⚡ AI datacenter demand needs double-to-triple current US electricity generation

“This generation's Manhattan and Apollo Project is energy production scaling in the United States. That's it. All the other shit flows.”

“Yeah, look at the number of times that Elon has said this. Like it's like the amount of energy that's available if we just harness it from the sun and redirect it with some storage, would solve all of our problems. But the problem is that we're at like just a little over a terawatt. We're inching along in gigawatts, whereas China is in the terawatts, adding terawatts.”

“The thing I do watch the most, Sacks, and that I do think, you and I have talked about this, that I do think is the critical linchpin for all of those other points of abundance being unleashed is the energy equation.”

“Well, I agree that you need to build out power, but I think the administration is committed to doing that. Drill baby, drill.”

“because we are rapidly scaling with regard to our own energy production. You have to be an energy intensive nation in order to produce AI. And so the United States has to play this game.”

“And you see that on the power side, right? You're going from 100 megawatt datacenters to 300 megawatts to, we're just starting to now see the first gigawatt power datacenters. I don't even think they're live yet, but this is where they're trying to get to. And I don't think it's beyond the realm of possibility that we could be at five or 10 gigawatt datacenters in the next, I don't know, several years.”

“But then underneath that AI thing, we have these two very complicated problems. One is a meaningful energy problem. I've been talking about this now incessantly. We don't have enough electrons. We just don't.”

“We know that AI needs a tremendous amount of power. And whatever you thought you knew, actually, you had to re-underwrite because what Elon has shown is actually you need to now create these mega clusters, 100,000 GPUs going to a million GPUs. All of the power forecasting that we have is miscast. It doesn't even account for this, number one. Number two, there are 35,000 applications into FERC to get approved to generate electricity. 35,000, that's going to meander through some administrative rigmarole. There is a five-year delay to get a gas turbine into America and online.”

“We have to win some of these key building block technologies, we want to win in chips, we want to win in AI models, we want to win in applications. He said we need to build, we need to unlock energy for these companies.”

“That's super interesting, Travis. That's like, it's almost like the idea that we all talk about today is data centers and data centers need their own power substations.”

“But then there's like this dark horse that nobody's talking about, which is, it's called electricity, it's called power. And all these vehicles are electric vehicles. And if you said, yeah, I just did some like quick back of the envelope calcs, if all of the miles in California went EV ride sharing, you would need to double the energy capacity of California.”

“But just today it looked like President Trump overturned it and now it's full seam ahead. If you have it permitted, then it's all about just get the power, however you need the power, which I think is the smart decision. Back to your earlier point, we are in a race here.”

“If you think of the GPU as the atomic unit of the future, right, and the key competitive advantage and the key kind of ingredient, you can't power a GPU without a data center, and you can't use a data center without power. So I do think at the end of the day, it really comes down to the grid and it comes down to power. So I think that's kind of constraint number one.”

“which is energy, electricity capacity buildout in the United States because it's critical. Building AI infrastructure means that we need to build energy infrastructure.”

“I mean, more nuclear, more better”

“But for me, it has to be such a priority that we accelerate a nuclear energy rollout in the United States, because that's what China is doing”

“I think by the time that you are rate-limited by energy, you'll have a plethora of solutions.”

“there is no way to meet the energy demand without nuclear”

“And the downstream is they can power more H100s, run bigger jobs.”

“We talk a lot about wanting to onshore manufacturing in the US., build out new manufacturing technologies, but ultimately all of these industries, particularly AI., are going to be driven by the cost of power.”

“The amount of electricity that is like double what we have right now and even triple what we have right now, they are, it's incredible how much they need to be the leader. And we're going to have to be able to do that.”