Kush Bavaria

"Look, we want to bring transparency and create an exchange for compute as a commodity. We believe compute will turn into every other commodity has existed. If you look at oil markets, electricity markets, any other commodity that exists has a market. It's traded right behind us. But compute doesn't have a sort of economy that exists around it."

"Look, I think compute has become more and more of an asset class. And if you look at it, every other commodity market such as oil, electricity, as you've mentioned, has a way to sort of transact and exchange that resource. And so, we're building the rails to be able to transact and exchange compute, like you do with electricity."

"I think finance of these data centers is just as much of a constraint as other resources are. If you don't have the money, if you don't have the tools to get a loan, to get financing for your data center project, there's no way that's going to be built."

"capital is the next bottleneck that's going to take place in data center economy. Look, we have power now. We have the chips. What's the next sort of thing? What's blocking? And really, if you look down to it is capital."

"And when there's uncertainty, there needs to be created a market so people can express their own opinions onto that space."

"I think inference is going to take a lot of workloads on compute. If you look how inference demand is shaped, it's very volatile. I don't know when I'm going to click the button. I don't know when I'm going to generate code. And so, it's creating the same way. It's like you consume electricity in a very volatile way. And I think inference demand is also very volatile, where you compare it to training where you know exactly when you're going to train the model."