NEW YORK — September 2, 2026 — Wall Street is preparing to financialize one of the most valuable resources of the artificial-intelligence era: computing power.
The global race to build artificial intelligence has created enormous demand for graphics processing units, data centers and electricity.
Now, the financial industry is taking the next step.
Rather than viewing advanced AI processors solely as technology equipment, financial institutions are increasingly treating access to GPU computing capacity — commonly referred to simply as compute — as an emerging asset class.
CME Group, one of the world’s largest derivatives marketplaces, is preparing to launch futures contracts tied directly to the rental prices of Nvidia GPUs.
The products are scheduled to begin trading on October 5, 2026, pending final regulatory review.
The development could represent an important milestone in the transformation of artificial-intelligence infrastructure into a full-scale financial market.
Nvidia H100 and Blackwell B200 Enter the Futures Market
CME Group and Silicon Data plan to introduce two initial compute futures products.
One will track rental prices for Nvidia’s widely used H100 GPU.
The other will track Nvidia’s newer Blackwell B200 architecture.
Rather than investors purchasing physical GPUs, the contracts will track indexes measuring the hourly cost of renting computing capacity.
Each futures contract will represent approximately one month of GPU rental exposure.
The structure creates something the AI industry has historically lacked:
a standardized forward price for computing power.
That is extremely important for companies building AI infrastructure.
AI laboratories, cloud providers and data-center operators frequently sign enormous contracts for computing capacity, but GPU prices can change dramatically as new processors are introduced and supply-demand conditions evolve.
Compute futures could allow companies to hedge that risk.
Compute Is Becoming the “Oil” of Artificial Intelligence
The concept is increasingly being compared with commodity markets.
Oil producers and airlines use futures markets to manage energy-price risk.
Farmers and food companies use derivatives to manage agricultural commodity prices.
Banks and investors use interest-rate futures to manage financing exposure.
Now AI companies may eventually use compute futures to protect themselves against changes in the price of processing power.
CME Group CEO Terry Duffy has described compute as the “new oil of the 21st century.”
The comparison reflects the increasingly central role computing capacity plays in the digital economy.
Every major AI model requires compute.
Every chatbot query requires compute.
Training increasingly sophisticated AI systems requires enormous quantities of compute.
And the global infrastructure required to supply that computing power is becoming one of the largest investment opportunities in modern technology.
ICE Is Building a Competing Compute Market
CME is not alone.
Intercontinental Exchange, the company behind the New York Stock Exchange and one of the world’s largest financial-market infrastructure providers, has also announced plans to develop GPU compute futures.
ICE has partnered with technology companies including Ornn and NATIVX to create products designed to track changing prices for AI computing capacity.
One approach tracks spot prices for GPU compute across major hardware categories.
Another seeks to create an energy-normalized compute benchmark linking the cost of AI processing with electricity consumption.
The entry of both CME and ICE indicates that compute derivatives are moving beyond an experimental concept.
Two of the most important derivatives-market operators in global finance now see potential demand for standardized AI-compute pricing.
A $3.6 Trillion Infrastructure Opportunity
The financial opportunity behind this new market could be enormous.
Boston Consulting Group analysts have estimated that approximately $3.6 trillion of AI infrastructure investment could be deployed between 2026 and 2030.
Much of that capital will be required to finance GPUs, networking equipment, data centers, power generation and supporting infrastructure.
Financing that equipment introduces a new challenge.
AI chips can cost tens of thousands of dollars each, while enormous data centers can contain tens or hundreds of thousands of processors.
But technological progress is extremely rapid.
A powerful chip today may become significantly less valuable several years later when a new generation of processors enters the market.
That creates uncertainty for lenders.
A standardized futures market could provide a mechanism for managing some of that residual-value and rental-price risk.
BCG estimates cited by Reuters Breakingviews suggest a reliable forward pricing curve for compute could potentially reduce total borrowing costs associated with AI infrastructure financing by approximately $116 billion between 2026 and 2030.
If that occurs, derivatives could become an important financial engine behind the next generation of AI data centers.
Nvidia Sits at the Center of the New Market
The biggest potential beneficiary may be Nvidia.
The company already dominates the market for advanced AI accelerators.
Its chips form the backbone of many of the world’s largest artificial-intelligence systems and data centers.
According to Nvidia CEO Jensen Huang, processors can represent roughly 60% of the cost of an AI-focused data center.
That means the financial value of the computing equipment inside AI infrastructure can be enormous.
If futures contracts based primarily on Nvidia processors become widely adopted, Nvidia hardware could evolve beyond being merely a semiconductor product.
Its GPUs could become benchmarks for pricing computing power itself.
The comparison with commodities is significant.
West Texas Intermediate and Brent crude provide reference prices for global oil.
Nvidia H100, B200 and future architectures could potentially become reference instruments for the cost of artificial-intelligence computing.
BlackRock, Goldman Sachs and Private Capital Move Into AI Compute
The emergence of compute futures is happening alongside a much larger wave of institutional investment.
Nvidia recently announced partnerships with some of the most powerful investment firms in the world to develop financing platforms for AI compute infrastructure.
Participants include BlackRock, Goldman Sachs, Apollo, Blackstone, Brookfield and KKR.
The initiative aims to mobilize more than $500 billion in third-party capital over time toward the construction and financing of AI infrastructure.
The strategy is designed to transform Nvidia-based computing infrastructure into an investable financial asset capable of attracting institutional capital.
That represents a major evolution in the AI investment cycle.
The first stage of the artificial-intelligence boom primarily rewarded semiconductor companies.
The second stage drove enormous capital expenditure into data centers.
The next phase could increasingly involve Wall Street creating financial products around the infrastructure itself.
GPUs Are Becoming Financial Collateral
Another major development is the use of graphics processors as collateral.
Some lenders have already provided financing secured against GPU infrastructure.
The concept is similar to other forms of asset-backed lending.
A company purchases expensive computing equipment.
The hardware generates revenue through AI workloads or rental contracts.
Lenders provide financing against the value and future cash flows generated by that equipment.
But there is a major problem.
The future value of GPUs is difficult to predict.
Nvidia releases more powerful hardware generations at a rapid pace.
A processor commanding premium rental prices today may become substantially cheaper when newer chips become available.
Compute futures could help lenders hedge that risk.
This could make GPU-backed financing more attractive and potentially unlock significantly larger pools of institutional capital.
AI Infrastructure Becomes an Investable Ecosystem
The transformation is creating a new investment ecosystem around artificial intelligence.
Investors are no longer looking exclusively at companies developing AI models.
Capital is increasingly moving toward the entire infrastructure stack.
That includes semiconductors, GPU financing, cloud computing, data centers, electricity generation, transmission networks and cooling infrastructure.
The financialization of compute could create additional investment products on top of those physical assets.
Futures may eventually be followed by more sophisticated derivatives, structured products, lending markets and investment funds.
The result could be the emergence of an entirely new category within global capital markets.
The Risks Remain Significant
Turning computing power into a standardized financial instrument will not be simple.
Unlike a barrel of oil, one unit of GPU compute is not necessarily identical to another.
Different Nvidia processors have different performance characteristics.
The location of a data center affects electricity prices.
Network connectivity affects performance.
Cooling infrastructure affects efficiency.
And the price of compute can vary substantially across geographic markets.
Competition is another risk.
Google, Amazon and other technology companies are developing proprietary AI processors, while AMD and other semiconductor manufacturers continue to challenge Nvidia.
If multiple chip architectures gain significant adoption, the compute market could fragment across competing benchmarks.
AI Meets Wall Street
Despite those challenges, the direction of travel is becoming clear.
Artificial intelligence is creating not only a technology revolution but a financial one.
Trillions of dollars will be required to construct the infrastructure needed to support increasingly powerful AI systems.
Wall Street wants to finance that expansion.
To do so efficiently, financial institutions need standardized prices, reliable collateral values and tools for managing risk.
Compute futures could provide part of that infrastructure.
For Nvidia, the development could be especially significant.
Its processors already dominate AI computing.
If Nvidia GPU rental rates become globally recognized financial benchmarks, the company may occupy an unprecedented position at the intersection of technology and capital markets.
Nvidia would no longer simply sell the machines powering artificial intelligence.
Its hardware could increasingly help define the market price of artificial intelligence itself.
The AI investment boom has already transformed semiconductor markets.
Now it is beginning to transform derivatives, lending and institutional finance.
And if Wall Street succeeds in turning compute into a globally traded asset class, the next major commodity market may not involve something extracted from the ground.
It may exist inside a data center.