OpenAI, a global leader in artificial intelligence (AI), said on Monday,If the U.S. wants to stay ahead in the AI race, it needs to significantly increase its investment in energy capacity. In recent months, the startup has signed a series of ambitious deals to build out AI infrastructure, which will require massive amounts of electricity. At a time when the U.S. power grid is already overwhelmed, these huge data centers will push the limits of local power supply.
OpenAI said in a blog post on Tuesday that the power needed to maintain dominance in the field of AI far exceeds the current supply capacity of the United States, and the growing power gap is threatening the United States' leadership in this technology.
OpenAI submitted an 11-page document to the White House Office of Science and Technology Policy encouraging the United States to commit to building 100 gigawatts of new energy capacity per year.
A gigawatt is a unit of measurement for electricity. Relevant analysis shows that 10 gigawatts is approximately equivalent to the annual electricity consumption of 8 million American households.
OpenAI noted that the U.S. added just 51 gigawatts of electricity last year and warned that this puts the U.S. at risk of falling behind.
If computing power is the "first half" of the AI competition, then electricity is the key to determining who wins the second half of the AI competition.
Goldman Sachs recently pointed out that the power consumption rate of AI server clusters far exceeds the pace of power grid expansion, and power supply may become the biggest bottleneck in the AI era. The bank believes that the key to determining who can build the next wave of data centers is not faster chips, but more creative power financing solutions.