- The UK data center boom is being held back by cabling bottlenecks, a survey of senior data center decision-makers has found
- AI is also progressing so quickly that data centers often require upgrades almost immediately after becoming operational
- Findings suggest UK data centers are not actually being built to perform
As the United Kingdom seeks to establish itself as a leader in AI tech, the buildout of data centers is encountering one key obstacle.
AI technologies are advancing so rapidly that developers are struggling to futureproof data centers to be able to handle the additional capacity or new computing requirements.
According to a survey by data center infrastructure specialist Onnec, data centers are struggling to navigate past cabling, latency, and rapidly advancing AI capacity requirements that take years to fix.
UK data centers are struggling to stay fit for purpose
Onnec found over three-quarters (77%) of UK data center operators were encountering critical bottlenecks in sourcing cabling to support their AI workloads, with 72% stating that the AI rush is creating additional cabling problems that could take years to fix. Furthermore, 37% of data center operators are encountering latency or bottleneck issues with AI training.
But the problems don’t end there. These trends are being worsened by the rapid development of AI technology. 65% stated that alongside these problems, their data centers are seeing AI capacity increase.
Moreover, 83% also stated that by the time their hardware is operational, AI requirements have progressed so much that upgrades are needed almost immediately to meet new demands and capacity. This compounds problems with cabling bottlenecks, with a further 33% unable to scale their data centers without significant cabling work.
In some cases cables have introduced significant project delays or additional costs, with 44% stating that insufficient upfront cabling planning had caused such issues.
“Everyone is focused on securing power and GPUs for AI, but the network connecting that hardware together is too often finalised only after those decisions have already been made,” said Matt Salter, Global Head of Data Centres at Onnec.
“The racks are set, the pathways are tight, and teams are left working around the design. You can have the power and GPUs in place, but they still need to communicate. If the network cannot support that traffic, the AI environment will not perform as expected.”
“The priority is to make cabling part of the design before layouts and pathways are fixed,” adds Salter. “Operators need enough capacity, access and flexibility to support more than the first hardware deployment. Otherwise, today's AI build-out will leave teams carrying out expensive work inside live facilities much sooner than expected.”
Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews, and opinion in your feeds.
