The climate impact of artificial intelligence — both in terms of power consumption and all the electronic waste that gadgets create — is a legitimate, growing concern. Research from the University of Massachusetts Amherst suggests the process of “training” neural networks to make decisions or searching them to find answers uses five times the lifetime emissions of the average U.S. car. Not an insignificant amount.
For many in the data center sector, one of the most pressing concerns is that much of the world’s data center infrastructure operates in a manner that is financially suboptimal and environmentally unsustainable. If a data center is only using a fraction of the available power, then the capital investment that is tied up in inflexible power infrastructure is impotent. The question is, who is paying for that stranded capacity and unused space?
The electricity that powers appliances and lights homes also generates small magnetic fields that exist everywhere. A research team, headed by Penn State scientists, built a device that delivers as high as 400% higher power output when compared to other advanced technology when working with low-level magnetic fields similar to those seen in buildings and homes. The technology holds major implications for designing smart buildings, which will need self-driven wireless sensor networks to perform things like remote control of systems and tracking energy and operational patterns, the researchers said.
A study published in the journal Science last week, notes that while global data-center energy consumption has risen over the past decade, a predicted explosion in power consumption by data centers has not manifested thanks to advances in power efficiency and, ironically enough, the move to the cloud. Data centers accounted for about 205 terawatt-hours of electricity usage in 2018, roughly 1% of all electricity consumption worldwide. This represents a 6% increase in total power consumption since 2010, while global data center compute instances rose by 550% over that same time period.
A new study from Greenpeace and North China Electric Power University reports that in five years, China’s data centers alone will consume as much power as the total amount used in Australia in 2018. The industry’s electricity consumption is set to increase by 66% over that time. Buildings storing data produced 99 million metric tons of carbon last year in China, the study finds, which is equivalent to 21 million cars.
The IoT is a broad term that defines anything connected as part of a network, capable of both sending and receiving data. The majority of this network is based on sensors, and it’s these devices that will power Energy-as-a-Service. Current energy production focuses on a centralized grid and while it has served us well in the past, decentralized microgrids are emerging to prevent power disruptions, monitor energy usage and even transition into green energy.