AI Compute Futures

Compute is the defining commodity of the 21st century. Global spending on AI infrastructure is expected to exceed $500B annually by 2028, yet the market for procuring and pricing compute remains fragmented, opaque, and bilaterally negotiated. There is no standardized unit of trade, no transparent pricing benchmark, and no reliable mechanism for future delivery.

Corn pit on the trading floor at the Chicago Board of Trade in 1988 (Greg Wass/Creative Commons)

Before we dive into what AIC is building and why, a little history is required on commodities and futures, and when the breakout moments supercharge these commodities. Let’s go back to 1983, and see the various breakout moments for oil futures and energy markets.

Global demand for compute is accelerating at an unprecedented pace, driven by AI, large-scale model training, and the expansion of inference across every sector.

Meeting this demand requires massive investment in data centers, GPUs, networking, and power. McKinsey estimates roughly $7T in cumulative global data center capex through 2030, forming the underlying infrastructure AIC contracts will clear against.¹

Despite this scale, the compute market remains structurally broken. Capacity is concentrated, pricing is opaque, and access is constrained. There is no unified mechanism for price discovery, nor an efficient way to hedge, trade, or gain exposure to compute.

In every prior commodity cycle, once a resource reaches sufficient scale and strategic importance, financial markets emerge around it. Oil, power, and metals all followed this path, evolving into deeply liquid markets where derivatives volumes exceed the underlying tens of times over.

Compute is now reaching that moment.

AIC is built for this transition, introducing the financial layer that enables price discovery, liquidity, and risk transfer, transforming compute from a constrained resource into a tradable asset class, and unlocking a market that scales to many multiples of the underlying infrastructure.

Futures unlock the financial layer, allowing participants to hedge risk, and trade exposure to an asset without owning it.

Units of the commodity can be shorted without owning the underlying and as long as it’s closed before delivery it is cash settled. This means that the commodity’s financial activity can grow far beyond its physical supply.

Applied to compute, this means GPU capacity and compute resources can be priced, hedged, and traded like any other commodity, transforming a constrained infrastructure market into a liquid financial one.

In mature commodity markets, financial volume scales to tens of multiples of the underlying physical supply. If deliverable compute reaches $1T annually by 2030 under central demand scenarios, applying conservative commodity-market multipliers implies a futures complex in the $10–25T annual notional range.

The key insight is that liquidity does not require widespread physical delivery, it requires credible physical delivery. The ability to settle against the real world provides true price discovery and financialization.

Commodity markets financialize when producers and consumers need to hedge.

Compute is now at that point: multi-billion-dollar infrastructure, multi-year deployment cycles, supply bottlenecks in power and interconnection, and buyers committing hundreds of billions, soon to trillions, against uncertain future pricing.

At the same time, demand is accelerating rapidly with the rise of AI, while supply remains constrained by physical infrastructure. This combination of scale, capital intensity, and price uncertainty is what has historically triggered the emergence of futures markets.

The conditions are already here. AIC is built for this moment, applying financial market structure, standardization, and clearing to a market that has reached the threshold but lacks the mechanism to scale.

Successful industry standard contracts often look niche at launch and obvious in retrospect. West Texas Intermediate (WTI), now the dominant U.S. oil benchmark, achieved that position only after physical delivery standards, market depth, and hedging behavior.

The same pattern emerges whenever a fragmented physical market becomes standardized and centrally priced.

The launch environment for AIC is much more favorable than the past. The underlying market already exists at scale, hyperscalers are investing hundreds of billions in capacity, NVIDIA’s data center business has reached massive scale, and enterprises are committing significant budgets to AI workloads. Taken together, this represents one of the largest coordinated infrastructure buildouts in human history, on par with the interstate highway system and the Apollo program combined.

Yet pricing remains fragmented and opaque, unlike other tradeable commodities that developed efficient multi-sided markets.

AIC is a multi-sided market connecting supply, demand, and financial participants.

On the supply side are data centers, GPU clouds, and infrastructure providers seeking better utilization, forward demand visibility and financing. On the demand side are AI labs, enterprises, and model builders that need reliable access to compute at predictable terms. A third group, market makers, speculators, and other financial participants, provides liquidity and enables risk transfer.

Each participant already exists and faces a clear constraint: fragmented access, opaque pricing, and no mechanism to hedge or trade exposure. AIC brings these participants into a single, standardized market.

The AIC platform is built as a vertically integrated system combining a derivatives exchange, a clearing interface, and a physical compute delivery network.

AIC Margined Market

A fully margined, less than 28 day market, that is physically settled with a central limit order book, real-time margining, and standardized contracts tied to GPU hardware and a normalized compute unit.

AIC CFTC Licensed Exchange

Expansion into a regulated futures exchange with full forward pricing, integrated clearing, and institutional participation.

Grand Central

A unified delivery layer that delivers compute in an agreed upon standardized unit of work to contract holders, enabling compute to be delivered as a commodity.

AIC’s product and technical architecture establish the system for pricing, trading, and delivering compute. Full technical architecture is included in the appendix.

AIC is designed to follow a staged path from initial market launch to full regulatory designation.

In February 2026, Xchange Alpha secured CFTC designation as a Designated Contract Market (DCM) in just 204 days, the fastest approval on record.¹⁹ This shows that exchange licensing can move quickly when built on credible systems, strong preparation, and alignment with established regulatory frameworks.

We will initially launch under existing short-term contract exemptions with a physically settled, margined market, and transition to a fully regulated exchange upon CFTC designation, establishing AIC as an institutionally recognized venue.

AIC’s go-to-market begins by solving a hard operational problem for real buyers: securing trusted compute from fragmented suppliers. Rather than launching an exchange and waiting for liquidity, we will build the market on top of real usage. Drawing on deep experience in developing highly liquid futures markets, AIC follows a structured path:

AIC will develop its market by onboarding participants in parallel, with each group engaged through tailored entry points.

Buyers

AI labs, model builders, and enterprise teams already face fragmented supply, uncertain availability, and limited cost visibility. They are engaged through direct sales and integration into procurement workflows, providing reliable access to compute, improved pricing transparency, and trusted execution.

Sellers

Data centers and GPU providers are onboarded through a curated process that prioritizes quality and reliability. AIC offers a new demand channel, improved utilization, and the ability to hedge future offtake, establishing confidence in delivery from the outset.

Financial Participants

Market makers and traders are introduced as the market develops sufficient depth in demand, supply, and transaction data. Their participation enables liquidity provision, price discovery, and risk transfer.

Trading & Clearing

Transaction and clearing fees from futures markets are high-margin, volume-driven, and scale with liquidity. CME generated $6.52B in revenue in 2025, supported by record average daily volume of 28.1M contracts and a 64.9% operating margin.

Market Data & Benchmarks

In mature exchange businesses pricing data sets, reference rates, and analytics are some of the most valuable revenue streams. At ICE, Fixed Income & Data Services generated $2.3B in revenue, representing roughly a quarter of the business, while at LSEG, Data & Analytics accounted for 43% of total income with high margins.

Settlement & Delivery Infrastructure

This is where compute is actually delivered into real workloads. It behaves more like neutral infrastructure than a traditional exchange. Equinix generated $9.217B in revenue in 2025 with a 49% adjusted EBITDA margin, providing a strong analog with the exception that we see a future where trillions of dollars of compute will flow through this infrastructure.

AIC is led by founders operating at the intersection of large-scale compute infrastructure and financial market design.

The compute market today resembles many commodities before their financialization: fragmented, brokered, and lacking standardized contracts, centralized clearing, and true price discovery.

Current participants fall into four categories but none create a true futures market:

This round enables AIC to build the final layer of the market: making compute tradeable, deliverable, and financialized.

Capital will be deployed across core product and infrastructure, team expansion, regulatory execution, and early supply and demand. The objective is to build the product and gain DCM designation.

Compute is becoming humanity’s foundational commodity, but an efficient market around it has not yet formed. Pricing is opaque, access is fragmented, and there is no standardized mechanism for forward delivery or risk transfer.

AIC builds the missing layer. By combining physical delivery, standardized contracts, and centralized market structure, it enables compute to be priced, traded, and settled like every other critical commodity.

We understand how to scale systems, operate in regulated environments, and bring new market structures into existence. But this moment is different and is happening now. The shift underway is redefining how intelligence itself is created and deployed, and compute sits at the center of it.

AIC participates in the growth of AI, by defining the system through which compute is accessed, priced, and traded.

Contact: GS@AIComputeFutures.com