Hyperscale Data Centres: Scale, Speed & Strategy

Yondr breaks ground on third phase of London campus

Yondr Group, a global developer, owner and operator of hyperscale data centres, has broken ground on the third building of its London data centre campus in Slough. Once completed, the 40MW data centre will bring the total capacity of the London campus to just over 100MW, making it one of the largest in the UK. The ground-breaking brings the site of a derelict paint factory back into commercial use, and the event follows completion of the first 30MW building at the London campus in July last year. An additional 30MW data centre is currently in the construction phase on site. Yondr’s third building on the London campus has been designed to BREEAM ‘Very Good’ standards, to deliver both reliable data capacity and sustainability. The new facility will be energy efficient in line with the company’s sustainability strategy and 2030 target for Scope 1 and 2 carbon neutrality. Solar panels will be installed on the building’s roof and the facility will boast industry-leading power utilisation efficiency (PUE). Plans for the building were developed in close collaboration with Slough Borough Council and the Canal & River Trust, with the goal of realising a building that actively enhances biodiversity and improves the canal side, with a focus on aesthetics and active transport access for local residents.A green wall facing out towards the canal is one of the ways in which aesthetics and sustainability have been integrated into the project, as part of a strategy of blending the building with its surroundings. The green wall has been designed on an independent structure to overcome the technical requirements of the data centre and it will provide both visual and acoustic shielding, in addition to contributing to biodiversity. Yondr’s commitment to sustainability will extend beyond the border of the site for the new data centre, with the planting of trees and shrubs which have been specially selected to be sympathetic to local plant species and wildlife. A new walking and cycle route will contribute to the health and fitness of the local community and help to reduce vehicular traffic by creating a convenient cut-through. In a further step to create a positive impact in the local area, Yondr conducted a community-needs assessment which highlighted challenges children face in accessing outdoor learning opportunities. To help bridge this gap, Yondr has provided funding for six classes from local schools for a day of hands-on learning at the Iver Environment Centre – allowing the children to immerse themselves in biodiversity, conservation, and the importance of protecting the environment. Peter Hill, VP of Design & Construction EMEA at Yondr, comments, “The plans for this third building on our London campus show a clear evolution of our data centre design and delivery capabilities. It demonstrates just how far we have come in embracing sustainability to deliver our carbon neutrality goals and bring forward exemplar data centre developments. “Breaking ground on this project is a milestone not just for this building, but also for our London campus and our increasingly strong presence in Europe. I’d like to thank everyone who has helped progress this project through the design and planning stage, as well as those who have joined us to celebrate today, and I look forward to a smooth construction process and handover next year.” The facility is expected to be completed in mid-2026. For more from Yondr Group, click here.

RWO wins ‘next generation’ data centre work

North East engineer, RWO, has won work to support the development of the first of a new generation of data centres, strengthening its position as a growing provider of services to the sector. The Newcastle-based firm is providing an undisclosed package of civil, structural and geo environmental engineering services for the Latos Data Centres hyperscale data centre in Cardiff, a Tier III asset designed to meet the needs of the most demanding global technology companies. The move follows the Stockton-based IT services and consulting specialists’ announcement to deliver 40 new data centres across the UK by 2030. Cardiff will be the first of its data centres to go fully live later this year, and it's set to offer a total of 90MVA across 50,400 square metres of floor space. Power for the site will come from a 100% renewable energy supply from the National Grid, supported with a backup feed from the neighbouring 1000MW Tremorfa Energy Park - which is one of the world’s largest battery energy storage facilities. RWO’s work, which is being undertaken in conjunction with Teesside architect, Create Architecture, sees the provision of engineering expertise to bring forward state-of-the-art facilities to meet strong demand for domestic data processing and storage capabilities. The data centre market is poised for significant growth in the coming years, with forecasts indicating a substantial increase in demand, particularly when it comes to Artificial Intelligence (AI) - with the European AI market expected to grow by 25.9% in 2024, with annual growth of 15.9% until 2030. The Cardiff project comes as RWO continues to expand operations beyond its traditional North of England heartland. Ross Oakley, Managing Director of RWO, says, “Securing this work is another big step forward for us as we continue to grow our presence in the burgeoning data centre design and build sector. Our involvement through the planning and pre-construction phase, coupled with our in-depth knowledge of highly-specialist industrial and temperature-controlled facilities such as this, has enabled us to bring forward an effective programme of engineering to deliver the project.” Latos plans to open its purpose-built data centres across the UK by 2030 as part of a mission to enable UK businesses to capitalise on the power of advanced computing, including AI. For more from Latos Data Centres, click here.

CapitaLand India Trust signs agreement with global hyperscaler

CapitaLand India Trust (CLINT) has signed a long-term agreement with a leading global hyperscaler for one of CLINT’s data centres under development. With this, CLINT is likely to pre-lease about half of its total gross power capacity under development of around 250 MW. Gauri Shankar Nagabhushanam, CEO of CapitaLand India Trust Management, says, "CLINT's decision to diversify into critical infrastructure such as data centres is validated by the strong interest shown by hyperscale and enterprise customers and positions us well to capitalise on India’s burgeoning digitalisation needs and drive long-term value for our unitholders. We are on track to complete the development of our data centres in India, accelerate leasing momentum, and unlock value through divestments of partial stakes to strategic investors. With the support of our sponsor, CapitaLand Investment (CLI), we are well-poised to meet the fast-growing demand for data centre capacity in India." Manohar Khiatani, Senior Executive Director of CLI, who oversees the group’s data centre business and is Chairman of the Trustee-Manager, adds, “India is a core market for CLI and we see strong growth potential across the asset classes that CLINT is active in, including data centres. This agreement with a leading hyperscaler is a milestone and demonstrates our ability to deliver world-class data centre solutions. It underscores our deep in-country expertise and strong vertically-integrated capabilities in acquiring greenfield land; designing, developing and completing construction on schedule; as well as leasing and operating data centres. "Digitalisation is the global megatrend of our times and CLI has significantly expanded its data centre business and capabilities worldwide in the last few years. With our 27 data centres across the globe offering more than 800 MW in gross power on a completed basis, we are well-positioned to cater to the expansion needs of hyperscalers and enterprises. “India is seeing a surge in demand for data storage and processing given the significant scale and growth of mobile and internet users and businesses widely adopting technologies such as AI and Internet of Things. The rapid pace of digital transformation, increased adoption of cloud computing and the need for data localisation will continue to drive demand for data centres. With 30 years of experience in India, strong technical capabilities and a global customer network, CLI is fully equipped to support CLINT’s growth in India.” CLINT is developing four state-of-the-art facilities in all the key data centre corridors of India (Mumbai, Chennai, Hyderabad and Bangalore). The company says that this strategic diversification strengthens its portfolio in India, enabling it to deliver quality risk adjusted returns to unitholders.

Tract Capital introduces data centre development platform

Tract Capital, an alternative asset manager focused on creating businesses that enable rapid scaling of digital infrastructure, has introduced Fleet Data Centers to programmatically meet the next phase of hyperscale data centre growth. Fleet Data Centers enters the market with a specific focus on mega-scale campuses with a prioritised target of single-user campuses. The Fleet Data Centers team consists of industry veterans who have already made a lasting imprint on the evolution of global digital infrastructure and believe they are uniquely capable of enhancing data centre development scale and operations in the face of rising demand. Tract Capital CEO and Executive Chairman of Fleet, Grant van Rooyen, comments, “We are focused on meeting the capacity and scale needs of tomorrow. Predictable and flexible data centre delivery on large-scale contiguous campuses is the logical solution for customers trying to navigate divergent demand forecasts. “The scope of our opportunity is expanding daily as customers search for new models to replace legacy building blocks and to meet the pace of demand for new infrastructure. Chris and I first worked together 26 years ago. I have immense admiration and respect for his body of work, his leadership anchored by a bias for action, and his character. It is a privilege to welcome him to our business family and to entrust the Fleet business into his capable hands.” Fleet Data Centers intends to build some of the largest data centre campuses in the world, combining gigawatt-level capacity with a campus-based commercial model that gives customers the ability to confidently secure capacity to meet their high-side demand forecasts. Fleet Data Centers’ campuses and operations will be designed in collaboration with customers to deliver seamless augmentation to their existing data centre fleets and access ongoing design innovation to intercept new technologies. Chris Vonderhaar joins Fleet Data Centers as President, most recently coming from Google Cloud where he served as Vice President of Demand and Supply Management. Prior to his time at Google Cloud, Vonderhaar served in a decade-plus tenure at Amazon Web Services (AWS) where he was responsible for the design, planning, construction and operations of the AWS global data centre platform, including energy and sustainability. Vonderhaar is joined by industry veterans from hyperscalers, wholesale data centre providers, network infrastructure providers and equipment vendors, who have collectively deployed dozens of gigawatts of data centre capacity across hundreds of data centres globally. Chris states, “Our team has direct insight into the challenges hyperscalers face as they scale. Designing, building and operating these platforms is getting harder while demand is accelerating. Hyperscalers need infrastructure that is going to deliver predictability and flexibility decades into the future. Our collaborative, long-term model backed by our engineering, system and operational excellence positions Fleet Data Centers to be an integrated extension of our customers’ data centre platforms. Grant and the Tract Capital team have been successfully starting new and innovative businesses for decades and have a proven and respected record of success. I am thrilled to be their partner.”

Harrison Street and PowerHouse sell Virginia data centre

A joint venture between Harrison Street, an investment management firm focused on alternative real assets, and PowerHouse Data Centers, a real estate developer for next-generation hyperscale data centres, announced the sale of ABX-1, a 265,580-square foot powered shell data centre with 60 MW of power, located in Ashburn, Virginia. The sale marks the successful completion of the joint venture’s first development together. Launched in January 2022, the PowerHouse and Harrison Street joint venture partnership has committed $3.2 billion to develop powered shell data centres representing nearly 6 million square feet of data centre space in Northern Virginia, Dallas and Reno. ABX-1 was fully leased to global data centre company, CyrusOne, ahead of the data centre's final completion in October 2023. "This project reflects the strength of the Northern Virginia data centre market, which has been the nation's fastest-growing data centre market for the past five years,” says Michael Hochanadel, Managing Director and Head of Digital Assets for Harrison Street. “Since forming our joint venture with PowerHouse in 2021, we've successfully executed our strategy of developing flexible powered shell facilities in Northern Virginia that meet the exacting specifications of leading cloud providers and colocation users. We look forward to continuing this partnership and expanding our portfolio across Northern Virginia and other strategic US markets where robust digital infrastructure is in high demand.” The sale of ABX-1 is a notable milestone for PowerHouse’s joint venture with Harrison Street. “PowerHouse’s innovation and expertise, coupled with Harrison Street's valuable partnership, enables us to deliver high-quality data centre infrastructure with unbeatable speed to market," explains Doug Fleit, Co-Founder and CEO of PowerHouse Data Centers. "With several projects underway across the country and a growing pipeline of developments, we remain laser-focused on serving the growing demands of hyperscale providers and creating long-term value for our communities and the broader industry." Following the sale of ABX-1, PowerHouse and Harrison Street have three campus developments completed or underway in Northern Virginia, with an executed lease, pre-lease, or letter of intent for 100% of the buildable square feet. In addition to the Virginia portfolio, the joint venture recently purchased land and started construction on PowerHouse Reno (in Reno, Nevada), and PowerHouse Irving (in Dallas, Texas), to develop new state-of-the-art data centre campuses to address the needs of these rapidly growing, highly connected data centre markets. Since 2018, Harrison Street has committed over $5.4 billion to powered shells, carrier hotels, colocation sites, and dark fibre platforms. Over this time, the firm has expanded its relationships with dedicated operating partners and hyperscale users, raised discretionary capital for a dedicated digital vehicle andrealised on its first data centre investments. The sale of ABX-1 continues to demonstrate the growing investor demand across alternative real estate sectors. For more from PowerHouse, click here.

IPTO and Serverfarm to develop data centres in Greece

The Independent Power Transmission Operator (IPTO) of Greece and Serverfarm, a global data centre developer and operator, announced today the signing of a Heads of Agreement regarding the formation of Gemini, a joint venture with the objective of developing and operating state-of-the-art, hyperscale-ready data centre facilities in Athens and elsewhere in Greece, on sites owned by IPTO. This strategic alliance marks a significant milestone in advancing Greece's digital infrastructure and fostering sustainable growth in the data-driven economy. Gemini will combine the collective expertise and resources of two leading entities in their respective fields. IPTO, as a crucial pillar of the Greek energy sector, plays a pivotal role in managing the country's electricity grid, ensuring stability, and embracing renewable energy solutions. Leveraging its extensive experience, IPTO will be providing reliable and sustainable power supply, essential for supporting data centre operations along with strategically located sites with access to power and optical fibre networks, as well as other operational resources. Serverfarm, as an expert in data centre development and operations globally, with a strong track record in commercial real estate ventures, will be bringing unparalleled industry experience and advanced design and operational know-how, ensuring that the joint venture will become a point-of-reference for data centre services in Greece. Gemini plans to construct and operate hyperscale-ready data centre facilities in the Greater Athens area, leading initially with a campus with committed power of 130MW, creating the foundations for establishing a robust digital ecosystem to meet the escalating demands of cloud service providers, content delivery networks and enterprises. The data centres will be designed with a focus on energy efficiency, utilising state-of-the-art cooling technologies and renewable energy sources to minimise their environmental impact, creating an unparalleled platform for hyperscale computing in Greece. The joint venture envisions Athens as a prominent digital hub, offering secure, reliable, and low-latency data centre facilities to national and international clients, bolstering Greece's position in the global data economy. Avner Papouchado, the Founder and CEO of Serverfarm, says, “As leading tech and hyperscale organisations continue to expand into Greece, we see a great opportunity to leverage our expertise in this region. Our strategic partnership with IPTO, underscores our mission to invest in transformative projects that create long-term value. The Greek data centre market is still one of the most under-served in Europe, but at the same time, its geographical location makes it ideal to serve as a data gateway between continents. “Our collaboration with IPTO in creating Gemini and our shared commitment to excellence and sustainability will enable us to leverage this immense potential to offer high quality, data centre services in Greece. Our goal is to cater to the growing needs of hyperscale customers in the area and elevate Athens as a major hub for the industry, shaping the digital landscape in the broader region.” Manos Manousakis, the Chairman and CEO of IPTO, adds, “IPTO is building critical infrastructure for tomorrow's electricity and telecommunications backbone networks throughout Greece and beyond, interconnecting the future. We are delighted to partner with Serverfarm, through Gemini, to deliver world-class data centre facilities in Greece. “Serverfarm’s vast experience in developing state-of-the-art data centres, coupled with IPTO's robust energy infrastructure and its strategic location at the crossroads of continents, will create an unparalleled platform for hyperscale computing in Greece, catering to the escalating demand for digital services in the region. This alliance symbolises a significant milestone for Greece's digital transformation, which is poised to build the foundation for a flourishing data-driven economy in Greece. “In this way, we fully exploit synergies and create win-win business opportunities, transforming Greece into a critical energy and data hub of high geopolitical value, at the crossroads of Europe, Africa and Asia.”

Nokia building IP network to support AI workloads

Nokia has been selected by Nscale, the hyperscaler engineered for AI, to deliver an IP network solution to support AI workloads at Nscale’s new data centre in Stavanger, Norway. The data centre, which is powered entirely by renewable energy and optimised for energy-efficient cooling, will enable cutting edge AI services, including Graphics Processing Unit as a Service (GPUaaS), a technology which Nscale is an expert in. With a growing global demand for AI-driven applications that consume a large amount of data, Nscale required a reliable, high-performance network to support training and inferencing on large-scale GPU clusters. Nokia’s Ethernet-based data centre fabric provides scalability, programmability, and low-latency performance - essential for handling the rigorous demands of AI workloads across industries including education, healthcare, government and finance. David Power, CTO at Nscale, says, “Our mission is to redefine the boundaries of AI and High-Performance Computing through innovative, sustainable solutions. Nokia’s data centre fabric enables us to scale our GPU clusters while maintaining the reliability and performance needed to serve our customers with cutting-edge AI services. The flexibility of Nokia’s solution ensures we can bring advanced AI capabilities to market faster.” Paul Alexander, Vice President and Country General Manager UK&I at Nokia, adds, “As a hyperscaler, Nscale is already delivering turnkey AI development and deployment solutions to businesses worldwide. With support from Nokia, the path to accessible and transformative AI innovation has never been easier. By combining advanced Ethernet technology with sustainability, we are helping Nscale deliver world-class AI services while supporting its commitment to renewable energy and environmental responsibility.” The solution, based on Nokia’s 7220 IXR and 7750 SR platforms, is designed with the most modern, open, and programmable architectures to ensure ultra reliability, seamless automation and adaptability. Nokia is helping cloud builders worldwide to build modern data centre networks that are highly reliable, secure and easy to operate – which is essential to meet the growing demands of AI workloads worldwide. For more from Nokia, click here.

Spirent selected for Ethernet infrastructure testing

Spirent Communications, a provider of test and assurance solutions for next-generation devices and networks, today announced that a leading AI-focused hyperscaler has selected Spirent to perform critical 800G Ethernet fabric testing and assurance, to ensure optimal performance for AI workloads. Spirent’s solution will enable the hyperscaler to balance efficiency and sustainability, while monetising its AI Ethernet infrastructure investments. “Artificial intelligence is transforming the industry and this hyperscaler is leader in building the next generation public cloud for enabling large scale AI workloads,” says Aniket Khosla, Spirent’s Vice President of Wireline Product Management. “At Spirent we understand the challenges that come with building sustainable AI data centre networks and supporting high-performance compute use cases, and we’re excited to support leading industry players as they address the performance demands of advanced AI applications.” Hyperscalers globally are continuing to scale rapidly to meet the dramatic demand for accelerated compute solutions critical for training and serving the most sophisticated AI models. Leveraging Spirent’s B3 800G Appliance and proven test methodologies, Spirent is enabling hyperscalers to optimise the performance of high-speed Ethernet infrastructure required to support AI-driven applications. The B3 800G Appliance is a high-density native 800G OSFP and QSFP-DD test platform, and the first in its class to support IEEE 802.3df specifications. It provides rigorous, high-scale testing to evaluate bandwidth and resource utilisation, helping to accelerate 800G deployments and enable the networking industry to power new applications running AI/ML. With the B3 800G Appliance, hyperscalers can test and validate their infrastructure’s readiness for bandwidth-intensive AI workloads, while maintaining optimal network efficiency. For more from Spirent, click here.

Nscale secures funding for AI infrastructure expansion

Nscale, the hyperscaler engineered for AI, has raised $155 million in Series A funding to accelerate expansion across Europe and North America. Led by Sandton Capital Partners, with participation from Kestrel, Bluesky Asset Management, and Florence Capital, this oversubscribed funding round signals strong confidence in Nscale’s path to building the hyperscaler, purposely engineered for AI. Nscale develops sustainable AI-ready data centres, deploys massive-scale GPU infrastructure, and delivers high-performance AI cloud services. This unique full-stack approach enables Nscale to co-design the data centre and GPU supercluster infrastructure to deliver unrivalled speed, performance, and efficiency on AI workloads. Nscale's full-stack AI private cloud platform is designed to support the entire generative AI lifecycle, enabling developers to run and train AI models. Nscale’s platform provides bare metal and virtualised GPU nodes, Kubernetes-native services (NKS), AI workload scheduling powered by SLURM, and advanced serverless and dedicated inferencing options. In addition, Nscale is launching a public cloud service in Q1 2025 that will provide developers access to purpose-built inference and training solutions in a flexible development environment. Since launching from stealth in May 2024, Nscale has experienced insatiable demand for AI infrastructure, quickly growing its pipeline of greenfield data centres across Europe and North America from 300MW to 1.3GW, with 120MW planned for 2025 development. These sites will be designed from the ground up to support large supercluster deployments for global hyperscalers and LLM platforms, leveraging the latest advancements in closed-loop direct liquid cooling to deliver maximum performance and efficiency, while minimising environmental impact. The hyperscaler is now uniquely positioned to capitalise on the market for large-scale AI infrastructure, and can deliver bespoke GPU clusters at any scale for governments, AI scale-ups, and global enterprises. “The AI market is scaling rapidly, and so are we,” comments Joshua Payne, CEO of Nscale. “Nscale manages every layer of infrastructure in the value chain to meet the intensive needs of large-scale AI customers. In particular, the largest risk to the market's ability to scale is the large contiguous tranches of electricity required to power these large GPU superclusters. Nscale has a 1.3GW pipeline of sites in our portfolio, which allows us to design from the ground up, the data centre, the supercluster and the cloud environment end-to-end for our customers. “This means we can deploy bespoke GPU clusters tailored to customers’ requirements, do this faster than competitors and at any scale, all while delivering superior unit economics. With the support of our investors, we can now reach more customers globally with bespoke designed, sustainable and cost-effective AI infrastructure that unlocks new AI capabilities, products, and services.” Rael Nurick, Co-Founder at Sandton Capital Partners, adds, “We are excited to double down on our investment in Nscale with our lead investment in the Series A round. With a notable founder track record, established industry partnerships, and a unique vertically integrated approach, Nscale is building the hyperscale AI platform to power enterprise AI at scale.” This investment is the latest in a series of significant milestones, reflecting Nscale’s continued commitment to driving innovation and delivering value to customers since it raised $30 million in seed funding in December 2023. For more from Nscale, click here.

Pure Storage and Kioxia form hyperscale DC partnership

Pure Storage, an IT expert that delivers advanced data storage technology and services, has announced a collaboration with Kioxia Corporation, a provider of memory products, to develop cutting-edge technology that addresses the growing demand for high-performance, scalable storage infrastructure among today’s hyperscale environments. Traditional storage solutions, particularly those relying on hard disk drives (HDDs), struggle to meet the demands of hyperscale environments due to their limited speed, scalability, reliability, and excessive power consumption. HDDs are ill-equipped to handle the massive, fast-growing volumes of data generated in these environments without introducing latency and bandwidth limitations, and are difficult to scale and integrate seamlessly into modern data centers. To eliminate these obstacles, the latest collaboration between Pure Storage and Kioxia will deliver a data storage platform engineered from the ground up to tackle the needs of hyperscale environments, allowing for rapid scale while lowering power consumption and reducing the overall physical footprint of hyperscale data centers. With the combination of Pure Storage’s advanced data storage platform with Kioxia’s QLC flash memory, hyperscalers can now keep pace with growing data demands without sacrificing performance. Benefits include: Increased Performance: Ultra-fast, low-latency performance for handling massive, data-intensive workloads, while enabling greater reliability to keep pace with rapidly growing customer demands. Improved Cost Efficiency: Optimised storage costs and lower overall total cost of ownership (TCO) by reducing the need for extensive physical resources, while offering better scalability. Lower Power and Waste: Reduced energy consumption significantly compared to traditional HDDs, lowering operational costs, leading to a smaller overall physical footprint, cutting hardware waste and supporting more sustainable data centre operations. Bill Cerreta, GM, Hyperscale, Pure Storage, comments, "Collaborating with Kioxia allows Pure Storage to bring the full potential of all-flash storage technology to hyperscale environments. Pure has a decade of experience in delivering systems that manage flash for enterprise businesses. Now we’re using those innovations, and Kioxia’s latest technology, to enable the hyperscalers. Together, we’re creating a solution that will empower these organisations to manage their data seamlessly, with speed and efficiency at the core." Caesar Ichimura, Chief Marketing Officer, Kioxia Corporation, adds, "Our collaboration with Pure Storage marks an exciting milestone in the evolution of hyperscale storage. As data volumes continue to explode, we’re committed to delivering a solution that combines high performance with lower operational costs. All-flash technology is the future of storage, and through this collaboration, we are driving the innovation needed to address the complexities of today’s hyperscale environments." For more from Pure Storage, click here.