Skip to playerSkip to main content
  • 10 hours ago
Transcript
00:00I want to just dwell a little bit on Core Scientific because you're very important to
00:04Core Scientific. You are their only customer. You'd hope to bring them on, keep using their
00:10compute. What happens now? How solid is that partnership? Oh, I think, first of all, thank
00:15you for having me and happy Halloween. Happy Halloween. So the partnership is very, very
00:22solid. You know, we continue to be a consumer of the services they provide. We continue to be a
00:29consumer for the next 10 or 15 years with a series of extensions of the infrastructure
00:38that they provide and that we require in order to deliver our products. So, you know, the
00:43relationship is in good shape. The shareholders voted down the acquisition yesterday. And,
00:55you know, from a strategic perspective, I think everybody on both sides of the fence really felt
00:59like it made sense. And it really just came down to price. And for us, you know, we put out a bid where
01:08we're, as you said, we're quite acquisitive. We've bought Marimo yesterday. We bought weights and
01:14biases, monolith, open pipe. We've really been buying and building the AI cloud. And there is a price
01:24point that made sense for us to move forward with that transaction. And we have a plan. And we are
01:31going to be very disciplined around that plan. And, you know, at approximately 10%, it made sense. And
01:38if it's going to be above that, then we'll continue to use them as a vendor.
01:44Well, Core Scientific now tries to diversify its own customer base other than just you.
01:49Does that give you any worry?
01:50No, not at all. Look, you know, our footprint within the Core Scientific ecosystem is about,
01:58I think it's about 580 megawatts worth of infrastructure. This quarter, we've signed
02:05over 600 megawatts of data center infrastructure outside and, you know, exclusively outside of
02:14Core Scientific. So, look, you know, they're going to run their business.
02:18You know, we hope that they are a strong operator. It's important to us that they are. We are hopeful
02:25that they will continue to invest in their business and their ability to execute and deliver
02:31infrastructure. And, you know, we continue to work with them on a go-forward basis, just as we have
02:35for the past five years.
02:36Look, they were a Bitcoin miner turned AI supplier. You two were that. There are some others out there
02:40as well, Tara Wolf and the like. Would you look to acquire any of those?
02:44So, I've always talked about acquisitions as strategic and opportunistic. My view is acquisitions
02:53like Marimo or Weights and Biases, Monolith, those are strategic acquisitions. They move the company
03:00and broaden our software solutions. The acquisition of a infrastructure provider is a opportunistic
03:10acquisition. We're currently building data centers within CoreWeave from the ground up to solve this
03:17problem of adding additional control over the infrastructure. So, you know, we've got a data
03:23center in Pennsylvania and Lancaster that we're building. We have a data center in New Jersey and
03:27Kenilworth that we are building. And so, you know, we're always, you know, reviewing the
03:34opportunities that exist within those two buckets, the strategic and the opportunistic bucket. And
03:39we're open to looking at things that move our company forward. But once again, it's got to be at
03:45the right price.
03:46Right price. Do you have to raise more capital to keep on building out your data center offering?
03:51So, you know, CoreWeave has been at the tip of the spear of raising capital
03:57for the AI build out, right? Like we were the first ones to do the GPU-based
04:05backed lending products. And our growth continues to rage along. I mean, it's just amazing how fast
04:16we're growing, how much interest there is for continued build out of our product, of our software
04:23solution and delivery to a broader and broader base of clients. And so, you know, we'll continue to
04:29raise capital to support that activity just as we have.
04:31And the reward is big enough. You know, the Gill lawyers of this world at DAO Deverson still saying
04:35the capital structure doesn't make him happy because the amount of 5% you get back on a 9% investment is,
04:40are you managing to rectify that?
04:42I, first of all, I just fundamentally don't agree with his analysis. So we can, we can start there.
04:47But more importantly, you know, we have built a business that generates great returns and provides
04:57the infrastructure that the world needs to build and deliver artificial intelligence. We think we
05:02have a great business plan. We have a very large and diversified shareholder base that agrees with
05:10us. And we're going to continue to expand on the business that we've built.
05:13And look, you're diversifying the customer base. I think Meta, it's interesting how Meta is getting
05:17a bit beaten up for the amount that they're plowing into AI data centers, whereas we're
05:21congratulating the likes of Amazon and indeed Google. What do you make of these anxieties around AI
05:26infrastructure build out?
05:27You know, it's interesting. When, when, when a big investor gets beat up for investing large sums of
05:36money into the infrastructure that we're going to have to provide, feels kind of like that's pretty
05:42bullish for us. We're going to have to provide that infrastructure. Look, Meta has a strategy.
05:48They have been incredibly disciplined around executing on that strategy. They are going to
05:56build their AI solution. And we are really excited about adding them to our long lists of very large,
06:05very active customers that are, that are really going to define what compute looks like and what the
06:10world looks like for the next 50 years.
Be the first to comment
Add your comment

Recommended