Building the intelligent data centre

Key findings from STL Partners’ global data centre intelligent operations survey (September 2026) reveal the evolving value for intelligent operations.

Overview of survey respondents

100 respondents by organisation type

1

The value of intelligent operations is shifting from resilience to optimisation and efficiency

Today

Main reason for using smart operations systems

28%resilience and reduced downtime
10%energy efficiency

In three years

Greatest source of operating cost savings

48%energy and cooling
15%reduced downtime and incidents
2

Operators want their data centres to run with greater autonomy and expect investment to rise to achieve this

Today

Operators see considerable headroom for progress and would already accept more automation than they currently run, across every deployed system.

Running now Would be comfortable with

In three years

That appetite is reflected in investment plans.

81%

expect to increase investment in software, control and monitoring over the next three years

3

Integration is the major hurdle to unlocking the value of intelligent operations

Integration is the top barrier

What is the single greatest barrier to more value?

It is also the top requirement for greater autonomy

What is needed for more autonomous operations?

4

Connecting cooling and IT data is becoming critical for optimising high-density operations

Today….

48%

say the biggest limit on cooling data is that it isn’t well connected to IT load or capacity data

….As AI and HPC densities rise

70%

expect to integrate IT workload data more closely with power and cooling

Recommendations

Integration must be tackled today to enable the intelligent data centre of tomorrow

Define data flows and ownership before procurement

Before buying or upgrading systems, establish which IT, power and cooling systems need to share data, and who owns each connection.

Favour open interfaces and backwards compatibility

Specify up front your requirements for open interfaces and compatible protocols and encourage vendors to support mixed and legacy estates.

Budget and test for integration explicitly

Account for integration effort in investment cases, rationalise overlapping systems and test new and existing systems together before implementation.

Build towards a single pane of glass

Ensure each investment strengthens one unified view across IT, power and cooling, building a more platform-centric approach rather than siloed, single stacks.

Cross-system integration is still a challenge. We’re often logging into one system for water, another for power, and then relying on separate spreadsheets as well.

DC operations manager, colocation and hyperscale operator, US

Find out more in our report

Building the intelligent data centre: unlocking resilience and efficiency