Energy rivalry between Artificial Intelligence and Bitcoin mining is heating up. As tech firms improve AI, they compete with Bitcoin miners for energy. This competition reshapes US energy consumption, as both sectors have driven unprecedented electrical demand.
AI data centers are leading in the race for energy consumption. These power-thirsty projects have forecasts indicating they will consume anywhere from 85 to 134 TWh of electricity annually by the year 2027. This is roughly equal to the yearly energy consumption of Norway or Sweden and demonstrates how massive an amount of power needs to be utilized in order to execute complex AI models such as ChatGPT.
Each one of these models runs on large farms of servers, and to run ChatGPT for every search conducted by Google, estimates are that the number needed would be over 500,000 servers, with estimated consumption of about 29.2 TWh annually.
According to estimates, Bitcoin mining uses 120 TWh of energy annually. Bitcoin mining consumed 0.4% of global electricity last year, which is a massive figure. Analysts expect AI to outperform Bitcoin miners in energy demand by 2027, shifting 20% of their power capacity to AI.
Competing For Resources
With the growth of both AI and Bitcoin mining, they are increasingly competing for some of the same energy resources. Competition is soaring, with major technology companies such as Amazon and Microsoft aggressively pursuing energy assets that, until recently, had been controlled by crypto miners.
Competition is warming up wherein, for some mining operators, cash is made through leasing and selling power infrastructure while for others, the risk of losing access to the electricity that keeps them in business becomes a reality.
This makes for such a mad scramble for energy that data centers are projected to gobble up as much as 9% of all US electricity by the end of the decade, more than doubling their current consumption.
Curiously, while crypto mining relies more on renewable energy sources-as it gets approximately 70% of its energy consumption from green sources-AI data centers mostly depend on fossil fuels.
Image: AsianInvestor
This therefore presents disparities in the views on the sustainability of both technologies. As AI demand continues to see an upward movement, tech companies weigh their carbon footprint against other source alternatives, including nuclear power.
The Road Ahead
The future of energy consumption in the tech industry is highly unpredictable. While AI continues to push the limits, its appetite for energy will see a corresponding increase. And unless efficiency starts to outpace growth radically, the environmental consequence is sure to be dire.
According to the International Energy Agency, the combined energy consumption of AI and Bitcoin mining would surge to 1,050 TWh by 2026-a quantity of electricity needed by an entire country.
But one question will always remain, as with most high-stakes energy races: will AI and Bitcoin mining be able to coexist, as in not sucking up all the earth’s resources?
How that is accomplished depends on how well these industries are able to innovate and adapt to the rising tide of sustainable energy solutions. As they race for power, the future of technology and the environment is hanging in the balance.
Featured image from Ken O./LinkedIn, chart from TradingView