COMMENT
In our digitally driven world, data has become an invaluable resource for businesses across all industries. Data enables intelligent products, services and operations. Data is also the unsung hero in today’s “age of artificial intelligence”. By 2030, the AI market will be worth more than $1.3 billion in revenue – growing 36.8% from the market size of $150.2 billion in 2023, and data is undoubtedly the catalyst driving this immense growth.
The current state of data centers
Data is not just an intangible asset stored in the virtual ether of the “cloud.” In reality, data and data storage are very much tied to the physical world. With the increasing introduction of artificial intelligence models, more data will be needed, which will then be stored in data centers.
These centers help keep data safe, accurate, available and, for all practical purposes, “alive.” Using switches, storage systems, servers, routers, and other equipment, data centers can store critical data sets around the clock. However, the extreme heat produced by the energy used to process and store the data causes overheating problems. Data centers therefore require energy-intensive cooling systems to ensure that equipment does not fail. Such failures can lead to loss of data, vital workloads and service interruptions, which have occurred in short order the Australian-based data center of a major technology company when a regional utility power outage shut down the center’s cooling mechanism.
Unfortunately, cooling comes at a high price, which represent almost 40% of the average electricity consumption of a data center. Many operators recognize the pressure traditional air cooling methods place on their finances and net zero commitments, which is why some are opting for liquid cooling systems. While liquid cooling is technically more energy efficient than air cooling, it can still have a negative impact on the environment and opens other doors to potential disruptions.
The dependence of data on water
When we talk about “liquid” cooling systems, in most cases we are referring to water. Just like living things, data needs water to survive, and so do AI models, software, and countless other technologies that rely on data.
Most data centers in the United States use utility-supplied fresh water, the same water supply humans rely on. THE an average data center uses 1 million to 5 million liters of water per day (paid item) and considering this 30% of the world’s data centers are located in AmericaThis means Americans are losing access to a significant amount of safe drinking water.
The United States is already experimenting with it alarming water shortage as a result of low rainfall fueled by climate change and increased demand from a growing population. Add in increased demand from data centers as operators try to cool facilities overrun with AI data, and America could experience a water crisis.
To make matters worse, America’s water supply faces an additional threat: cybersecurity attacks. Bad actors are increasingly targeting U.S. water infrastructure. The primary concern is the health and safety of Americans, but these attacks also threaten data centers and the technology that depends on the data. Hackers have historically targeted cooling systemsand will no doubt continue to do so by identifying weaknesses in data center water infrastructures, as well as security gaps regional water services serving data centers in the United States.
Two-pronged approach: sustainability and safety
So how can data center operators effectively store data critical to the booming AI market and virtually every aspect of the digital world, while limiting water consumption and protecting their water infrastructure?
From an environmental point of view, this will mean using other water sources in addition to the fresh water provided by American utilities. Some operators are exploring the use of wastewater, industrial water and seawater to reduce pressure on the United States’ depleting freshwater supplies. Improved liquid cooling system monitoring tools Monitoring your water consumption is also essential. And companies do it wisely using federal funds to explore efficient cooling options and should continue to collaborate with the public sector moving forward.
Additionally, operators should consider introducing new liquid cooling systems, as there are now more advanced designs that use less water. And as new data centers become necessary (and we know they will) companies should consider building these facilities in cooler climates as well as in areas with non-overloaded water basins.
From a cybersecurity perspective, many operators will need to prioritize defensive measures that specifically protect water infrastructure, which has likely not been a key security focus in previous years. This will require a strategic reset in which leaders identify water, energy and other external dependencies and extend risk assessments to environmental resources and other assets beyond the “logistics-based supply chain” that could disrupt business or operations . If companies have not already done so, they should also apply zero trust principles to all water infrastructure on which they depend.
Organizations need to prepare for a variety of scenarios: What if the data centers you rely on for business sustainability have a water shortage that forces them to shut down? Does your crisis plan consider a catastrophic loss of water, and therefore data? If water resources and infrastructure are not part of your security plans, chances are you could very well face these scenarios.
To avoid such potential threats, keep this simple mantra in mind: Define. To protect. Stick up for. Assess your risk so you can mitigate it and monitor and maintain your defenses.
That said, water utilities must also do their part to help prevent attacks that could severely impact critical data centers and compromise water supplies. Experts highlight a renewed focus on cyber resilience through public-private partnerships, improved centralized patch management systems, risk assessments, and interim controls to address immediate vulnerabilities in critical systems and legacy infrastructure.
Change is needed now
Data center operators face a considerable challenge. They must meet the need to store a surprisingly growing amount of data, while using a dwindling water supply to cool their facilities. Business, technology and sustainability leaders must work together to formulate strategies that not only protect the environment, but also protect their data.
Without water there is no future for humans or the life-changing technology we rely on. And make no mistake: We are potentially headed for a water supply crisis if data center operators, utilities, and the U.S. government don’t implement preventative measures now.