Data Centres
within racks, considering the number of connections now required.
Compounding these challenges, supply chain constraints bottleneck how quickly operators can deploy new facilities and upgrade existing locations, which in turn affects equipment purchasing decisions.
Supporting AI at this scale must place unprecedented strain on data centre infrastructure. What impact does this have on energy consumption and power delivery? Power consumption remains a pressing concern. Data centres are energy-intensive, not only to process these massive workloads, but also to ensure proper cooling. The latest IEA (International Energy Agency) report revealed that data centres consumed 460 TWh in 2022 and projected this number to surpass 1,000 TWh by 2026.
This level of demand is creating regional capacity constraints, which affect both the distribution and transmission networks within the data centres. This is rendering the large cluster buildout model that dominated the last decade to be increasingly unsustainable. Instead, we are seeing more distributed buildouts located closer to high-density end-user locations with enough power availability.
With data centres moving closer to end users, how does this impact local communities?
There have been a lot of justifiable pushbacks from local residents, some of which have led to legal action. One such case in France resulted in the revocation of a data centre provider’s operating license after the $150 million facility was built.
As you can imagine, this is forcing operators to be very careful when deciding where to build new data centres, especially in Europe.
You have cases from places like Frankfurt, Germany, where the city’s data centre capacity grew from 50MW to 150MW in just three years. That is a rapid change. It prompted the local municipal council to introduce what it calls “the Frankfurt Data Centre Masterplan,” which aims to prevent the development of new data centre clusters. It changed urban planning rules to enable the strengthening of existing clusters while curbing further expansion.
It comes down to communication and engaging with communities to explain the benefits, speaking with leaders, and supporting local initiatives to build trust.
Beyond location, what other geographic considerations are influencing deployment strategies? Data sovereignty is a key concern. Many regions already enforce strict data laws and regulations that dictate regional controls, mobility standards, and protections. And these will continue to evolve alongside AI’s own advancements, making it vital to incorporate flexibility and scalability into physical infrastructure investments. This enables operators to feel confident in supporting any new topologies and data rates for next-gen servers and switches, even as regulatory landscapes change.
How are all these new factors affecting talent acquisition? Hiring remains a challenge, and the talent gap is significant. Roughly ten per cent of
data centre roles are unfilled, which is more than twice the national average for any given industry. And that is before accounting for expansion, which will only increase the demand for talent.
These are highly technical roles. Only fifteen per cent of applicants meet the minimum job qualifications, which means open positions can remain vacant for months, impeding any plans to scale rapidly. One approach that is proving effective is to offload complexity. Many are now leaning on offsite integrators to supply pre-populated, complete racks. This allows for a faster buildout pace, putting less strain on the technical expertise needed at new distributed data centre locations.
Newly developed products are also helping lessen deployment challenges. For example, smaller, multi-fibre connectors address densification issues, and multi-gang connectors, with over 100 fibres per single connection, enhance the speed and efficiency of deployment. Another strategy is to consolidate connection points by using higher fibre-count cables, which can significantly reduce the number of touchpoints and cabling congestion.
The challenge is that AI is forcing system designers and operators to rethink every layer of design and deployment. The areas we have touched upon are only the beginning. Finding the right deployment strategies and system architectures, hardware, and integration partners, along with the required talent, is paramount. Scalability, resilience, and efficiency are central to overcoming these challenges. How operators respond will shape the future of data centres for years to come.
https://www.molex.com/en-us/home
www.cieonline.co.uk Components in Electronics July/August 2026 31
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48
