DATA CENTRES The liquid leap
Chris Coward, director of project management at BCS discusses how AI is driving last minute design changes and market uncertainty.
T
he proliferation of artifi cial intelligence (AI) has already revolutionised industries, from healthcare to fi nance, by enabling advanced data processing and decision- making capabilities. Central to this technological leap were AI data centres, specialised facilities that housed the hardware and software necessary for AI computations. Everyone in the data centre sector knew it was coming, but it is fair to say that most people underestimated the speed of its arrival and the size of the demand. As an industry, we had to once again demonstrate our ability to pivot and change in order to deliver the crucial specialist data centre facilities needed to support the growing proliferation of artifi cial intelligence. In particular, this meant having to adapt and embrace the technology that supported it. In fact, over the past few months, we supported clients who had taken the decision to ensure their facilities were AI-ready, despite the fact that they were already at various stages of design or even construction. The key issue was around cooling as AI required far more processing power than standard computing, which in turn needed more cooling. So, this meant fl ipping the previously
22 November 2024 •
www.acr-news.com
designed cooling systems from air-cooled to L2L (liquid to liquid). The irony here for those of us who had been in the industry for a while was that we had spent 10 years getting water out of our facilities. We had four clients that were changing their cooling systems. One had a facility that had already been built and the others had all the building permissions and were already quite far down the route of tendering out the construction. Although liquid cooling had been around for many years, it hadn’t been widely used in recent times with air cooling generally the method of choice to achieve the lowest PUE. This raised questions about its initial costs, environmental impact, lifecycle costs and overall ROI. All of these elements needed to be considered and factored into the project, in some cases retrospectively. For existing facilities, this might have meant adapting existing chilled water implementations to deal with AI, including air cooled, direct to chip and immersed solutions. This created unwelcome uncertainties in the market around whether these changes would need updated planning permissions, and the potential delayed construction start.
Download the ACR News app today
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
