Kevin Dede: Good morning, Wes. Hi, David. Thanks for taking my question. I was – yes, I was wondering, I apologize if I sort of missed the boat on this one, but could you peel the onion back a little bit on the contracts themselves. I’m wondering, like, with Character A.I. for instance as an exclusive binding, is there any penalty for termination – early termination on either side?
Wes Cummins: It’s a standard contract. So we’ve walked through kind of how these – they’re structured and the first customer has – it’s these option periods and the first customers executed the options so we have the full contract for the $180 million. It’s a non-cancelable contract except for kind of standard contract terms if we fail to perform on the contract, but other than that, it’s standard – it’s not exclusive for either one of us. It doesn’t, obviously, doesn’t preclude us from doing business with anyone else and it doesn’t stop any – our customers from going and getting compute power elsewhere, but that’s the basic structure.
Kevin Dede: Well, reserve capacity sort of implies, Wes, and I apologize I read too much in this, but it implies that you’re providing capacity as a second-tier supplier. Is that not right way –
Wes Cummins: No. So there’s two models in this market, if you go and look at just the vernacular for the market. Reserve capacity means that the customer has reserved for that capacity exclusively from themselves, don’t think of it as like the reserves the Coast Guards or the Air National Guard reserves, right? It’s not – don’t think of it is that. They’ve reserved the capacity exclusively for themselves, and so, think of it as a take-or-pay contract. So they have to pay for it, whether they’re using it or not. And then as you move over the other model in this space is on demand, so people just pay for usage. And so if they use an hour, they use several thousand hours, they just pay for the hours they use. The on-demand market, the pricing is at a – call it, a 30% premium.
That’s not exactly that number, but a premium to the reserve capacity market because when you’re in the on-demand market, obviously, you have the risk as an operator of not having all of your capacity utilized. But the reserve contracts on – for our side are much safer because it’s basically take-or-pay guaranteed for the length of the contract. And generally, what we’re doing is signing contracts with the borrowers especially the first that pay us back for the equipment purchases in the life of the contract. So it’s doesn’t leave us thinking out. And then – so I think on the first one, we have two years on the contract and then we should have three to four years of useful life on the H100s maybe the innovation cycle leads up significantly, but the NVIDIA introduced the A100, I believe, in 2019.
And if you have A100s operating now, you can still book those for pretty attractive rates. So that’s kind of the life cycle of those.
Kevin Dede: Well, thanks Wes. Appreciate the color. Can you, I guess, dive a little bit deeper on exactly how much colo power you’re penciling in and maybe an idea sort of on the spreads there, what you’re paying for power. And then I guess, what you are charging your customers in sort of rough terms?
Wes Cummins: So which segment are you talking about, Kevin, are you talking about our HPC data center side?
