00:00Meta, building a new gigawatt-sized data center in Texas to advance its AI efforts.
00:05Let's bring in Bloomberg's Kurt Wagner, who joins us now from Seattle.
00:08So, Kurt, put this into perspective for us.
00:11What kind of size data center is this, and how will this exactly advance what Meta already has currently?
00:20Yeah, Meta is building several large data centers to try and keep up with all of this AI demand that they are creating.
00:28This one, as you mentioned, will ultimately have the capacity, one gigawatt of capacity when it's done, expected to be completed in 2028.
00:36It is big. It is not the biggest thing they are building.
00:38They're also working on a data center in Louisiana that they expect could be as high as five gigawatts of capacity.
00:44So, while this is sort of the latest in a number of investments they're making around AI infrastructure,
00:50it is not the largest, but again, it shows that they are really, really committed to keeping up with OpenAI, Google, and others in this AI race.
00:58And how does this fit in with how much Meta has committed to spending on AI-related infrastructure projects?
01:06So, in the release this morning, they said that this particular data center in El Paso is going to represent about $1.5 billion of investment.
01:14So, certainly sizable, but again, nowhere near as big as this one in Louisiana.
01:20President Trump actually met with Mark Zuckerberg and then came out and said he thinks that data center in Louisiana could cost as much as $50 billion.
01:28So, this is big. It is not $50 billion big.
01:31But again, I think it's sort of just the steady drumbeat we've seen from Meta in terms of investing in AI infrastructure, and this is just the latest.
01:38Right. And I ask that because Meta has committed to spending, I think, up to $72 billion this year on AI.
01:44And this is something that investors look very closely at when the company reports earnings.
01:48It will do so on October 29th. Have investors become a little bit more forgiving in terms of the capex on AI?
01:55I think when there was a version of that number that was released in Q1, I think it was very eye-popping.
02:02That was back in January. We hadn't quite seen the level of deal-making that we've seen across the AI industry in big tech.
02:09And so, I think at that time, it was sort of like, whoa, what's going on here?
02:13It's now been several quarters since then.
02:16And I think that number, that $72 billion capex number you mentioned, is sort of accepted at this point because Meta continues to deliver on the ads business.
02:25And as long as they're continuing to print money from advertising, they can afford to go out and take these big swings on AI.
02:30And that's sort of the game plan right now.
02:32I think people are getting comfortable with this idea from them in a way that maybe they weren't 10 months ago.
Be the first to comment