Architect Plans First Exchange-Traded Futures Linked to AI Compute Pricing

Architect aims to turn AI compute costs into a tradeable market using regulated, perpetual futures for institutions.

Architect Financial Technologies is preparing to enter a new segment of financial markets tied to artificial intelligence infrastructure. Rising demand for computing power has pushed GPU and memory costs into the spotlight as major risks for AI-focused businesses.

To address that shift, the firm plans to launch derivatives contracts that allow institutions to trade and hedge compute pricing in a regulated setting.

Architect Develops First Exchange-Traded Futures Tied to Compute Pricing

Architect Financial Technologies, founded by former FTX US president Brett Harrison, said its institutional exchange AX will soon add perpetual futures linked to compute pricing. 

Contracts will track daily rental prices for GPUs used for AI training and inference, as well as dynamic pricing for random-access memory. He also added that launch timing depends on regulatory approval, according to a statement shared with The Block.

According to the firm, these products would mark the first exchange-traded futures contracts tied directly to compute. Trading will take place on AX, a Bermuda-regulated platform that serves non-U.S. institutional clients. 

Additionally, positions will be marginable and can be funded with U.S. dollars or dollar-backed stablecoins. Compute pricing benchmarks for the contracts come from Ornn Data. For clarity, Ornn builds indices using live transaction data from several spot GPU markets. 

Chief executive Kush Bavaria mentioned that HydraHost is the largest data contributor through its GPU brokering activity. Data reflects real trades rather than quoted estimates, aiming to give clear price signals in a fast-changing market.

Target users include AI labs, data center operators, lenders, and hardware vendors. These groups face sharp swings in GPU prices as demand for training and inference grows. Architect revealed that it is in talks with market makers to support liquidity at launch, as well as with compute buyers and data centers looking for hedging tools.

AI Compute Pricing Draws Interest as Firms Seek Hedging Tools

Growing spending on AI infrastructure has created pressure for clearer pricing and risk management. Harrison said companies now need standard contracts and centralized order books to manage exposure tied to computing costs, which have become a major capital expense.

Key features of the planned compute futures include:

• Perpetual contracts with no fixed expiry date.
• Daily price tracking for GPU rental and memory usage.
• Margin trading supported by cash or dollar-based stablecoins.
• Benchmarks based on live transaction data.
• Access limited to non-U.S. institutional participants.

Bavaria described compute as a core commodity for the AI economy. He admitted that limited transparency has made it hard for institutions to manage risk. Ornn’s indices aim to support price discovery and give buyers and sellers clearer reference points.

According to Bavaria, AI labs need stable pricing rather than sudden GPU market swings. Infrastructure firms holding large GPU inventories also lack ways to hedge depreciation. He added that traditional commodity traders are increasingly watching compute markets due to limited transparency and growing demand.

Architect Extends Crypto-Style Perpetual Futures to Real-World Assets

Architect’s plans build on a broader effort to apply crypto-style perpetual futures to traditional and emerging assets. AX began offering perpetual contracts tied to equities and foreign exchange in October. The platform does not list cryptocurrency-linked derivatives, focusing instead on real-world assets under regulated structures.

Because U.S. rules restrict perpetual futures tied to traditional assets, AX operates through a Bermuda subsidiary and serves overseas institutions. The structure enables continuous trading while remaining within regulatory boundaries.

Architect is backed by investors including Coinbase Ventures, Circle Ventures, and SALT Fund. The company raised $35 million in a Series A round in late 2025, led by Miami International Holdings and Tioga Capital. 

Funding followed earlier raises of $5 million in 2023 and $12 million in 2024. And this brought the total valuation to about $187 million, according to a source familiar with the matter.

Market context around the expansion includes:

• Rapid growth in AI-related capital spending.
• Rising demand for GPU capacity across sectors.
• Limited hedging options for compute infrastructure.
• Interest from institutional traders beyond crypto.
• Ongoing regulatory review of derivative structures.

Harrison, who left FTX US in September 2022, has positioned Architect as a bridge between crypto derivatives mechanics and regulated global finance. As interest widens beyond digital assets, the firm is betting that compute pricing will become a tradable market in its own right.