Phil Alsop
Verified
Is this you? As a journalist, you can create a free Muck Rack account to customize your profile, list your contact preferences, and upload a portfolio of your best work.
Claim your profile
Articles by Phil Alsop
AI Enters the Infrastructure Era
For years, we thought that competition in artificial intelligence would be determined by model quality. Each new version seemed destined to reshuffle the market, fuelling a constant race for better benchmarks, higher parameter counts and ever-greater computing power.
WHY SUPPLY CHAIN AGILITY IS KEY TO DATA CENTRE CONSTRUCTION EFFICIENCY
Data centre construction and capacity are exploding, in fact faster than the industry can keep up. It’s a level of growth that has not only left as many as half of planned AI data centres running behind schedule, but also revealed our industry’s vulnerabilities in procurement. It’s a problem which continues to cause a crippling bottleneck: a balancing act between our evolving, complex machines and maintaining supply chain agility to ensure we keep pace with rapid technological changes.
Honest conversation, not blanket prohibition, is the key to fair and flexible energy grids
Data centers are rapidly becoming political footballs, caught between those who recognise their essential role in underpinning future growth and those rightly concerned about their impact on the environment and local energy prices. Binary thinking and polarisation will benefit no-one. What is required are new approaches, innovative ways of working and genuine collaboration. At the heart of this shift must be a recognition that data centers can be more than a significant load on energy grids.
Buying GPUs is the easy part: The infrastructure challenges behind enterprise AI
As organisations race to accelerate AI projects by investing in GPUs, a key question is being overlooked. What happens once this powerful, expensive hardware enters the data centre? The industry is solving the GPU problem. Supply chains have adapted, capacity is being secured, and budgets have been approved. What’s being underestimated is the GPU operations problem: the power, cooling, networking, and lifecycle realities that determine whether any of that compute produces value.
The data centre industry has solved cooling… now comes the hard part
The data center industry has made its decision on liquid cooling. Facilities running high-density AI workloads led the way. The major hardware vendors have designed their latest AI processor hardware around it. Direct-to-chip cooling, where liquid is routed through cold plates directly onto the processors, is now the reference architecture for any new AI-optimized facility. The question is no longer whether liquid cooling will become standard. It already has.
Power efficiency is a strategic growth enabler
Historically, CIOs have been in charge of managing servers, storage, networking and the associated budgets. Today, they must also manage something else - something that has become just as valuable as those resources - access to power. As enterprise data continues to grow exponentially and AI workloads place increasing demands on infrastructure, power availability is emerging as one of critical constraints affecting growth potential and digital transformation rates.
Building Europe’s AI hub in the Nordics
Artificial intelligence is becoming foundational to Europe’s industrial competitiveness, energy transition, security and public services. But AI leadership is not only about algorithms and talent. It is about infrastructure: power, land, cooling and increasingly high-performance connectivity. Global and European technology players are already moving. Google has invested heavily in Finland, Denmark and Norway. Microsoft has established Azure regions in Norway and Sweden.
The data centre industry has a trust rather than a water problem…
Britain has spent this extra-hot summer discovering, it would seem, that data centres use water - and a lot of it. They do of course, but in context the water-usage scare stories are overblown and misleading. Water UK’s own figure is that data centres in England draw around 6.6 million litres of drinking water a day, rising to something near 20 million if capacity trebles as government ministers intend. But set that against the system it sits inside.
AI’s New Battleground is Power
For years, the AI race has been a battle of chips, algorithms and talent. But now, there is a more fundamental resource determining who wins: electricity. Every query and model runs on power, and increasingly, it is the availability of that power, not computing hardware, that decides where AI infrastructure gets built.
“FAIL TO PREPARE, PREPARE TO FAIL” - WHY WEATHER RISK MANAGEMENT NEEDS TO BE TAKEN MORE SERIOUSLY
With experts predicting an intense and prolonged El Niño weather phenomenon, the world is on track for droughts, floods and more extreme weather events. In the UK, experts are anticipating a particularly stormy autumn of heavy rain and strong winds, followed by a harsh winter of frozen conditions.
Zinc-plated steel vs. zinc die cast fittings
The United States has thousands of operational data centers, and that number is expected to grow substantially in the years ahead, with thousands more new data centers currently in various stages of development nationwide. As a result, both retrofitting current operations for demand growth as well as the creation of entirely new facilities require an enormous amount of cable and conduit fittings as they play a substantial role in datacenter infrastructure execution.
Microchip and Navitas enable 800V transition
As AI data centers scale to support high-power GPU clusters, the industry is shifting toward 800V DC rack power architectures to improve distribution efficiency, increase power density and support next-generation server designs. To help accelerate this transition, Microchip Technology and Navitas Semiconductor have collaborated on an 800V DC-to-6V DC reference design for AI data center rack power applications.
The grid is changing: Why data centres must evolve with it
The energy grid is changing. Large, centralised power stations were traditionally located in areas with high demand, and our grid was built around this layout. Today, the shift to renewables means power generation is increasingly concentrated in areas with the strongest wind and solar resources, often far from where it is needed. This is happening as our demand for electricity grows. Transport is becoming electric. Heat pumps are replacing boilers.
Siemens and TD SYNNEX embark on global partnership
Siemens and TD SYNNEX have agreed on a joint framework to establish a global distribution partnership.
The case for modularisation in modern data centre design
It can be debated whether we’re still in the data centre “boom”. Capacity upgrades and planning applications have both soared in a post-pandemic world. That may be a world we’re still getting to grips with, but you could argue that data centres are now an essential part of our daily lives and the “boom” is therefore the new norm. Their recognition as Critical National Infrastructure in September 2024 says as much. Two years on, however, demand shows no signs of letting up.
Considerations for Heat Sensor Customization in AI Data Centers
According to ABB, a single server rack is now consuming 30, 50, or even 100 kW — far exceeding the traditional 5 to 10 kW per rack. With 10x to 20x more power in the same rack, the power loss density and temperatures are going up. Not only in the rack itself, but the temperature can fluctuate in the data halls where racks are located. If all these factors aren’t accurately monitored and adapted to, it can be detrimental to data centers.
The gas hazards hiding inside a data centre
A data centre does not look dangerous. There is no flare stack, no tanker, no process vessel. Just cold air, sealed rooms and a control room. It is one of the tidiest industrial environments most engineers will ever walk into. That appearance is the problem. The risk does not sit where the people are. It sits in the spaces nobody visits, created by the systems that make the facility reliable: battery backup, mechanical cooling, standby generation, electrical distribution and gaseous suppression.
Liquid cooling without the rip and replace
For decades, air cooling has been the default approach in the data centre and, for a great deal of equipment, it continues to do the job perfectly well. The challenge is that AI, high-performance computing and other processor-intensive applications are driving much higher rack densities, placing new demands on cooling environments that were often designed for very different workloads. For data centre managers, that does not mean liquid cooling has to become an all-or-nothing decision.
Device Sprawl Is a Business Reality; Tool Sprawl Is an IT Choice
By Apu Pavithran, CEO & Founder, Hexnode.
Breaking Down Silos: Why Integrated Planning Is the Key to Sustainable Data Centre Growth in APAC
The rise of artificial intelligence, cloud computing and digital services is driving unprecedented demand for data centres across the Asia-Pacific region. From Singapore and Malaysia to Indonesia, Thailand, Japan and Australia, governments and investors are racing to expand digital infrastructure to support economic growth, digital transformation and increasingly data-intensive technologies.
Small but mighty – micro data centres are the foundation of security and disaster recovery
Digital services now underpin almost every aspect of modern society. Banks process millions of transactions every minute, insurers rely on their IT systems to manage policies and claims, while governments deliver essential public services online.
Delivery certainty will define the next phase of data centre growth
Demand continues to dominate conversations about data centre growth. Delivery capability will increasingly determine which projects progress and which remain part of the announced pipeline. Investment remains strong, AI is increasing the requirement for capacity and developers continue to pursue ambitious growth plans. However, the strength of the opportunity can obscure a more important question.
Finland and Cambridge deepen collaboration between two leading quantum ecosystems
According to the ECIPE Quantum Clusters Ranking, Cambridge ranks first globally and Greater Helsinki, mainly Espoo, second, underlining the strength of both ecosystems and their potential for closer collaboration in quantum technology and deep tech.
Automated intelligence: Closing the gap between the rounds
Data center operators already fully understand the pressures of maintaining uptime, resilience, and power and cooling demands. Faults can move quickly, and every intervention carries operational consequence. Regular inspection and maintenance are essential, but evidence gathering often does not keep pace with changing conditions within data center facilities.
The ‘anytime, anywhere workforce’ model that could help America win the AI race
The US risks losing its leading position in the global AI race with $64 billion worth of projects recently blocked or delayed. Skills are emerging as the next major bottleneck. Data centers are already straining power and water supplies, but the construction skills pipeline is also running dry. Shortages of key roles from electricians to site supervisors are driving some contractors to turn down projects, pushing up labor costs by 10% for some roles and delaying .
Risks of exposed penetrations on data centres
The effective weatherproofing of exposed building penetrations creates a strong and durable seal helping to provide long-term protection to the buildings and the sensitive equipment that they contain. What are the risks of exposed penetrations? When a building penetration is exposed, it becomes vulnerable to moisture ingress.
Rubicon Professional Services surpasses one gigawatt of battery energy storage capacity
As artificial intelligence (AI) fuels unprecedented demand for data centers and resilient electrical infrastructure, Rubicon Professional Services (RPS) has surpassed one gigawatt (1 GW) of installed battery energy storage capacity across projects nationwide, reinforcing the company's role as one of the infrastructure firms quietly enabling America's rapidly expanding AI economy.
Copilot is exposing years of Microsoft 365 governance debt
Why Copilot readiness depends on the Microsoft 365 estate beneath it By Andrea Sivieri, Chief Product and Technology Officer, CoreView.
Your GPUs are only as fast as your network fabric
For most of the data centre's history, the network quietly did its job; as long as it delivered reliable connectivity, it rarely attracted the same level of attention as other parts of the system. That era is over, and AI workloads have pulled the network into the spotlight. The industry is catching up to what that means for design. Traffic has changed direction Traditional data centres were built around north-south traffic.
Rethinking the network of last resort: How PON is transforming data centre out-of-band management
Out-of-band management (OOBM) is the network of last resort. When the production network goes down, OOBM provides a pathway to the management ports of rack devices, enabling IT teams to reset, reconfigure and troubleshoot equipment. Industry research puts the average cost of data center downtime at around $500,000 per incident (1). That figure does not include reputational damage or lost customer trust.
SMA introduces dedicated solutions to accelerate time-to-power for AI-driven data centres
The company is introducing a portfolio of data center power infrastructure solutions specifically designed for AI and hyperscale applications. Built on SMA's proven utility-scale technology platform and decades of grid integration expertise, the portfolio addresses one of the industry's most critical challenges: accelerating time-to-power while delivering reliable and scalable operation.
World Water Week: designing AI infrastructure with freshwater security in mind
Last week, the global water community gathered in Stockholm for World Water Week (23–27 August), under the theme Water for People and Progress. For the data centre industry, that theme could hardly be more relevant. AI is accelerating demand for high-density digital infrastructure around the world. As the industry plans the next generation of capacity, the question is not simply how much compute we can build, but how intelligently we can design the infrastructure that supports it.
Why Modern Defence Operations Demand Resilient Data Centre Infrastructure
The digital battlespace challenge The UK and its allies operate in an increasingly digital, data-driven battlespace. Armed forces require resilient, secure, and effective computing at the tactical edge, with infrastructure required to be deployable - quickly and effectively to diverse locations, including conflict zones. This digital backbone must operate reliably in all environments, from arctic cold to desert heat, while supporting a continuous sense-decide-effect (SDE) cycle.
It's time to rethink what we mean by a legacy data centre
One of the questions we're increasingly asked by operators, investors and asset owners is whether an existing data centre is still fit for purpose. The question is usually prompted by a specific challenge. Sometimes it's driven by AI deployments and higher rack densities. Sometimes it's part of a due diligence exercise during an acquisition. Sometimes it's because cooling systems are beginning to struggle during periods of sustained high temperatures.
AI DATACENTRES NEED A WATER-STRESS TEST BEFORE APPROVAL
Britain’s drought has pushed a familiar technology debate into the water sector. On August 28, the government rejected a call to halt most new AI datacentres, arguing that they are important infrastructure and account for less than 1 percent of industrial water use in England. At the same time, the water system is under visible stress. Water Magazine reported that Wessex Water imposed its first hosepipe ban in 50 years, while restrictions now affect millions of customers.
Data centre physical infrastructure market forecast to reach $120 billion by 2030
According to a recently published report from Dell’Oro Group, worldwide Data Center Physical Infrastructure (DCPI) manufacturer revenue is projected to grow at a 22 percent compound annual growth rate (CAGR) from 2025 to 2030, reaching $120 billion by the end of the period. This growth is driven by net additions to installed IT capacity, which account for the large majority of the forecast.
KROHNE dedicates production capacity and U.S. Center of Excellence
Sunday, 30 August 2026 Log inAdvertiseAboutContact KROHNE announces a major commitment to the fast-growing data centre sector by dedicating a significant portion of its annual production capacity to meet escalating demand. This strategic move is designed to support the infrastructure behind AI and cloud growth, where accurate, reliable flow measurement is essential to mission-critical cooling and process operations.
Circular Thinking: A New Imperative for Data Center Cooling
The AI revolution has spurred heated conversation about data center cooling. As compute density climbs to support ever-larger AI workloads, the systems that keep that hardware from overheating have become a top priority. Refrigerant decisions are emerging earlier in the value chain, spurring pressing procurement and sustainability conversations that, until recently, were not part of the data center planning process.
What AI-driven growth means for the gas systems powering US data centres
One question dominates conversations across the US data center market: where will all the power come from? AI is accelerating investment in digital infrastructure, but its impact extends beyond electricity demand. As workloads fluctuate, they are also changing the engineering requirements placed on the fuel gas systems supporting engine-driven generation.
Location, location, location – The strategic requirements to unlock data centre delivery
Data centres have been outlined by the UK Government as key to its latest industrial strategy, published in November 2025 . In fact, the document highlights that in order to support the delivery of these and the subsequent economic benefit they could bring, data centres along with gigafactories and new laboratories can opt-in to be designated Nationally Significant Infrastructure Projects, which provides key benefits such as reducing the average pre-application period.
How data centres can counter rising energy costs
With more than 500 data centres operating across the UK, these facilities underpin the country’s modern digital world. They support everything from everyday cloud services and online payments to complex Internet of Things (IoT) ecosystems - and their commercial value is rising fast. According to the industry body techUK, data centres currently generate £4.7 billion in Gross Value Added (GVA) each year, with the potential to unlock an additional £44 billion by 2035.
Data Centres can – and should - be good neighbours – here’s how
Data centres are an essential, arguably the essential, infrastructure for the modern digital economy. However, for all the many conveniences that we take for granted relying on the computing power that data centres provide, they are not necessarily welcome neighbours. Indeed, opposition to data centres has become an increasingly prominent trend, both in the UK and internationally.
The AI Corridor: How Europe, the Middle East, and the United States Are Reshaping Global AI Infrastructure
Artificial intelligence is no longer defined solely by advances in models and software. Increasingly, competitive advantage depends on the physical infrastructure that enables AI at scale—including data centres, energy supply, fibre connectivity, semiconductor access, and supportive regulatory frameworks.
Super-grid transformers: The hidden engine driving Europe’s AI data centre boom
In the race to power artificial intelligence (AI) at scale, attention often falls on graphics processing units (GPUs), liquid cooling and rack densities that now exceed 100 kW. Yet the component that ultimately decides whether those racks can be energised, cooled and kept online sits outside the white space – that is the transformer.
How to build full end-to-end stack AI infrastructure and why it matters
AI is not the same as the computing that came before it. Previously, data had to be stored in structured form and retrieved through specific queries. Today, AI systems can analyse all kinds of data, including static, dynamic, structured and unstructured data in formats such as video, images, audio and text.
How are AI data centres becoming grid participators?
Hyperscale operators are securing direct grid connections and long-term capacity agreements while also participating in flexibility trials. In the UK, National Grid’s AI grid trial shows that data centres can adjust demand dynamically in response to system conditions, improving grid responsiveness. At the same time, system operators are expanding demand-side response programmes and developing faster-acting balancing markets that increasingly include large industrial loads.
The Green Opportunity in Data Centres - Why Your Data Center's Green Strategy Might be Missing Half the Equation.
Data centers now consume nearly 2% of global energy. As the mainstay of the digital economy, that level of consumption is a problem but also an opportunity at the same time. If you push efficiency forward here, even modestly, the outsized environmental payoff scales accordingly. So why isn't it happening faster? Many corporations pledge zero-carbon goals, a target year, and a whole load of optimistic language, only to produce rather underwhelming results.
From Cooling Hardware to Cooling Strategy for High-Density AI Racks: Managing the Technology Loop
To sustain the thermal demands of next-generation hardware, data center operators, senior IT managers, and engineering teams are rapidly adopting direct-to-chip (D2C) liquid cooling architectures as the new infrastructure standard. This is good news for scaling computing power. Successfully navigating this transition, however, requires a proactive strategy that moves beyond simple hardware deployment to embrace precise coolant management, ongoing fluid testing, and specialized operational service.
AI is forcing a rethink of how data centres are built
Data centre cooling pipework. The race to deploy artificial intelligence is transforming data centres into some of the most complex and resource-intensive projects in construction. GlobalData is currently tracking a pipeline of data centre construction projects worth more than $2.3 trillion worldwide, with Western Europe alone accounting for over $500 billion of planned investment. That broader surge is echoed in Europe’s own market outlook.
Schneider Electric and AMD release first Helios Platform Reference Design
Schneider Electric and AMD have announced a jointly developed and validated reference design for the AMD Helios rackscale solution that provides a scalable blueprint for deploying high-density AI environments faster and with reduced risk and complexity. The reference design marks the first milestone of the collaboration between Schneider Electric and AMD and delivers upon the companies’ joint focus to create an easier path to AI Factory deployment.
Show More
loading
Actions
Get in touch with Phil
Contact Phil, search articles and posts on X, monitor coverage, and track replies from one place.
Learn more about Muck Rack