Neoclouds: The Category Reshaping Data Center Demand
If you haven’t come across the term “neocloud” yet, you probably will soon. What looked like a niche corner of the market a couple of years ago is now becoming a serious force in data center development.
Neoclouds are specialist cloud providers delivering on-demand GPU infrastructure. In practice, that means the computing power behind AI training, inference, and scientific workloads. Unlike hyperscalers, which offer a much broader cloud platform, neoclouds are focused on high-performance compute. That focus lets them move quickly, compete on price, and build around the density requirements of AI rather than trying to adapt older data center models.
The growth figures are hard to ignore.
- The segment has recorded an 82% five-year CAGR in revenue growth.
- CoreWeave ‘s stock has nearly tripled since its IPO.
- Nebius secured $700 million in backing from NVIDIA and Accel in 2024, followed by a further $1 billion raise in 2025.
- S&P Global has reported over $10 billion allocated to the sector in the past year.
For MEP engineers and the firms that employ them, the important point is what these facilities actually need. Many are operating at power densities above 100 kW per rack. They need liquid and direct-to-chip cooling, stronger structural allowances, specialist fluid networks, and power infrastructure that most existing buildings cannot support without major retrofit. They also need it delivered quickly.
The talent implications are obvious. MEP engineers with experience in high-density liquid cooling, critical power systems, or fast-track data center delivery are among the most in-demand people in the industry right now. Although the neocloud boom is not creating entirely new jobs, it is increasing the urgency and volume of roles the market was already struggling to fill.
Water Is the Next Frontier for Cooling Engineers
The industry’s shift toward liquid cooling is well understood. Less discussed is the operational complexity that comes with it, particularly around water sourcing, quality, discharge, and management at commissioning.
Vertiv ‘s launch this week of its PurgeRite NearZero service is a direct response to that challenge. The service integrates engineered flush planning, mechanical flushing, water treatment, filtration, reverse osmosis, and continuous water-quality monitoring to recycle flushing water through the commissioning of closed-loop hydronic systems. In selected deployments, it has reduced total water consumption by up to 78%, water haul-off volumes by up to 91%, and discharge management costs by as much as 34%.
One installation eliminated approximately 300 tanker trips for water delivery and removal.
The environmental angle is important, but the wider point is that liquid cooling is becoming a more specialized market. As high-density cooling becomes standard in AI infrastructure, fluid system commissioning needs its own skills and methods. MEP engineers who understand both the thermal side and the water treatment side are becoming more valuable. We are seeing that reflected in what clients are asking for.
The Power Gap Is the Market’s Biggest Constraint, and Talent Is Part of It.
The neocloud story is compelling. The infrastructure story behind it is more complicated.
Several analysts have warned that a significant portion of announced data center capacity may face delays due to power constraints, utility interconnection challenges, and equipment shortages. Capital and GPU availability remain factors, but power is now the main constraint, from grid interconnection and transformers to the wider utility infrastructure needed to bring facilities online.
For neocloud operators, this is a direct challenge to their core proposition. Speed to market is the competitive advantage. If facilities cannot connect to power on meaningful timescales, that advantage is neutralised.
The firms best dealing with this tend to have strong expertise in utility engagement, power procurement, brownfield sites, and retrofit options. They are also the firms that have invested seriously in MEP and critical systems engineering talent. From feasibility through to energisation, the quality of the technical team can decide whether a project moves or stalls.
By most measures, we are still in the early stages of the AI infrastructure build-out. Demand is strong, and the investment committed is significant. But the physical layer still matters. Engineers, commissioning teams, cooling specialists, and power infrastructure professionals remain the bottleneck that capital alone cannot solve.
That is ultimately what MEP Pulse is about: the people building the infrastructure that the next decade of technology depends on.
The Community Dimension Is Now a Project Delivery Risk
There comes a point in every major infrastructure cycle when public resistance stops being background noise and starts affecting delivery. For data centers, we appear to be at that point.
Environmental activist Erin Brockovich launched a nationwide data center tracking platform in late May, inviting communities to submit reports about facilities near them. By early June, the map had received more than 5,000 submissions. The concerns include water consumption, electricity demand, noise, and pressure on local infrastructure. They are also being heard by local governments.
Seattle has passed a one-year moratorium on new data center development. Multiple counties in Kentucky have introduced similar temporary measures. Cities across Georgia, including Atlanta, and Monterey Park in California, have approved development bans this year.
The data center industry has often treated community engagement as a planning formality. That is becoming harder to sustain. The regulatory environment around new builds is tightening, and the sites most likely to progress are the ones where developers have taken community relationships, environmental commitments, and transparency seriously from the start.
For MEP and infrastructure professionals involved in site selection and early-stage development, this is increasingly part of the brief.
Closing the Lifecycle Gap
One notable development this month is the launch of Global Certification and Verification Group, a newly aligned group of independent specialists focused on commissioning, verification, and safety governance for mission-critical infrastructure, including data centers.
The firm’s view is that a gap has opened between design and operation. Build timelines have shortened, technical specifications have become more demanding, and independent oversight across the project lifecycle has not always kept up. GCV Group is positioning itself around what it calls “lifecycle continuity”, bringing commissioning, HV electrical, certification, and verification specialists under one framework rather than treating them as separate workstreams.
It is a useful way of describing something the market has been feeling for a while. The current build cycle is moving faster than the quality assurance structures that supported more measured delivery programmes. In neocloud facilities operating at extreme densities, the consequences show up clearly at commissioning, energisation, and handover.
From a hiring perspective, this fits with what we are seeing in the market. Commissioning engineers and technical assurance specialists with data center experience remain in short supply. Firms growing in this space are investing heavily in that capability, whether through senior hires, graduate programmes, or acquisitions of specialist teams. That demand is unlikely to ease.
Donnacha Mescal
Donnacha specializes in recruiting talent across the MEP and Commissioning sectors, connecting industry experts with leading organizations. Through a strong network spanning HVAC, plumbing, electrical systems, and commissioning, he helps companies secure professionals who deliver technical excellence and support strategic growth across complex building and infrastructure projects.
At Metric DCX, our work in the data center market covers the full MEP and critical systems landscape, from design and build through to commissioning, operations, and the specialist technical disciplines that AI-dense infrastructure now requires.
We work with neocloud operators, hyperscale developers, and the specialist consultancies and contractors to support their hiring needs. Our focus is on placing engineers, technical leads, and senior professionals who understand what this generation of data centers requires.
If you are building a team, responding to demand growth, or simply want to understand where the talent market is moving, I’m always happy to talk.
