How The Massive Power Draw Of Generative AI Is Overtaxing Our Grid

67,207
0
Published 2024-07-28
There’s more than 8,000 data centers globally, but it’s not nearly enough to keep up with the power needs of generative AI. One ChatGPT query takes about 10 times as much energy as a typical Google search. Training one large language model can produce as much CO2 as the entire lifetime of five gas-powered cars and use as much water as a small country. Even if we generate enough power, our aging grid is increasingly unable to handle transmitting it to where it’s needed. That’s why data center companies like Vantage are building closer to where power is generated, while the industry invests in alternate energy source and creative ways to harden the grid.

Chapters:
1:58 Chasing Power
6:06 Hardening the Grid
8:45 Keeping it cool
11:56 More work per watt

Produced and Shot by: Katie Tarasov
Edited by: Evan Lee Miller
Supervising Producer: Jeniece Pettitt
Animation: Christina Locopo, Jason Reginato, Emily Rabbideau
Additional Camera: Andrew Evers
Production Support: Lisa Setyon
Additional Footage: Apple, Arm, Getty Images, Google, Helion, Meta, Microsoft, Nvidia, OpenAI, UC Riverside

» Subscribe to CNBC: cnb.cx/SubscribeCNBC
» Subscribe to CNBC TV: cnb.cx/SubscribeCNBCtelevision

About CNBC: From 'Wall Street' to 'Main Street' to award winning original documentaries and Reality TV series, CNBC has you covered. Experience special sneak peeks of your favorite shows, exclusive video and more.

Want to stop worrying about money? Sign up for CNBC’s new online course Achieve Financial Wellness: Be Happier, Wealthier & More Financially Secure. We’ll teach you the psychology of money, how to manage your stress and create healthy habits, and simple ways to boost your savings, get out of debt and invest for the future. Start today and use code EARLYBIRD for an introductory discount of 30% off through Sept. 2, 2024: cnb.cx/3WiasGB

Connect with CNBC News Online
Get the latest news: www.cnbc.com/
Follow CNBC on LinkedIn: cnb.cx/LinkedInCNBC
Follow CNBC News on Instagram: cnb.cx/InstagramCNBC
Follow CNBC News on Facebook: cnb.cx/LikeCNBC
Follow CNBC on Threads: cnb.cx/threads
Follow CNBC News on X: cnb.cx/FollowCNBC
Follow CNBC on WhatsApp: cnb.cx/WhatsAppCNBC

#CNBC

How The Massive Power Draw Of Generative AI Is Overtaxing Our Grid

All Comments (21)
  • @saiedanam
    Honestly this sounds like a more reasonable origin story for The Matrix than "We burned the atmosphere to block the sun"
  • @pnketia
    This is why so many people are investing in home solar and battery backup systems and using the grid as a backup! Every year my energy bill goes up yet I use less energy, buy energy star rated appliances an run my home ac less despite record temperatures every year. Now that I have solar on my roof and multiple storage batteries in my garage I no longer rely on the grid and only use it as a backup.
  • @bars4745
    Most people are not aware that AI applications require data centers and high energy. Companies investing in AI are likely to face more energy and grid problems.
  • @johnyracercat
    So they want to get rid of gas engine vehicles to save the plant, while at the same time rolling out AI that is 5 to 10x worse than gas vehicles. Makes sense.
  • @alb12345672
    I am an EE, took a mechanical engineering class in the early 90s, was not too interested in it. But the professor was working on thermal chip cooling systems, from what I understand is common tech in laptops now. Your laptop chip heats a fluid in a sealed tube, then it boils off and cooled by a fan, turning it back to a liquid. It is taken for granted but without such a solution, a slim laptop would not exist. He pretty much stated without advances in thermodynamics and cooling, eventually all the EE tech will come to a dead halt. It is true.
  • @ap0ria
    Awesome Vantage Commercial!
  • @Kevin-oj2uo
    I dont understand why these companies cant invest the money on their own power which will make them profit. Why do regular people have to pay for that?
  • @Gabeyre
    AI companies should be forced to build and only use solar power generation and storage. Atleast if AI fails, pple can use the renewable energy infrastructure they leave behind.
  • @1ycan-eu9ji
    LLMs should all be behind paywalls, if you don't need it enough to pay for it then you shouldn't use it, that's the key thing, many people just don't wanna do google searches or whatever and instead ask chatgpt directly, that shouldn't be the case, we are turning a world that was already pretty hard to sustain with renewables into a much harder world to sustain with renewables
  • @mack-uv6gn
    Hopefully they are adding solar panels to these facilities to help with power usage. No government subsidies either, they can afford it.
  • @SevoraLF
    Report recently came out that said OpenAI is facing bankruptcy due to generative AI losses
  • this why so many people being charged more electricity. due to data centers energy usage.
  • At the 5:30 mark, there has to be a mistake, fusion energy from 2028? If true then this itself deserves a video.
  • @arnechino
    While the mass adoption of AI has transformed digital life seemingly overnight, regulation of its very physical impacts has not kept pace.
  • @nyx211
    The human brain uses about 10 to 20 Watts of power to function. Brain cells, like all cells, had to be energy efficient from day one. AI researchers don't consider things like energy usage or cost when designing their models.
  • @KeliK1
    I'll be darned! I had no idea. It makes sense. More processing results in more heating and more power usage which results in overtaxing the grid. Never heard of water cooling of computing components before either. I learned something new today. So AI is the new "gas guzzler". Maybe we should stop using it for the sake of our planet and think for ourselves instead of making a computer do so.
  • @ScrapKing73
    I don’t understand why any energy hungry building with a large unobstructed roof isn’t covered in solar.
  • @stargazer3828
    These data centers are already considering switching to SMR's (small modular nuclear reactors) to supply their energy needs because the cost per kilowatt is much cheaper when you factor in the overall costs versus the power output.