforces Qatar’s position as a regional digital gateway and enables enterprises to operate with greater performance, reach, and competitiveness, accelerating digital transformation across Qatar and the wider GCC.” The companies say the development reflects continued investment in interconnection infrastructure to support growing demand for cloud services and international data exchange. For more from DE-CIX, click here.

Carrier opens €12m Montluel HVAC testing facility

Carrier, a manufacturer of HVAC, refrigeration, and fire and security equipment, has opened a new testing facility at its European Centre of Excellence in Montluel, France, to support the development of cooling and heating technologies for data centres, industry, and large commercial buildings. The €12 million (£10.4 million) investment expands the company’s research and development capacity, with a focus on high-performance systems aligned with electrification trends and the use of lower-impact refrigerants. Testing at the site follows Eurovent-certified performance methodologies. The expansion comes as demand for data centre infrastructure continues to grow across Europe. According to JLL’s 2026 Global Data Center Outlook, the EMEA region is expected to add 13GW of new capacity by 2030, driven by hyperscale deployments and artificial intelligence workloads, particularly in markets such as London, Frankfurt, and Paris. Increased capacity for HVAC system testing The new laboratory is designed to support testing across a wide range of operating conditions. It enables evaluation of air-cooled chillers up to 3,200kW, air-source heat pumps up to 1,500kW, and water-source systems up to 6,000kW. The facility can simulate temperatures ranging from −20°C to +60°C, with humidity control, and supports water flow rates of up to 1,600m³/h. This allows for testing under varied and demanding conditions relevant to real-world applications. Bertrand Rotagnon, Executive Director, Commercial Business Line and Data Centres Europe at Carrier, comments, “With these new test laboratory facilities, we’re raising the bar on how we support customers and partners in Europe. “The combination of higher test capacity and advanced environmental control lets us validate performance with zero tolerance, earlier, and bring solutions to market faster, giving customers the confidence to move ahead on high-efficiency cooling and heating for data centres, industry, and district heating.” Nicolas Fonte, Director, Systems Engineering at Carrier Climate Solutions Europe, adds, “The new testing facility expands our engineering team's ability to test and validate chillers and heat pumps for very wide and [the] most critical operating conditions. “This new equipment enables us to validate performance, with high precision, of next-generation chillers and large heat pump platforms supporting [increasing] customers' requests for future infrastructures.” The development forms part of the company's stated ongoing investment in HVAC technologies to meet increasing performance, efficiency, and regulatory requirements across European markets. For more from Carrier, click here.

EPRI, OCP aim to advance DCs as flexible grid resources

EPRI (the Electric Power Research Institute), an independent, non-profit energy research and development organisation, and the Open Compute Project (OCP), a non-profit organisation that develops and shares open hardware standards and designs for data centre infrastructure, have announced a collaboration focused on developing data centres as flexible resources for power systems. The initiative aims to support digital infrastructure growth while improving how data centres interact with electricity networks, particularly as demand increases from artificial intelligence and other compute-intensive workloads. By working together, the organisations intend to support improved integration between data centres and power systems while developing technical frameworks to enable more flexible operation. Arshad Mansoor, President and CEO of EPRI, comments, “We’re in the midst of an energy revolution, and it must be smart, flexible, and innovative to keep rates affordable for customers across the globe. “Through this collaboration with OCP, EPRI is combining rigorous power system science with open, scalable data centre innovation to advance practical solutions that enable data centres to operate as flexible, grid-supporting resources - strengthening reliability and affordability for all.” Developing flexible data centre energy models The collaboration brings together stakeholders across the energy and data centre sectors, including a European group involving DCFlex, National Grid, NESO, PPC, RTE, and RWE. This group is working to develop frameworks that reflect operational requirements, with a focus on improving resilience and scalability as data centre capacity expands. Activities include work on shared standards, testing environments, and implementation guidance for flexible data centre operations. Zane Ball, Chief Technology Officer at OCP, notes, “With a growing member base and top-tier data centre expertise coming together with a single vision, our collaboration creates opportunities for harmonised standards, shared testing environments, and coordinated guidance for implementing flexible, resilient, and affordable data centre solutions.” EPRI says it is also supporting the work through field demonstrations at data centres in Europe and the United States, exploring flexible load approaches that could support grid stability and reduce barriers to connection.

LS Electric wins $115m data centre contract

LS Electric, a South Korean manufacturer of electrical equipment and automation systems, has secured a $115 million (£84.9 million) contract to supply power infrastructure for a series of data centre developments across North America. The projects will support major technology companies expanding capacity for artificial intelligence and other compute-intensive applications, where consistent and high-quality power is required. Under the agreement, LS Electric will deliver switchgear and distribution transformers designed for continuous operation in high-demand environments. Expanding North American manufacturing footprint The deal comes at a time as data centre operators are increasing focus on power systems that offer reliability, adaptability, and long-term support as facilities scale to meet rising workloads. Large-scale developments of this kind also require suppliers able to meet strict technical standards while maintaining consistent delivery across manufacturing, logistics, and on-site coordination. LS Electric says it will support the projects from design through to commissioning. To fulfil the contract, LS Electric will utilise its growing industrial presence in North America, including operations in Utah and Texas, such as MCM Engineering II and its Bastrop campus. These facilities will support production and system integration, as well as ongoing regional expansion in engineered power infrastructure. LS Electric states it will continue to expand its offering for the sector, focusing on technologies that support reliable and energy-efficient data centre performance. For more from LS Electric, click here.

Mission Critical Group invests in WattEV

Mission Critical Group (MCG), a critical power infrastructure company, has announced a strategic investment in WattEV to support the development of 800V DC power infrastructure for AI data centres. The partnership focuses on advancing power delivery systems designed to meet the increasing demands of high-density AI workloads, including generative AI and inference applications. As part of the agreement, Mission Critical Group will support the industrialisation and deployment of a medium-voltage solid-state transformer (SST) platform. This technology is intended to enable the transition to 800V DC architectures within large-scale data centre environments. The companies state that traditional AC-based power systems are facing limitations as AI workloads scale, driving interest in alternative approaches to power distribution. The proposed 800V DC architecture enables direct conversion from medium-voltage AC, with the aim of improving efficiency and reducing system complexity. The modular design is intended to support flexible deployment, faster installation, and easier expansion. High-density power delivery Jeff Drees, CEO of Mission Critical Group, says, “We are building the next evolution in modular power delivery. The investment in WattEV highlights our commitment to advancing solutions for ultra-high-density AI workloads, including generative AI and inference.” Michael Maiello, SVP of Innovation at Mission Critical Group, adds, “We are moving beyond incremental improvements to a fundamentally different power architecture. "By converting the ultra-high-power demands of AI directly from medium-voltage AC to 800 VDC, we unlock the full efficiency and performance benefits of 800 VDC distribution.” Salim Youssefzadeh, CEO of WattEV, concludes, “Our technology is already proven in high-power, real-world applications where efficiency and reliability are critical. Together with MCG, we’re bringing that performance into the data centre to accelerate the adoption of 800 VDC architectures with confidence and speed.” The companies state that the collaboration aims to support the deployment of scalable power infrastructure for next-generation AI data centres. For more from Mission Critical Group, click here.

Carrier launches AquaEdge chiller

Carrier, a manufacturer of HVAC, refrigeration, and fire and security equipment, has introduced the AquaEdge 19MV4 centrifugal chiller, designed to support cooling requirements in high-density AI data centres. The system forms part of the company’s QuantumLeap portfolio and is intended for use in environments where increasing compute density and rising temperatures place pressure on existing cooling infrastructure. The chiller is designed to deliver between 2.1 MW and 3.3 MW of cooling capacity, supporting workloads driven by high-performance GPUs. It is also engineered to operate with chilled-water temperatures of up to 35°C and condensing temperatures up to 55°C, aligning with liquid cooling approaches such as direct-to-chip and rear-door heat exchangers. Designed for high-density cooling environments Carrier states that the system uses a variable-speed centrifugal compressor capable of operating between 10% and 100% load, allowing it to respond to fluctuating AI workloads without frequent cycling. Marti Urpinas, Senior Technical Manager, Vertical Markets EMEA, DC Applied at Carrier, comments, “AI workloads are reshaping data centre specifications, pushing our customers to seek greater thermal headroom without sacrificing power stability. "That sounds like a tall order, but the AquaEdge 19MV4 isn’t a ‘standard’ chiller; it’s a variable-speed centrifugal platform that delivers cooling continuity for high-density racks, even as operators push chilled-water temperatures higher to support direct-to-chip architectures.” The unit is designed to restart within 150 seconds following a power interruption, supporting thermal recovery and reducing the risk of overheating in high-density environments. It also incorporates harmonic filtering to limit electrical distortion and protect associated infrastructure, including uninterruptible power supplies (UPS). Carrier reports that the system can achieve a coefficient of performance (COP) of up to 6.75 and an integrated part load value (IPLV) of 11.4 under AHRI test conditions. The chiller is available with refrigerants including R-1234ze and R-515B, supporting compliance with EU F-Gas regulations. Additionally, noise levels are specified at below 80dBA under defined operating conditions. For more from Carrier, click here.

ZIEHL-ABEGG highlights ZAbluefin fan

German ventilation manufacturer ZIEHL-ABEGG has outlined the performance characteristics of its ZAbluefin centrifugal fan, designed for HVAC and air handling unit applications. The fan uses a biomimetic blade design, including a corrugated leading edge and twisted geometry, to improve airflow efficiency. A serrated trailing edge is intended to reduce turbulence and noise while maintaining stable performance under varying airflow conditions. According to the company, the design supports energy efficiency at typical operating points, particularly in environments where airflow may be disrupted. Focus on efficiency and low-noise operation The ZAbluefin fan is designed to reduce sound output, with a focus on minimising tonal noise, making it suitable for noise-sensitive environments. Its performance curve allows for a wider operating range without flow separation, enabling system designers to meet different requirements without oversizing equipment. The fan is also intended to support compliance with current and future efficiency regulations. The product range covers diameters from 250mm to 1,120mm, with airflow capability of up to around 90,000m³/h and static pressure up to approximately 2,500Pa. This allows use across both compact and large-scale HVAC systems. ZIEHL-ABEGG has also developed a one-piece mounting system to support installation. The mount is designed for multiple orientations, including horizontal and vertical configurations, and is intended to simplify installation and reduce component variation. The company states that the combined fan and mounting design aims to improve efficiency, reduce noise, and simplify deployment across a range of HVAC applications. For more from ZIEHL-ABEGG, click here.

Nokia recognised in data centre switching report

Finnish telecommunications company Nokia has been named a Leader and Outperformer in the 2026 data centre switching Radar Report by GigaOm, marking the fifth consecutive year it has received this recognition. The annual report evaluates data centre switching platforms based on technical capabilities, product development plans, and business criteria. In the 2026 edition, Nokia’s Data Center Fabric is positioned in the Innovation/Platform Play category and classified as an Outperformer. GigaOm assessed 10 vendors in the market, comparing their offerings against a range of operational and technical benchmarks. Nokia’s platform received five-star ratings across all key feature areas reviewed, including hardware, switching and routing functionality, operational management, NetDevOps alignment, and traffic security. Focus on automation and AI-ready networking The report also highlights Nokia’s capabilities in areas such as AI-focused networking features and its microservices-based network operating system architecture. Andrew Green, Analyst at GigaOm, says, “GigaOm’s analysis highlights Nokia’s consistent innovation in data centre switching and its strong feature delivery over the past year. "Its Data Center Fabric stands out for the depth of its hardware and software capabilities and strong support for automation that are designed to address the requirements of modern AI-driven data centre environments.” The recognition reflects broader trends in the data centre sector, where organisations are updating infrastructure to support data-intensive workloads and distributed cloud environments. Michael Bushong, Vice President Data Center at Nokia, comments, “Being named a GigaOM data centre switching Leader and Outperformer for the fifth year in a row is validation of two things: "First, we are doing the hard work of providing highly reliable data centre solutions for cloud, enterprise, and service provider customers tasked with building in this AI era. "And second, we are keeping pace with all the innovation being driven by AI. Openness, automation, scale, and reliability are more than buzzwords; they require constant care and feeding, and this recognition represents that.” Nokia’s Data Center Fabric combines switching platforms, the SR Linux network operating system, and an event-driven automation platform to support automated data centre environments. It is designed to be deployed in both new and existing infrastructure and to integrate with a range of network environments. For more from Nokia, click here.