AI forces data centers to use more energy. Software can help

Tech giants are racing to avert a carbon time bomb caused by the massive data centers they are building around the world.

A technique pioneered by Google is gaining traction as increasingly power-hungry artificial intelligence comes online: using software to search for clean electricity in parts of the world with excess sun and wind on the grid, then ramping up data center operations there . This could reduce carbon emissions and costs.

There is an urgent need to understand how to manage data centers to maximize the use of renewable energy, said Chris Noble, co-founder and CEO of Cirrus Nexus, a cloud computing manager that leverages data centers owned by Google, Microsoft and Amazon. .

The climate risks triggered by AI-driven computing are far-reaching and will worsen without a major shift from fossil fuel-based electricity to clean energy. Nvidia Corp. CEO Jensen Huang said artificial intelligence has reached a “tipping point.” He also said the cost of data centers will double within five years to fuel the spread of new software.

According to the International Energy Agency, data centers and transmission networks already account for up to 1.5% of global consumption. Together, they are responsible for emitting as much carbon dioxide as Brazil each year.

Hyperscalers, as the largest data center owners such as Google, Microsoft and Amazon are known, have all set climate targets and are facing internal and external pressure to meet them. These ambitious goals include decarbonizing their operations.

But the rise of artificial intelligence is already devastating these goals. Graphics processing units were central to the emergence of large language models and use more electricity than central processing units used in other forms of computing. According to IEA estimates, training an AI model uses more energy than 100 households in a year.

“The growth of artificial intelligence is far outpacing the ability to produce clean energy,” he said.

Additionally, AI power consumption is volatile and more like a sawtooth graph than a smooth line that most data center operators are accustomed to. This makes decarbonization a challenge, not to mention ensuring grid stability.

The growth of AI is driven by North American companies, which keep computing power – and energy consumption – concentrated there, said Dave Sterlace, account manager for global data centers at Hitachi Energy. This is a trend that was not expected two years ago.

To reduce data center CO2 emissions, hyperscalers and other data center providers have financed large amounts of solar or wind farms and used credits to offset emissions. (In the case of credits, some have failed to have a significant impact on emissions.)

But this alone will not be enough, especially as the use of artificial intelligence is increasing. That’s why operators are turning to the strategy used by Google, a unit of Alphabet Inc., called load shifting. The idea: reduce emissions by revolutionizing the way data centers work.

Today, most data centers try to operate in a “steady state,” so that power consumption is fairly stable. This leaves them at the mercy of the grid they are connected to and whatever the daily mix of natural gas, nuclear and renewable energy production is given the lack of transmission lines between the regions. To break their dependence on dirtier grids, tech giants are looking for opportunities to shift daily or even hourly operations of data centers around the world in a bid to absorb excess renewable energy production.

Google has launched the first effort to match energy consumption in some data centers with zero-carbon energy on an hourly basis, in an effort to run its machines on clean energy 24/7. No one has yet fully achieved this objective. And, to be sure, the strategy of moving cargo around the world could be complicated by countries pushing for data sovereignty policies that attempt to limit and safeguard the flow of data across borders. But what Cirrus Nexus and Google are testing could still be a critical piece of the puzzle to reducing emissions.

Manhattan-based Cirrus Nexus scans the world’s power grids and measures emissions in five-minute increments to find the least polluting computing resources for itself and its customers in industries ranging from pharmaceuticals to accounting. The company had a chance to put this research into practice last summer.

The Netherlands found itself in the midst of its sunniest June on record, sending the cost of solar energy plummeting on the grid. This made it cheaper and less carbon-intensive to run the servers. Cirrus Nexus then moved its computing load to California once the sun set in the Netherlands, allowing it to tap into solar power as it came online for the day in the Golden State.

Chasing the sun from Europe to the US West Coast and back, according to company data shared with Green Bloomberg. Making operations flexible to do so comes with both benefits and risks.

Being able to pursue zero-carbon megawatts of backup can help reduce stress on grids, such as during a heat wave or freezing winter storm. But data centers must work with utilities and grid operators because large swings in demand can throw electrical systems into chaos, increasing the likelihood of blackouts. Dominion Energy, which is seeing increased demand for data centers at its utility in Virginia, is working on a program to leverage load shifting at data centers to relieve stress on the grid during extreme weather conditions.

In recent years, Google and Amazon have tested the shift in using data centers for their own operations and for customers using their cloud services. (Cirrus Nexus, for example, uses cloud services offered by Amazon, Microsoft and Google.) In Virginia, Microsoft signed an agreement with Constellation Energy Corp. that ensures that more than 90% of the energy for its data center in the area it will be zero-carbon energy. Reaching 100%, however, remains a formidable goal for this and other hyperscalers.

Google’s data centers run on zero-carbon energy about 64% of the time, with 13 regional sites running at 85% and seven at just over 90% globally, said Michael Terrell, who leads the Google’s 24/7 carbon-free energy strategy.

“But if you don’t replace fossil resources, you won’t fully meet your climate goals,” Terrell said.

Sign up for the Eye on AI newsletter to stay up to date on how artificial intelligence is shaping the future of business. Sign up for free.

Source link

Leave a Reply

Your email address will not be published. Required fields are marked *