Data centers are making the grid more efficient, not less.
- Energy Secretary Chris Wright : “Ultimately AI is not going to make electricity more scarce and expensive in the country, it’s going to make electricity more abundant, more reliable, and ultimately more affordable.”
- According to Goldman Sachs, Nvidia chip compute power increased 15X while energy demand increased just 2X — proving the AI industry is delivering exponentially more computing power without a commensurate spike in energy use.
- As data centers grow larger, they become more energy-efficient per unit of compute — using less electricity per calculation — while enabling the density and speed that advanced AI applications require. Bigger data centers are better for consumers, the economy, and the environment.
Data centers are not leading to higher utility costs, and, in many cases, can reduce energy bills for everyday customers.
- Data centers represent just 4% of America’s total electricity demand — yet they power Americans’ daily digital lives, from banking and healthcare to communications and daily streaming habits. By comparison, residential air conditioning systems consume about 30% more power than data centers today.
- Studies have found that the addition of large loads to the power grid, such as data centers, can contribute to lower retail utility prices at the state level. A study conducted by Charles River Associates found that “where rates have increased most, data centers were not the cause.”
- S&P Global found “no clear correlation between the change in state-level electricity prices over the last five years and the change in state-level data center power consumption” — debunking the claim that data centers are responsible for rising electricity costs across the country.
- Duke Energy reports that each gigawatt of data center capacity signed under a 15-year contract generates up to $1 billion in savings for the broader customer base.