The Trillion Dollar AI Market: How Compute is Becoming a Commodity

Let's get to the big story.

Breakdown.

While AI has evolved into a $1 trillion market by enterprises and developers have long faced a major risk the highly volatile cost of GPUs, the computing power that runs AI models, and now the financial tools are emerging to commoditize the space.

Ice is partnering with financial infrastructure firm to launch US dollar denominate futures based on Orange's proprietary indexes and at the same time. has backed a landmark on chain institutional block trade on poly market while joining us live here at the New York Stock Exchange to discuss what all of this means is Kush Bavaria, co-founder & CEO of ORNN, great to have you here.

Thank you so much for joining me.

Great to have you here as well.

Well, of course when we're looking at the US equity markets, we continue to hear and see record highs here, but this does come on the focus on AI both as a technology.

As well as an investment theme.

So when it comes to what you're doing at Orn, give us an idea of compute as a commodity class, the how and the why.

So I can explain it very simply.

We basically build indices that track the price of compute.

You can think of it very similar to WTI or BRE as we talked about earlier.

And in that sense we track compute as a commodity of the future, and we believe that this will become a commodity just like oil, natural gas, or electricity traded.

Yes, and I'm sure you've been getting this question a lot, but when it comes to the H100 index, tell us about the data that you actually source and where it comes from.

So we run our own exchange, but we also work with every single broker and other exchanges in the market, and you can think of these as spot exchanges, the same way that S&P builds their natural gas indices or WTI crude is built off of Cushing, Oklahoma's oil reserves.

Yeah, so when it comes To the work that you're doing, there has been a lot of focus on GPUs.

So what have you been seeing since you've launched the index?

I think we've been seeing crazy high demand right now for GPUs.

There's an insatiable demand from both AI companies and different startups that are looking to procure GPUs, and the supply side just hasn't caught up yet, so prices are skyrocketing for even chips that are about 6 to 5 years older.

And of course you're here at the New York Stock Exchange.

You recently launched an announcement with ICE.

So tell us what's going on.

We're working with ICE to launch futures in the US and along with ICE's other futures markets.

Those will set to go live very soon, so we're excited to work with them on this.

Yes, and also another thing that I want to ask you about, there's been a lot of focus on prediction markets here.

So give us your take on what's folding and unfolding from a regulatory perspective.

So I think prediction markets have done an incredible job so far like we work with Poly Market and Cai at this point.

Cai just recently had their HRPS launch within the US, so I think it's Excited to see prediction markets more and more institutionalized, and we think there will be a natural place where compute will be traded as well.

And for people out there who are watching right now, they might not be familiar with this demand that we're seeing in AI for compute for GPUs.

What do you expect to see in terms of the growth?

I think the easiest way to see it is you look at Nvidia stocks today.

See how it's hitting record highs almost day over day, and you look at that and you say, OK, what is the commodity that Nvidia is producing?

It's these H100 chips, it's these B-200 chips that are then plugged into neo clouds such as Coreweave, Nebius, and other sorts of neo clouds, and we think that the price of compu will continue to increase and will sort of behave like a commodity market in a sense going into the future.

I do want to get your take on what you make of liquidity that you're seeing in the market.

I think liquidity is starting off very strong early on.

It's going to exponentially grow.

We've just seen month over month tripling, doubling, honestly quadrupling this month, so liquidity is definitely increasing a lot and just the compete space in general.

Yes, and there are other players out there, not just here in the US, but also abroad that are looking to capture what you're working on right now.

So what is the Long term vision when it comes to.

I think look, we want to focus on building this market in the US.

We believe that America has the strongest financial markets, and today we're the only company that's based in the US that's building this for American markets.

So we're proud to work with both America in general to advance this AI race with China.

And when we're talking about the AI race, Kush, we all know that a lot of time and investment needs to go into this, as we're hearing from the big players. including the Mag 7 names.

So what does this build out actually look like long term?

I think we've been seeing projections that reach around $7 trillion to $10 trillion reaching into 2030.

On the fact that data centers, we look at all this cap, there's no real way to hedge this cost or this exposure, and we think a futures market for compute is probably the best way to hedge this exposure.

And finally, before I let you go, I do want to get an understanding of how you're looking at growth.

Your organization, I think the team right now is 18 of us total, and we're looking to expand the team to around 30 to 50 people coming in next year.

We've honestly tripled our team size in the last 6 months.

It's funny, the company started 8 months ago and we've already reached such a large headcount and our revenue numbers are just continuing to increase.

Well, Christian, it was great having you here at the New York Stock Exchange.

Thank you so much for joining us, and I look forward to talking to you again for having me.

Thank you so much.