CLICK ON COVER TO VIEW THE LATEST ISSUE

The relentless growth of artificial intelligence (AI) is fuelling a massive data centre revolution. Although this is generally a very positive sign that points to the long-term health of the sector, it brings with it significant issues that also need to be addressed.

The chief concern is whether enough existing data centres can cope with this enhanced demand. The bottom line is that many of them simply weren’t designed and built to accommodate AI and are limited in their ability to provide sufficient power and cooling for this type of compute. It means many owners and operators are now having to scale their data centres accordingly in order to have the robust digital infrastructure required.

This isn’t a one-dimensional challenge though. The vast amounts of data required by AI are seeing mass adoption of graphics processing units (GPUs), due to their capability to concurrently process multiple computations more efficiently.

Although GPUs are more power efficient per byte of data processed than central processing units (CPUs), the total power consumption is expected to rise. Data centres will therefore need to carefully manage environmental and regulatory concerns. Inside_Networks has assembled a panel of experts to examine whether data centres are scaling quickly enough to keep up with the demands that AI presents and whether they can do this while also cutting carbon emissions.

This issue also contains special features on intelligent building network infrastructures and testing and test equipment, with experts from Siemon, LMG, AEM and Sudlows providing fascinating insight into a range of subjects. These include the need for smart retrofit in existing buildings to meet environmental obligations, the benefits of high-speed Wi-Fi, the various network infrastructure configurations found in modern intelligent buildings and why a robust testing strategy is vital.

Rob Shepherd
Editor