Fluke Corporation study finds two-thirds of data-center operators are not fully prepared for AI’s growing electrical demands, as power-quality issues and skills shortages threaten expansion
-
asserted
issues → find → expansion
78 percent experienced electrical or power-system issues contributing to disruption, delays, downtime or serious near-misses in the past year.
asserted
percent → experience → year
- Only 33 percent say their electrical infrastructure is fully ready for higher-density AI workloads.
asserted
infrastructure → say → workloads
- 34 percent say electrical skills shortages have delayed infrastructure upgrades or new capacity.
asserted
shortages → say → upgrades
- 29 percent say power-quality issues have reduced their ability to deploy or operate AI/GPU workloads.
asserted
issues → say → workloads
The global race to build and expand AI data centers could face a major setback, with new research from Fluke Corporation raising questions about whether the infrastructure behind the AI boom can deliver on the industry’s ambitious growth plans.
uncertain
infrastructure → build → plans
While operators race to secure additional grid capacity, weaknesses in the electrical infrastructure inside data centers are threatening to delay the deployment of new AI workloads and disrupt existing operations.
asserted
weaknesses → race → operations
Just one in three data center operators (33%) say their existing electrical infrastructure is fully prepared for higher-density AI workloads, while nearly eight in ten (78%) have experienced electrical or power-system issues contributing to disruption, delays, downtime or serious near-misses in the past year.
asserted
ten → say → year
At the same time, six in ten (60%) say recruiting qualified electrical technicians has become more difficult, with 34% reporting that skills shortages have delayed infrastructure upgrades or new capacity.
asserted
shortages → say → upgrades
Together, the findings suggest that the challenge for the industry is no longer simply securing power from the grid, but ensuring data centers have the infrastructure and expertise to turn that power into reliable computing capacity.
asserted
centers → suggest → capacity
Commenting on the findings, Parker Burke, President of Fluke Corporation, said: “The industry has spent years asking where the power will come from.
asserted
power → comment → where
But our findings show that the focus should also be on how effectively data centers can use the power they already have.
asserted
they → show → power
As AI drives unprecedented demand for computing capacity, the electrical infrastructure inside data centers is becoming a critical bottleneck.
asserted
infrastructure → drive → centers
Without addressing this last-mile challenge, the industry risks having the power to expand, but not the infrastructure to deliver.
asserted
industry → address → power
The consequences are already being felt.
asserted
consequences → feel → ?
More than one in four data center operators (29%) say power-quality issues have reduced their ability to deploy or operate high-density AI or GPU workloads, while 29% report that electrical problems are only identified after performance or equipment failure, or downtime, has occurred.
asserted
failure → say → workloads
Limited visibility into critical electrical systems may be adding to the challenge.
uncertain
visibility → add → challenge
Fewer than half of operators (45%) report real-time visibility across those systems, potentially affecting their ability to identify and address problems before they impact operations.
asserted
they → report → operations
The internal infrastructure challenges identified by Fluke come alongside wider constraints on electricity supply.
asserted
challenges → identify → supply
Three-quarters of data center operators (76%) say access to sufficient power and grid capacity is moderately or severely limiting their ability to expand, while operators report an average 12-month wait to secure additional grid connection or power capacity for AI and GPU-related expansion.
asserted
operators → say → AI
While securing additional electricity remains a major challenge, the findings highlight the importance of ensuring that existing electrical infrastructure can reliably support growing power demands.
asserted
infrastructure → secure → demands
Parker Burke continued: “Grid constraints are largely outside an operator’s control.
asserted
constraints → continue → control
What is within their control is how effectively they monitor, maintain and manage the electrical infrastructure they already have.
asserted
they → monitor → infrastructure
Greater visibility into electrical systems and a proactive approach to maintenance will be increasingly important as operators look to maximize existing capacity, reduce disruption and support growing AI demand.”
asserted
operators → look → demand
About the survey
The study, commissioned by Fluke on behalf of Censuswide, surveyed 1,300 data-center industry operators and professionals across six countries, namely the US, UK, Germany, France, India and the Kingdom of Saudi Arabia to understand the power, infrastructure and workforce challenges facing the industry as AI demand accelerates.
asserted
demand → commission → industry
About Censuswide
Censuswide is an international market research consultancy headquartered in London, with offices in New York, Dubai, Bristol, and Glasgow.
asserted
Censuswide → headquarter → York
Our research supports insights, communication, PR and marketing teams across consumer, B2B/Corporate and Healthcare sectors.
asserted
research → support → sectors
Find out more here: https://censuswide.com/
About Fluke
Founded in 1948, Fluke Corporation is the world leader in compact, professional electronic test tools and software for measuring and condition monitoring.
asserted
Corporation → find → measuring
Fluke customers are technicians, engineers, electricians, maintenance managers, and metrologists who install, troubleshoot, and maintain industrial, electrical, and electronic equipment and calibration processes.
asserted
who → install → equipment
###
FLUKE is a registered trademark of Fluke Corporation.
asserted
FLUKE → register → Corporation
What are the biggest electrical infrastructure challenges facing AI data centers?
asserted
challenges → face → centers
A: AI data centers face growing challenges around electrical capacity, power quality, infrastructure reliability and access to skilled electrical technicians.
asserted
centers → face → technicians
As AI workloads increase power density, operators must ensure that existing electrical systems can safely and reliably support higher demand.
asserted
systems → increase → demand
Fluke research found that only 33% of data center operators consider their electrical infrastructure fully prepared for higher-density AI workloads.
asserted
infrastructure → find → workloads
Q: How does poor power quality affect AI data center performance?
asserted
quality → affect → performance
A: Poor power quality can cause equipment malfunctions, overheating, unexpected shutdowns and downtime.
asserted
quality → cause → malfunctions
Issues such as voltage fluctuations, harmonics and electrical disturbances can affect the reliability of critical data center systems.
asserted
Issues → affect → systems
Fluke research found that 29% of operators say power-quality issues have reduced their ability to deploy or operate high-density AI or GPU workloads.
asserted
issues → find → workloads
How can data centers improve electrical reliability and reduce downtime?
asserted
centers → improve → downtime
A: Data center operators can improve electrical reliability through regular electrical testing, power-quality monitoring, thermal inspections and preventive maintenance.
asserted
operators → improve → testing
…and 9 more, not listed.