China’s Edge in the AI Race May Come Down to Power
Haley Zaremba

China’s secret weapon in the global race to lead AI innovation is its enormous electricity resources. While China quietly plugs away at ramping up its development and integration of large language models, Western countries are facing increasing energy security tradeoffs to keep data centers running.

The increased energy demand and grid capacity required to run AI’s ever-expanding computational load are putting enormous strain on power grids in Western countries. Power grids were already in a precarious position in the United States and Europe as electrification, and the breakneck addition of solar and wind energies have outpaced investment in supportive infrastructure. The result has been an increasing occurrence of major power outages and painfully high energy prices. 

“As energy demand surges and grid modernization accelerates, utilities and developers are dealing with the twin pressures of global trade uncertainty and grid reliability,” wrote Jeremy Furr, senior vice president of strategic sourcing at Stryten Energy, in a recent op-ed for Utility Dive.

But, paradoxically, China – the world’s biggest installer of renewable energies and the world’s first electrostate – is not having the same issues, or at least not with the same existential magnitude.

One key difference is that China has invested considerably in expanding and updating its power grids, matching grid capacity with demand much more successfully than its western counterparts. While China experienced a slew of regional outages early in the decade, the country hasn’t suffered any serious blackouts since then. 

All of this puts China in a very strong economic and geopolitical position. The relative abundance, stability, and affordability of energy in China could give the country a major leg up in the global AI arms race. “In terms of electricity capacity, the differences between China, the US and the EU are substantial, and China is clearly staying ahead of the curve across scale, structure and growth momentum,” Alberto Vettoretti, managing partner at consultancy Dezan Shira and Associates, recently told the South China Morning Post.

Moreover, China is now subsidizing energy for data centers after Chinese tech companies complained of energy costs “caused by the domestic semiconductors they are obliged to use, which are less efficient than Nvidia’s,” according to a new report from the Guardian. This subsidy is only one of many, many incentives for AI firms in China. “With dominance of AI now a core government policy, every city and region is offering incentives to AI start-ups,” Time reported this week. 

While China is still lagging behind the United States in terms of AI development and design, Chinese tech firms will likely catch up and surpass Silicon Valley in the not-too-distant future. The Chinese government’s AI+ Initiative, announced in August of last year, states its aim to “reshape the paradigm of human production and life,” and to integrate AI into 90 percent of China’s economy by 2030. 

For the second-largest economy in the world, that’s going to require staggering amounts of energy and grid capacity. But Beijing can also harness the power of AI to optimize the power grid. AI could be instrumental in efficiently regulating a grid that is increasingly powered by variable energy resources, and which is simultaneously faced with skyrocketing around-the-clock demand. Large language models can use supply and demand data to calculate minute fluctuations in supply and demand in real time, and do so more affordably than other computer models could.

“If AI models are well trained to give accurate forecasts on how much renewable electricity will be generated throughout the day and how much electricity will be needed at corresponding times, then the grid operator can decide how to balance the demand and supply in advance more efficiently and safely,” Xi'an Jiaotong-Liverpool University’s Fang Lurui told Reuters earlier this month. 

China is planning to comprehensively integrate AI into its power grid by next year.

By Haley Zaremba for Oilprice.com

Haley Zaremba is a writer and journalist based in Mexico City. She has extensive experience writing and editing environmental features, travel pieces, local news in the Bay Area, and music/culture reviews.

oilprice.com