I can't help but poke fun at a classic /r/PeopleLiveInCities/ map. It's not surprising that people more frequently experience power issues in places where there's a large demand for power; ie,...
I can't help but poke fun at a classic /r/PeopleLiveInCities/ map. It's not surprising that people more frequently experience power issues in places where there's a large demand for power; ie, large metro areas. It may or may not have to do specifically with AI, but even before the AI boom, we've known that data centers were going to continually expand, and we've also known that the US electrical grid is in dire need of a massive overhaul; mostly because "I'll spend half our GDP on transformers, substations, and powerlines" isn't exactly a sexy or popular campaign slogan.
It was always going to reach a boiling point, AI or not. Before AI it was crypto. Before it was crypto it was cloud computing. After AI it will be something else, but we're always going to be chasing new innovations that are likely going to use a massive amount of power. Unfortunately, until people start experiencing regular brownouts and electrical issues, fixing the problem will never be politically popular.
Anytime there’s a big swing in energy supply or demand it can create stress on the grid. With more granular-level readings of what is happening with harmonics on the residential level, it becomes easier to understand the risks to households, said Carrie Bentley, CEO of Gridwell Consulting.
“If you know it exists, it is easy to fix,” she said. “So if this is a problem, it is a nice problem.”
Some solutions are already in the works.
Most new data centers in Virginia require their own substation and transformer to be able to come online and this isolates them from nearby distribution circuits, limiting their impact on other customers, Dominion spokesman Ruby said. The Virginia utility is building a new transmission line into Loudoun County to address power quality and reliability constraints overall. Devices like filters and capacitors can help address harmonics issues around data centers.
I can't help but poke fun at a classic /r/PeopleLiveInCities/ map. It's not surprising that people more frequently experience power issues in places where there's a large demand for power; ie, large metro areas. It may or may not have to do specifically with AI, but even before the AI boom, we've known that data centers were going to continually expand, and we've also known that the US electrical grid is in dire need of a massive overhaul; mostly because "I'll spend half our GDP on transformers, substations, and powerlines" isn't exactly a sexy or popular campaign slogan.
It was always going to reach a boiling point, AI or not. Before AI it was crypto. Before it was crypto it was cloud computing. After AI it will be something else, but we're always going to be chasing new innovations that are likely going to use a massive amount of power. Unfortunately, until people start experiencing regular brownouts and electrical issues, fixing the problem will never be politically popular.
From the article:
Archive